101
Copyright © 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271 JOURNAL OF THE TECHNICAL UNIVERSITY AT PLOVDIV, BULGARIA FUNDAMENTAL SCIENCES AND APPLICATIONS VOL. 13 (3) 2006 ANNIVERSARY SCIENTIFIC CONFERENCE 2006 THE SCIENTIFIC REPORTS Electronics

JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

  • Upload
    others

  • View
    4

  • Download
    0

Embed Size (px)

Citation preview

Page 1: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright © 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

JOURNAL

OF THE TECHNICAL UNIVERSITY AT

PLOVDIV, BULGARIA

FUNDAMENTAL SCIENCES AND APPLICATIONS

VOL. 13 (3) 2006

ANNIVERSARY SCIENTIFIC CONFERENCE 2006

THE SCIENTIFIC REPORTS

Electronics

Page 2: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 2 -

ORGANIZING COMMITTEE Conference Co-chairs: Prof. Kamen Veselinov,

PhD Rector of TU - Sofia

Prof. Dimitar Katsov, PhD Director of the Plovdiv Branch of TU - Sofia

Members: Prof. DSc Vassili Loumos Greece Prof. DSc Mark Himbert France Prof. DSc Panayiotis Frangos Greece Prof. DSc Reinhart Verschoore Belgium Prof. DSc Yasser Alayly France Prof. Dr. Dr.h.c.mult. Uwe Heisel Germany Acad. Prof. DSc Yuri Kuznetsov Ukraine Prof. DSc Alexander Tsiganenko Russia Prof. DSc Victor Baranov Russia Prof. DSc Edward Evreinov Russia Prof. DSc Okyay Kaynak Turkey Acad. Prof. DSc Kiril Boianov Bulgaria Corr. Memb. Prof. DSc Mincho Hadjiski Bulgaria Corr. Memb. Prof. DSc Petko Petkov Bulgaria Prof. DSc George Popov Bulgaria Prof. DSc Marin Nenchev Bulgaria Prof. DSc Mincho Minchev Bulgaria Prof. Angel Vachev, PhD Bulgaria Prof. George Stoianov, PhD Bulgaria Prof. Drumi Bainov, PhD Bulgaria Assoc. Prof. Pepo Yordanov, PhD Bulgaria Assoc. Prof. Valentin Kirchev, PhD Bulgaria Assoc. Prof. Kostadin Iliev, PhD Bulgaria Assoc. Prof. Valyo Nikolov, PhD Bulgaria Assoc. Prof. Peyo Stoilov, PhD Bulgaria

Page 3: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 3 -

CONTENTS Page

JEAN-PIERRE WALLERAND, MARC HINBERT, SUAT TOPCU, YASSER ALAYLI, MARGARITA DENEVA, MARIN NENCHEV

DEVELOPMENT OF ORIGINAL TECHNIQUES FOR REALISATION AND STUDY OF THE

PECULIARITIES OF TWO-WAVELENGTH CW DIODE - PUMPED Yb-DOPED CRISTAL AND

RED SEMICONDUCTOR LASERS ……………………………………………………………………..… 5

PLAMENA TODOROVA, MARGARITA DENEVA, MARIN NENCHEV

EFFECTIVE TWO COAXIAL ACTIVE MEDIA Nd3+:YAG – DYE FLASH LAMP PUMPED

LASER ……………………………………………………………………………………………………….. 23

ANTONIY PETROV, BOYKO PETROV OPERATOR IDENTIFICATION USING UNIVERSAL MODULE ….……….……………………..… 33

SVETOSLAV IVANOV, ELENA KOSTOVA, MINCHO PEEV

PULSE CHARGE OF RECHARGEABLE BATTERIES WITH THE ENERGY HEAPED IN A

PARALLEL RESONANCE SET …………………………………...……………………………………… 39

IVAN NEYCHEV, PETAR GYOSHEV, RUMEN KAROV FLYBACK DC-DC CONVERTER WITH IMPROVED PERFORMANCES.…………………………. 47

RUMEN KAROV, PENKA LENTCHEVA, DANIEL KAROV

CONNECTION OF INVERTERS TO JOINT LOAD ………….…………………………………...…… 55

NELI VARDEVA, GALIDIA PETROVA

STAND-ALONE MODULE FOR MULTICHANEL EEG SIGNAL DIGITAL FILTTERING USING

ADSP-21065L …………………………………...……………………………………………………………. 63 ALEKSANDAR GEORGIEV, RUMEN POPOV, MARIN PAMUKCHIEV, KLIMENT PETKOV SOFTWARE DEVELOPMENT OF SUN TRACKER REAL TIME CONTROL ALGORITHM….… 71

TSVETANA GRIGOROVA – SHTARBEVA

A UNIFIED ANALYSIS OF A SERIES RESONANT CONVERTER ………………………………….. 79

TSVETANA GRIGOROVA – SHTARBEVA SOFT SWITCHING CAPACITOR VOLTAGE CLAMPED SERIES RESONANT DC/DC

CONVERTER ……………………...…………………………………………………………………….…. 87

KATYA ASPARUHOVA, TSVETANA GRIGOROVA, SNEJANA TERZIEVA

MATLAB/SIMULINK IGBT MODEL ...………………………………………………………………….. 97

Page 4: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 4 -

Page 5: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 5 -

Journal of the Technical University at Plovdiv “Fundamental Sciences and Applications”, Vol. 13(3), 2006 Anniversary Scientific Conference’ 2006 BULGARIA

DEVELOPMENT OF ORIGINAL TECHNIQUES FOR

REALISATION AND STUDY OF THE PECULIARITIES OF

TWO-WAVELENGTH CW DIODE - PUMPED Yb-DOPED

CRISTAL AND RED SEMICONDUCTOR LASERS

JEAN-PIERRE WALLERAND, MARC HIMBERT, SUAT TOPCU, YASER ALAYLI, MARGARITA DENEVA, MARIN NENCHEV

Abstract: We present original solutions of CW diode-pumped Yb:doped crystal -Yb:GGG and Yb:YAG (first realizations in our knowledge) and red semiconductor lasers that produce simultaneously in a single beam or in separate beams emissions at two

independently tunable wavelengths (in the experiments in the ranges 1.02 - 1.06 µm ;

0.628 -0.632 µm). We describe and analyze simple and effective interference-wedge based schemes for producing such mode of laser operation. The low-loss passive-self injection locking controls using such type of resonators are presented. The peculiarities of the two-wavelength emission are studied both theoretically and experimentally. The advantages in comparison with use of two separate lasers (low necessary pump power for producing of the two emissions, high efficiency, and output in a single beam) and limits, due to the wavelength competition for such operation, are discussed. On the base of the investigations, the cases of competitive applications of the system developed are carried out.

Key words: tunable lasers, CW operation, Yb-doped crystal lasers, red semiconductor lasers, diode pumping

РАЗВИТИЕ НА ОРИГИНАЛНИ ТЕХНИКИ ЗА РЕАЛИЗАЦИЯ И ИЗСЛЕДВАНИЯ ВЪРХУ ОСОБЕНОСТИТЕ НА ДВУВЪЛНОВИ

НЕПРЕКЪСНАТИ ДИОДНО-ВЪЗБУЖДАНИ ИТЕРБИЙ-АКТИВИРАНИ КРИСТАЛНИ И ИЗЛЪЧВАЩИ В ЧЕРВЕНАТА СПЕКТРАЛНА ОБЛАСТ

ПОЛУПРОВОДНИКОВИ ЛАЗЕРИ

1. Introduction

The two-wavelength lasers and what justify the interest to them. Aim of the work.

The natural possibility of the wide gain lasers is to be operated simultaneously and in controlled manner at two or more independently tunable wavelengths in essentially different spectral intervals. This mode of operation (to-wavelength lasers) is earlier realized and studied in details for dye [e.g.1-3], F-color centers and Ti:Saphhire lasers [4-6]. More recently, by applying our original and effective schemes of two-channel laser resonators, we have realized

Page 6: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 6 -

and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]. The Yb-dopped crystal lasers are very convenient for CW diode pumping and with very low energetic difference between the transient and upper laser levels (and respectively very low conversion losses). In addition, it provides the tunable emission in very interesting spectral range of 1.02 – 1.06 µm, that permits after frequency doubling coincidence with the reference iodine lines (except the other applications, discussed below). The red semiconductor lasers are promising simplest sources for applications in spectroscopy, in metrology end in holography.

As it is well known, many practical applications of the lasers relate of the use of laser light at two independently tunable wavelengths [e.g. 2,]. We can note here, as a first example, the important laser technique called Differential Absorption Spectroscopy (DAS) that permits to monitor different composants of the complex substances including the atmosphere pollutants such as NO2, SO2 (the known its LIDAR variant). In this application, the light at one of the wavelength is tuned at the absorption maximum of the corposant under monitoring and the second – outside, at the absorption minimum. By comparing the powers of the two received light that passes or are scattered by the substance studied it is possible to obtain information about the presence and the concentration of the composant of interest. The laser light at two wavelengths is used in the modern and efficient method of laser isotopes separations, in nonlinear frequency mixing experiments to produce a sum and difference of wavelengths, in holography, in metrology (with one emission fixed as a reference line and the other scanned) . The best situation for such applications is when the two emissions are produced in a single beam and simultaneously. In some cases, the two wavelength lights can be in closely disposed parallel beams and emitted in nano- or micro-second time length sequences.

As it was noted above, an approach to solve this laser-technique problem is to operate one wide-gain laser at two independently controlled wavelengths. The second - trivial technique, to obtain laser light in a single beam at two independently tuned wavelengths is to superimpose the emissions of two separate tunable lasers using different type of amplitude or polarization beam splitters. The two noted above manners to obtain two-wavelength laser light present their advantages and disadvantages. Below, briefly, we will discuss the last questions. This is important to justify the active development, also actually of the two-wavelength lasers. Our discussion will be especially related with the two-wavelength CW diode pumped Yb-dopped crystal and CW semiconductor lasers.

1.1. Why the two wavelength lasers are of interest - their advantages in

comparison of the use of two separate lasers? Firstly, in technical aspects, what is evident, this is the low price to construct one laser

at the place of two separate lasers. The typical complex CW Yb-dopped crystal laser system contains as an essential parts a pump high power (~ 10 and more Wats) semiconductor laser(s) with respectively high price, a high price special dichroic input mirror, AR coated pump focusing optics, high price AR coated high quality Yb-doped crystal with stabilized cooled subsystem. The doubling of these elements in two separate lasers doubled also their high price. The analogical technical questions attend the realization of two independent external cavities semiconductor lasers.

Secondly, in basic aspects, there are two essential advantages:

� Being the threshold phenomenon, the laser operation needs of some minimal power to be produced. The important moment for the two wavelengths, a single-beam emission laser is that the emissions at the two wavelengths can be produced simultaneously by pumping the laser with the threshold power for single wavelength operation. Only the laser output power is devised between the two emissions. Consider a practical example. Let we have two separate CW diode pumped Yb:GGG lasers with threshold pump power of 3 W each. Let we

Page 7: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 7 -

have also a pump laser diode with the output power of 5 W. Evidently, in this situation we cannot obtain laser emissions from the two lasers. Contrary, if we construct the two-channel resonator laser, that produces the emissions at the two-wavelengths in the single pumped Yb:GGG crystal volume, we can obtain easily a notably powerful laser emissions at the two-wavelength.

� As it is known [10], the CW laser output power is proportional to the difference (Ppump/Pth-1), where Ppump and Pth are the pump and the threshold power, respectively. This simple relation shows directly that the output power in the each emission for the two-wavelength laser is a few times higher that this one obtained if two separate lasers are used by pumping of the same pump source. As an example let our diode pump system can assure an 8 Watts pump power (high quality and price pump laser). Lets our Yb:GGG lasers are with 3 W threshold. Thus, it is very easy to estimate that in this practically real situation the two wavelength laser can emits a light at each wavelength with about 3 times higher power that these ones, produced by the two separate lasers.

1.2. About the disadvantage of the two-wavelength mode of operation. The disadvantage for the most interesting case of generation in one beams of the two

emissions is relate with so called mode-competition effects [10], very essential when the laser active medium is homogeneously widened. The two generations, produced at a single laser transition compete nonlinearly and the small unbalance of the losses produces suppression one of the generations by the other. This limits the tuning of the generations in the two-wavelength laser in a relatively small spectral range (few end more times lower than for the separate lasers.

Having en essential experience [3,4,7-9 ] on the techniques of two-wavelength laser realization and on theoretical and experimental studies of the two-wavelength laser operation we will present and discus in the rapport, on the basis of our proper results [7-9], the two-wavelength CW diode pumped Yb-doped crystal and CW red semiconductor lasers, developed in the last years. We will concern the developed original techniques of realization, the approach and results of theoretical study and the results of real laser operation. Our practical experience concern the Yb:GGG and Yb:YAG crystal lasers. The laser operation in these two active media do not differ essentially.

2. New solutions of the two-wavelength lasers using an original two-spectral-

selective-channels resonators (application to CW diode-pumped Yb:YAG AND CW red

diode lasers).

The cases of CW diode-pumped Yb:YAG laser is typical for operation with pumping only a few times exceeding the lasing threshold. The laser system used in our investigation is with typical characteristics for this type of lasers. The pumping is at 970 nm with a diode

system with a nominal pump power of ~ 8 W (maximum power of 10 W) and with a lasing threshold for the nonselective cavity of the Yb:YAG with a threshold of ~ 3 Watts (given by the quality of the laser crystal). The red diode laser permits also to increase the pump current near two times (threshold of ~ 60 mA, nominal laser operation at 80-90 mA and maximum pumping to 100-110 mA). In our consideration we have used a red AR coated front surface

Fig.1. Calculated transmission and reflection of the Interference Wedge versus the wavelength (Ref. 4)

Page 8: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 8 -

laser diode of AMS Technology (model SAL-635), optimized for wavelength of 635 nm. Especially, the Yb:YAG laser was very sensitive to additional inserted losses. The detailed description of the realized by us laser systems are given in the published our articles [7-9]. Here, as a first point, we will generalize the schemes of the producing of two-wavelength emission in the discussed lasers.

The main part of the realized schemes are related to the basic idea to use in convenient manner an Interference Wedge (IW, or Fizeau Interferometer) as a spectral selector and tuning – coupling element for creation two spectral-selective channels laser resonators. Very important property of the IW is that the wavelength of its transmission can be tuned by a simple translation in its plane. This permits, during the tuning, to conserve the direction of the reflection and the transmission of the incident beams. Thus, using in convenient manner a series of Interference Wedges we can realized a very attractive and compact Wavelength

Devising and Multiplexing system (WDM) with spectrally tunable input-output. Except the application for creation of the two-wavelength laser resonator, this WDM system presents essential potential for applications in optical communication systems. Its basic principle is published earlier by us [11] and is developed recently by the authors of the present work in direction of the noted important applications in the lasers and as potential as WDM device.

The IW is optical element, in construction very similar to the well-known Fabry-Perot Interferometer (FPI). The difference is that in IW the reflective surfaces (e.g. multi-dielectric mirrors or layers) consists a small angle of ~ 10-5 rad. The properties of this device

for laser applications, when the IW is illuminated with a limited diameter beams such the intracavity beams in the laser pumped lasers is studied theoretically and experimentally by us in Ref. 12, 13. In general, they are similar to these ones of the FPI with variable thickness by translation of the IW. The resonant transmission for the illuminating beams with a diameters of tens parts and lower ( ~ 1 mm, 200 µm) of the distance between the lines of equal declination can reach 0.8 -0.9 and more in combination with large tunability (~ tens nm, depending on the wedge thickness) and transmission line-width of parts of nm and less. The IW presents also the important peculiarities that will be not discussed here [13]. The essential practical interest presents the technology, actually under progress, to fabricate the very compact IW of “sandwich” type that presents a sequences of disposed on an optical support sequences of a multi-dielectric reflective layer (reflection Rm of ~0.8-0.9), a transparent ~ few µm layer-separator and a second reflective layer, similar to the firs one. The example of the calculated

Fig.2. The measured transmission of the

interference wedge used in the work for λ=633nm. The experimental points are plotted on the curve.

0

0,25

0,5

0,75

1

0 2 4 6 8 10IW translation, mm

IW transmittion, ~

λ1

??2

?4?i

IW

IWIW

λ

λλ

λ

λ1

??2

?4?i

IW

IWIW

λλ

λ

IW

1

23

4

tun. λ4

tun. λ3

tun. λ2

tun. λ1

λ1… λi

Fig. 3. Principle scheme of the WDM system with Interference Wedges.

Page 9: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 9 -

transmission and reflection curves vs. wavelength is shown in Fig.1 for the IW with thickness of 5 µm, R=0.8 with angle of 3.10-5 rad is illuminated with 1 mm diameter He-Ne laser beam (at 630 nm). The experimentally measured dependence of the transmission for the last case vs. the IW translation is shown in Fig.2.

In Fig.3 are plotted the principle scheme of the discussed above WDM system that uses the series of Interference Wedges as the described this one.

Now, we realize and use successfully the two types of two-channel resonator schemes with IW. The first type is for application of low-gain lasers, where it is not possible to introduce in the resonator directly the selector due to the introduced losses. This is the case of the diode-pumped CW Yb:YAG. The laser cannot operate also for 10 W diode pumping when the described type of “sandwich” IW that we have with real resonant transmission of ~ 0.8 is introduced in the cavity. In the last case we applied co-called “passive self-injection technique” (PSIL) in combination with our two-wavelength IW structure [6,7]. We close the resonator, shown in Fig.4(a), in front of the two-wavelength selector (IW1-M3 and IW1 as an intermediate reflector or the no- resonant for it wavelengths-IW2 – M4) with partially transmissive mirror M2, thus forming an interference structures [7], each tunable by IW1 and IW2 translation respectively. As we have calculated

in Ref. 4 and Ref. 7, the reflectivity of such structure increases to be closely near 1, as can be seen from the Fig.5 (at the place of reflectivity without M2 of ~ 0.5). This type of system permits to obtain tunable laser emission with low gain lasers. Using this PSIL-IW resonator, we have obtained a laser operation at two independently tunable wavelengths, emitted in a single beam. The described two-wavelength laser scheme can be modified to produces the two-wavelength in just-imposed beams. To obtain such modification the two edged screens can be introduced conveniently in the resonator to screening the convenient half of each laser beams as it is shown in the Fig.4(b).

Fig. 4. Optical schemes of the two-wavelength cw Yb-doped crystal lasers. PSIL controlled cavities for single-beam emission at two wavelengths from (a) a single-lasing volume and (b) in parallel beams. (c) Cavity with an angular separation of the channels. M1, M3, plane mirrors; M1 is a dichroic mirror; M2, M4, M5, M6, concave mirrors; GP, the near-Brewster’s angle glass plate laser output; IW1 and IW2, interference

Page 10: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 10 -

The disadvantage of the described high efficient PSIL laser arrangement is the decreasing (to be half and less) of the tuning range also for the case of single-wavelength operation This is as results of competition with the non-selected lasing in the resonator closed by the mirror M2 when the tunable wavelength is at the low-gain wings of the gain curve. In addition, the selection outside the gain-maximum region can be accompanied with increasing of the background emission.

An addition, if the high transmission elements – FPI or IW with resonant transmissivity of ~ 0.95 and more are available, we can recommend the very simple and attractive scheme for two-wavelength operation shown in Fig.4(c). In this scheme, we obtain very easily, with non-precise alignment two-wavelength operation with different output for each generation. The two generations use also a single pumped volume in single crystal-cooled system-input dichroic mirror arrangement with concave mirror having a radius of curvature equal of its distance of the crystal, as can be shown in the figure.

When the laser produces a high gain that permits operation with introduced in the resonator relatively high losses IW-es, the preferred schemes is with the direct introduction of the IW-es two-wavelength structure as a close reflector of the laser resonator, as it is shown in Fig.6. The reasons were discussed in the previous point of the rapport. The noted last situation is realized with the red semiconductor laser, used by us.

Fig. 5. (a) Computed equivalent reflectivity Req of twowavelength interference structures M3–IW1–M5

and M3–IW1–IW2–M6 as a function of the wavelength and (b) transmission functions of the interference wedges for single pass TIW1,2 and double pass TIW1,2

2 for maximum wedge transmissions TIW1,2* = 0.7. The thick solid curves in (a) were calculated for the transmission maximum of each interference wedge TIW1,2* = 0.7 and for two values of R3, 0.92 and 0.98 (R3 = 0.92 corresponds to our experimental value. The thin solid curve in (a) represents the Req for high-quality IW with TIW1,2* 0.92 and for R3 =0.8. (Ref. 7)

IW1

IW2

tune λ1

tune λ2

low-power signaloutput at

low-power signaloutput at λ

2

laser output

1 +

2

M1

M2

M3

M4

λ1

λ2

AR front facediode laser

LIW1

IW2

low-power signaloutput at λ 1

low-power signalλ2

λ1

λ2

IW1

IW2

tune λ1

tune λ2

low-power signaloutput at

low-power signaloutput at λ

2

laser output

1 +

2

M1

M2

M3

M4

λ1

λ2

AR front facediode laser

LIW1

IW2

low-power signaloutput at λ 1

low-power signalλ2

λ1

λ2

Fig. 6. Optical scheme of two-wavelength diode-laser that uses a two-channel resonator with a single interference wedge as a tuning element and channels coupler. IW1 and IW2 are interference wedges, M1, M2, M3 and M4 – HR flat mirrors. (Ref. 12)

Page 11: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 11 -

For the 10 cm length nonselective resonator, composed of laser diode, AR coated 0.7 cm collimating lens the emission in parallel beams and 0.84 reflectivity (0.16 transmissivity) output coupler the threshold current was ~ 60 mA. The currant – output power curve for this resonator configuration is shown in Fig.7. The laser operates with the introduced 75% transmission IW1 and 95 % rerflectivity end mirror M1 with the threshold of ~ 80 mA. The output is taken from the reflection of the IW1 and is ~ 50 µW. When we reinjected the reflected beam in the opposite direction of the output with a second totally reflected mirror M2

, as it is shown in Fig.8, the laser output power output power reaches 250 µW for 90 mA pumping power. The last scheme can be considered as a variant of realization of PSIL technique. The laser can be tuned in both cases at ~ 5 nm around 630 nm with a linewidth of the emission of ~ 0.4 nm without PSIL and of ~ 0.7 nm with PSIL control. The high threshold of the parallel beams IW selection diode laser, evident is related with high diameter of the intra resonator beams, of order of 3-5 mm. We have improved the laser energetic characteristics essentially, using the modification of this scheme, shown in Fig.8 By replacing the diode-laser lens we focalized the diode emission on the flat 0.95% output mirror. The laser threshold rests near the same as for the parallel intra resonator beam case, however, as it was expected from the general considerations, the threshold with the same IW, placed near the output mirror, decreases to be ~ 64 mA. This is due to the narrowing of the beam diameters, illuminating the IW and respectively strongly increasing its transmissivity for the selected wavelength ( to ~ 85 %) [12]. The two-wavelength scheme can be easily understood from the shown Fig. 8

The described above two schemes are modified in the manner shown in Fig.9. Using a low reflectivity layer IW with a transmissivity between the resonant maximums of ~ 10 % (Fig.10) we can use simultaneously this IW as a tuning element, as a beam coupler and as an output coupler. Some problem is caused by the variation of the output coupling with the wavelength tuning that leads to strong variation of the ratio between the output powers at the two emissions.

Fig. 7. The red used diode laser output power (µW) versus the pump current (mA).

IW1

tune λ1

M1

M2

λ1

AR front facediode laser

LIW

tune λ1

laser output at λλλλ1

M1

M2

λ1

AR front facediode laser

LIW1

tune λ1

M1

M2

λ1

AR front facediode laser

LIW

tune λ1

laser output at λλλλ1

M1

M2

λ1

AR front facediode laser

L

Fig.8. An IW - PSIL effective tunable laser. M1, M2

– flat totally reflecting mirrors, IW1 – interference wedge.

Page 12: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 12 -

Fig.10. (a) The simpler for practical realization scheme variant, witch conserves the basic idea to use a single interference wedge as a tuning and coupling element. However, to form the second channel is formed using an 1800 line/mm diffraction grating in Littrow arrangement. (b) Schematic diagram of the transmission of low-reflectivity mirror interference wedge used in the scheme in Fig.9.

In all described schemes, we have obtained simultaneous two-wavelength laser operation, the main peculiarities that are described in Section 4 and more details in Ref. 7.

3. Approach for theoretical consideration of the cw two-wavelength operation of

the considered lasers and main results.

For the general case (pulse and cw), the two-wavelength multimode operation in a single pumped volume of the PSIL controlled DP Yb3+: laser can be described on the base of the rate equations system for quasi-three level laser schema [7]. The PSIL control can be included by use of the equivalent reflectivity Req of the end two-wavelength reflector in the schemes in Figs. 4(a) and 4(b). A similar system describes the action of the diode laser. For R3 = 0 the system describe the standard selection with intra-cavity selectors. The adapted by us set of rate equations is :

IW1

IW2

tune λ1

tune λ2

low-power signal

output at λ1

low-power signal

output at λ2

laser output

λ1 + λ2

M1

M2

AR front facediode-laser

L

L1

IW1

IW2

tune λ1

tune λ2

low-power signal

output at λ1

low-power signal

output at λ2

laser output

λ1 + λ2

M1

M2

AR front facediode-laser

L

L1

Fig.9. The two-channel resonator with focalization of the intra-resonator beam. The action of the scheme is similar to this one in Fig.3 with the advantage that the illuminated IW1,2 beam is with decreased diameters thus assuring additional laser line with narrowing.

IW1

tune λ1

tune λ2

low-power signal

output at λ1

M1

λ1

AR front facediode-laser

L

LGλ2

laser output

λ1 + λ2

0 order

IW1

tune λ1

tune λ2

low-power signal

output at λ1

M1

λ1

AR front facediode-laser

L

LGλ2

laser output

λ1 + λ2

0 order

0.1

0.9

λ,λ,λ,λ, nm

(x arb.units)

625 660

transmission, ~

λλλλ1

λλλλ2

tune λλλλ1

0.1

0.9

λ,λ,λ,λ, nm

(x arb.units)

625 660

transmission, ~

λλλλ1

λλλλ2

tune λλλλ1

(a) (b)

Page 13: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 13 -

where Nt is the total number of the active Yb-ions in cm3 (5.5*1020 cm3 ; here and below in parentheses we will give the value of the parameter corresponding to our experiment). The populations of the upper and lower laser levels in cm3 are noted as N1 and N2. The term Rp

(1.78*1023 cm-3s-1) is the pump rate and for the diode pumped laser, following Ref. [10], it can

be given by the expression Rp=ηpPp / hνpAl, where A (4.75*10-4 cm2) is the cross-sectional

area of the pumped volume of the active medium and l (0.3 cm) is its length, hνp (2.05*10-19

J) – the energy of the pump photon and Pp (5.5 W) – the pump power, ηp = 0.95- quantum efficiency The generated photons in the cavity is denoted as q1, q2 and qf for the two

selected wavelengths λ1 and λ2 and for the wavelength λf of the gain maximum, respectively. The qf gives the non-desired free lasing in the main cavity that, in the non convenient conditions, can accompanied or suppressed the selected emission in the passive self-injection locked schemes [7]. The output powers Pout1,2,f are proportional to q1,2,f , respectively [10].

Here ( ) ( ) eaakekakek LVlcB σ= (s-1) where σek (cm2) and σak(cm2) are the emission and absorption

cross-sections for the laser transitions, respectively andekakkf σσ= (σe, σa and fk are

functions of the wavelength given below). The index k = 1, 2, f is for corresponding

wavelength λk ; c (cm/s) is the speed of the light in vacuum. Va (1.43*10-4 cm3) is the active volume and Le (16 cm) is the optical length of the cavity. The term

keck cL γ=τ is the photon

life time, where

( )( )[ ]2k2k1ikkγ+γ+γ=γλ

presents the total losses for generation [10]. Here

γ1k ≈ - ln (R1) and in our case is ≈ 0; γ2k ≈ -ln (Req) where Req is function of the wavelength as it

is discussed in Section 2. In our schema γi is sum of three parts: the losses introduced by the

laser crystal γc (0.09), the losses of the transmission of the mirror M2 - γM (0.01) and the

useful losses of the output plate G - γout (0.02 for the optimized combination output power- tuning range).

Using this system, we have treated the threshold condition for simultaneous operation, the two-wavelength tuning range and the relative powers in the CW laser emissions at both

wavelengths. We treat the simultaneous stationary generation at two arbitrary wavelengths λm

and λn (m=q , n=t) The threshold conditions follows of equalizing the derivates dqm/dt and dqn/dt to zero. Thus, we can obtain the needed ratio between the losses for generation of each wavelength as a function of the corresponding simulated emission cross-section. Using this procedure, we have obtained for the losses [10] the following condition for simultaneous two-wavelength operation.

(6)

lN tanen

amemamem

anen

amemnm

+

++−

+

+=

σσ

σσσσ

σσ

σσγγ

( ) ( ) ( )

( )

( )

( ) )5(

)4(

)3(

)2(

)1(

12

2

212222

2

1

111211

1

2121222211211

2

12

cf

ffeffa

f

cea

cea

feffeep

t

qNfNBqV

dt

dq

qNfNBqV

dt

dq

qNfNBqV

dt

dq

NNfNBqNfNBqNfNBqR

dt

dN

NNN

τ

τ

τ

τ

−−=

−−=

−−=

−−−−−−−=

+=

Page 14: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 14 -

Here m and n are the total losses in the cavity at m and n, respectively and σem and σam

and σem and σam are the absorption and the emission cross-section for the corresponding

wavelength λm and λn, respectively, Nt is the total number of active Yb: ions and l is the length of the active medium.

If we directly consider the CW operation equalizing the derivates to zero we loss an information about the output power at each emission. Important idea in our consideration that permits the calculations of the power in each emission consists of solving the system (1) - (5) using the pumping with slowly increased power (~ ms) and raising the stationary value corresponding to the stationary pumping power. An example [7] of the calculated output powers at two wavelengths using this procedure are plotted in Fig. 11. Thus, as it is made in our work Ref. 7 we can calculate the noted above dependences of interest. Here in Fig.12 with solid lines we have plotted the calculated by the system (1) – (5) variation of the output power at each wavelength during the tuning with correction of the losses using equation (6). Analogical dependence like the (6) are obtained for semiconductor lasers [9]. The condition for simultaneous generation at two wavelengths is given by:

where [10] LR .ln 1 αγ +′−=′ , LR .ln 1 αγ +′′−=′′ , 2

2

2 D

′+

′=′ ,

2

2

2 D

′′+

′′=′′ ,

λπ

′−=′

dnnD .).(2 2/12

221 ,

λπ

′′−=′′

dnnD .)(2 2/12

221 .

Here σ’ and σ” are the corresponding of λ’ and λ” gain coefficients, n1 and n2 - the refractive indices of the active medium and cladding layers for Ga As laser (n1 =3.6, n2=3.4), d ~ 0.1

µm– the thickness of the active layer, α is the loss coefficient and is of ~10-1 cm, R1 resonator

length of L=300 µm, R1’ and R1” are the reflection of the corresponding at λ’ and λ” of the external mirrors. These reflectivities must be varied to be satisfied the condition (6) to obtain

generation at the two wavelengths λ’ and λ”. In practice these reflectivities can be varied by introducing a neutral filter with variable transmission or by their slowly misalignment (the last produces decreasing of the retroreflected beam). In the practical realization this is achieved experimentally both manners without prior calculations.

4. Two-wavelength Cw Yb :GGG and diode-laser performance - experimental

4.1. Yb:GGG laser

Single-Lasing-Volume Case with Emission in a Single Beam. The two-wavelength cavity was obtained from the cavity described in Section 2 by the addition of mirror M6 and wedge IW2. We then implemented the schemes depicted in Figs. 1(a) and 1(b). Here we consider the performance for the scheme of Fig. 1(a). When mirror M6 is blocked by a screen,

the laser operates as a single-wavelength laser at λ1. If we block mirror M5 and unblock

mirror M6, a similar single-wavelength operation at λ2 is obtained. If we remove the screen in front of M5, both channels are opened. Because of the well-known mode competition effect,

the laser starts to operate at wavelength λ1 or λ2 for which the lasing conditions are more

favorable. If this wavelength is λ1, we misalign M5 slightly to induce loss in this channel to satisfy the condition in Eq. (6). During misalignment, laser oscillation at the second

wavelength λ2 appears and reaches a maximal value. A spectral analysis of the output spot along its transverse section shows that both wavelengths are emitted in a single beam. In

γσ

σγ ′

′′

′′′′=′′ .

Г.

Г. (7)

Page 15: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 15 -

accordance with the theoretical prediction [7], simultaneous oscillation of a pair of wavelengths was obtained experimentally over a limited spectral range. The experimental output power for each wavelength is given in Fig. 12. We obtained these points by choosing

λ2 =1023.3 nm and tuning λ1. Two-wavelength operation can be obtained for a wide range of a few nanometers, but output powers at the two emissions are similar only at 1.5 nm around the wavelength that corresponds to near-equal powers. The relative intensities of the two wavelengths were measured separately through the monochromator. The power at each wavelength was then deduced by a measurement of the total output power. Fig.12 shows the good agreement between the calculations and the experimental results. The experimental tuning range over which the powers of the two channels remain similar ratio less than 20%, for example is larger than the theoretical prediction. This might also be related to the spatial hole-burning effect as discussed above for single-wavelength operation. This difference can also be attributed to saturation effects that are not considered in the ideal quasi-four-level laser system18 used. Meanwhile, we can conclude that the simplified system employed for the calculations provides a useful quantitative guide for many of the parameters of interest in the two-wavelength operation. If the losses are not adjusted during the tuning, a smaller spectral range is obtained. The experimental tuning curve is plotted in Fig. 13. One of the wavelengths

λ2 =1023.3 nm remains constant whereas the other λ1 is tuned. The tuning range is approximately 0.5 nm. Out of this range, the generation of one wavelength suppresses the other. Two-wavelength lasing operation can be re-obtained from the noted range if one rebalances the losses in both channels to satisfy the relation in Eq. 6. We have investigated the temporal characteristics of the emission at both wavelengths. Fig.14 shows oscilloscope traces

of the generation at two wavelengths λ1 = 1027 nm and λ2 = 1024.5 nm with the output power obtained at 0 for different wavelengths 0 are given in Fig. 13 by the filled triangles. The same dependence was obtained for the output power filled circles in Fig. 13 for the second emission

when λ2 was tuned. Curve (a) was calculated with laser parameters that correspond to our experimental conditions and shows good agreement with the experimental results.

Fig. 11. Examples of the calculation of the cw output powers for two pairs of wavelengths by use of the system of Eqs. (1)–(5) and the condition in Eq. (6). Pumping rate RP was modeled by a function with a linear part from 0 to 5.5 W during 0.5 ms and a constant value of 5.5 W there after. The laser parameters are given in the text.

Page 16: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 16 -

4.2. The two-wavelength red diode-laser operation.

In the investigation, we use values of the laser parameters, given on the picture. We will describe here only the results, obtained using the scheme in Fig.10. The two-wavelength operation in the other two schemes leads to similar results of this one, obtained in the scheme in Fig .10. More details are given in our article Ref.9.

Fig.13 Experimental tuning curves Pout,1 and Pout,2 obtained for fixed losses in the

two channels. Wavelength λ1 is tuned and λ2

is fixed.

Fig. 14. Oscilloscope traces of the power at each wavelength, both obtained with the scheme of Fig. 1 (a). The total output power is 0.4 W equally distributed in each wavelength. Because of the emission in a single-lasing volume, the mode competition

effect is strong and evident in (b).

Fig. 12. Calculated output powers Pout,1 at λ1

tuned and Pout,2 at λ2 fixed at 1023.3 nm when the two wavelengths are generated simultaneously. The circles and triangles present the corresponding experimental points. The losses are balanced in each channel during

the tuning of λ1 by an adjustment of M3 to satisfy the condition in Eq. (6).

1

1

2

2

1

1

2

2

Page 17: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 17 -

Fig.15. Tuning curves for two-wavelength generation. One of the

wavelengths (λ1) is fixed and the

second (λ2) is tuned. For each

wavelength λ2 the losses in the two channels are appropriately balanced to produce two-wavelength operation.

In the described experiment, the diode-laser emission was produced for the constant

currant of 80 mA. The laser threshold was ≈ 67 mA. Firstly, we block the operation in the grating channel by inserting a screen in front of the grating. The laser output was at single wavelength, selected in the firs channel by the IW3. The maximum output power was ~ 200 µW at 633 nm. By translation of the IW3 we perform the tuning of the single-wavelength emission of ~ 5 nm around 633 nm. As a second step, we block the mirror M1 by inserting the screen in front of it and after this unblock the operation in the grating channel. We obtain tuning range approximately from 633 nm to 637 nm by slowly rotation of the grating. When the two channels are unblocked simultaneously, as a rule, the generation in one of them – this one with the IW3, suppresses the lasing in the other. By slowly tilting the mirror M1 or introducing the variable-transmission (0.5 -0.9), low-absorbing filter and by vary its transmission we obtain the two wavelength emission. The wavelength competition was exceptionally strong and needed very precise balance of the introduced losses. When this balance was achieved, the laser operates long time at the two wavelengths. The two-wavelength operation was possible only near the gain maximum – in a range of ~ 4 nm near 633 nm. The linewidth of each emission is evaluated to be approximately 0.1 – 0.3 nm when a single selector is used. The maximum sum energy in the two-wavelength emission is slowly lower than of the emission at a single wavelength (e.g. 150µW-170 µW). For each tuning of one of the channel, one of the laser emission is suppressed by the other and to reproduce the two wavelength operation it is necessary a new precise balance of the losses. The beam has

diameter of ≈ 2.5 mm and specific distribution with near rectangular high intensity part along the direction of the diode p-n junction, superimposed on the near homogeneous illumination

of the beam spot and a divergence of ≈3mrad. The typical tuning curves, obtained when one of the wavelengths - this one, selected by the IW1 is fixed and the second is tuned are plotted

in Fig.15. The two cases are shown – for relatively near threshold pump current of 80 mA and for short time pump with a current of 110 mA that is not recommended for operation of our diode-element. In both cases, the strong competition effect can be seen between the generations and this effect is more essential for low-pump of 80 mA. For all pair of points the losses in the channels are precisely balanced. The noted phenomenon of the wavelength competition in the homogeneously broadened medium, such as the diode-laser, is well known and widely discussed in the literature for such type of operation [8-10]. The wavelength competition is a limitation of the application of this attractive technique. However, in many practical cases, one appropriately tuning and arrangement of two-wavelength diode-laser operation can be very useful. In our experimental conditions, the strong competition is due to work not higher exceeding the threshold. We hope that the noted strong competition will be reduced using convenient high gain laser operating essentially over the threshold. This is the aim of our further investigations. Earlier, using a grating or polarizing schemes

experimental realization of the two-wavelength operation in infrared (~830 nm) semiconductor laser is reported in Ref.15. The authors are observed similar strong

0

20

40

60

80

100

120

140

160

633 634 635 636 637

����1

����2

P����1,����2, ����W

����2 , nm (tuned)

����1

����2

for p.c. 80 mA

for p.c. 110 mA

���� 1 fixed ( 633 nm )

Page 18: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 18 -

competition between the two emissions and discuss the reason that permits the simultaneous two-wavelength operation in homogeneously broadened active medium. They suggest some reasons – the known spatial hole burning effect, filamentation, mode hoping and balanced gain for two emissions. In their point of view, the mode jump process may explain double-mode oscillation. However, they are not presented the temporal investigations. Earlier we have reported the two-wavelength operation in CW Titan:Sapphire laser [5] and we have used successively this operation for two-step excitation of atoms using non-metastable level. This shows that the two wavelengths are produced simultaneously because of the wavelength jump for less a few nanoseconds is not probable in 2-meters resonator (our experiment). As we have reported above in this report, the temporal investigation of the two-wavelength Yb:GGG laser with microsecond resolution (Fig. 14) shows an absence of the wavelength jumping. We have investigated the temporal behavior of our two-wavelength red semiconductor laser with resolution to 20 ns. Our conditions permit to observe the temporal behavior of one of the emissions during the two-wavelength operation. This is shown on the oscilloscope traces in Fig.16(a) – for 50 ns/div and in Fig.16(b) – for 20 ns/div. A 300 MHz LeCroy oscilloscope and ns-resolution detector are used, however below 50 ns the oscilloscope trace correspond of numerical mode of registration with change the real form of the signal. Nevertheless in all cases it can be seen that the wavelength jumping not present. We attribute the possibility for two-wavelength operation with a mechanism of equalized difference gain minus losses for both generations. This is confirmed by noted in this section needed precise balance to obtain two-wavelength operation.

(a (b

Fig.16. The temporal behavior of the laser emission at the one of the wavelength during the two-wavelength operation. At the each oscillogram the upper trace correspond to the temporal development of the laser emission and the low trace – to the receiver signal in absence of a laser generation. The time scale for left oscillogram is 50 ns/div and for the right – 20 ns/div. Not seeing laser emission intensity jumping.

Page 19: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 19 -

Fig. 17. Actual photograph of a two – wavelength CW red diode laser, realizing the scheme given in Fig. 9 with visualization of the two intra – cavity beams. The labels are given in the figure in the inset.

5. Conclusion

In the present work, we have summarized and generalized the solution, the approach for theoretical treatment and for experimental test data for two-wavelength operation of two actual lasers – CW Yb-doped crystal laser and CW red semiconductor lasers. We have discussed the area of applications of two-wavelength laser emission. It was shown that this lasers can be operated at two wavelengths simultaneously that represent a very attractive opportunity to obtain two-wavelength operation.

The two-wavelength lasers are a simplest and cheaper way to obtain such emission with maximum power for a given pump source used. In addition, what is very important, as it was shown here for a given pump source with limited pump power this can be a unique way for producing laser light at two independently tunable wavelengths –the division of the pump radiation for pumping two separate lasers in real practical cases can be insufficient for pumping the two lasers.

Except the noted advantage for the two-wavelength lasers, there are limitation of the tuning possibility – the tuning at both wavelengths is limited in relatively narrow spectral

Page 20: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 20 -

range of a few times narrower that this one obtainable with two separated lasers. This results of strong mode competition effects.

Nevertheless this limitation the developed two-wavelength lasers present, also potential, interest for many practical applications that do not need of wide tenability – in DAS monitoring of the atmospheric composants, in isotope separation, in non-linear investigations, in holography. In addition, the proposed our schemes are very simple for realization. Also, they permit with simple modification to be operated one expensive laser sources (e.g. as the considered lasers) as two separate lasers in just imposed beams without wavelength competition effects. However, the output power decreases, as it is typical for producing a two-wavelength light using separate standard lasers. In the last two-wavelength laser realization, the advantage is only technical – cheaper and simple laser construction.

ACKNOWLEDGEMENT

The work is supported by INM/BNM – CNAM, LIRIS Universite de Versailles Saint-Quentin and NFNI Bulgarian Contracts VUF-12 and Ph-1305.

REFERENCES

1. H.Lotem, R.T. Lynch. Double-wavelength lasers, Appl.Phys.Lett., 27, 344-346, 1975 and the literature therein; M.Schutz, U.Heitmann, A.Hese. Developmentof a dual-wavelength day- laser system for the UV and its application to simultaneous multi-element detection”, Appl. Phys., B61, 339-343 (1995)

2. W.Demtreder. “Laser Spectroscopy (Basic concepts and and Instrumentation)”, Second Edition, Springer, 1996

3. M.N.Nenchev, Y.H.Meyer. Two-Wavelength Dye-Laser Operation, Using a Reflecting Fizeu Interferometer, Appl. Opt., 24, 7-9 (1984) and the literaure cited therein; M.N.Nenchev, M.M.Martin, Y.H.Meyer. Alternate Wavelength DIAL Dye Laser Using Reflecting Interference Wedge, Appl. Optics, 24, 1957-1959 (1985) and the literature cited therein

4. M.Neveux, M.Nenchev, R.Barbe, J.C.Keller. A Two Wavelength Passively Self-Injection Locked, CW Ti3+:Al2O3 Laser, IEEE J. Quant.Electron., 31, 1253-1260 (1995) and references therein

5. R.Scheps, J. Myers. Doubly resonant Ti:Sapphire laser in Advanced Solid-State Lasers, L.L.Chase, A.A.Pinto, eds., Vol. 13 of OSA Proceedings Series (Optical Society of America, Washington, D.C., 1992), 60-63

6. F.Siebe, K.Siebert, R.Leonhardt, H.Roskos. A fully tunable dual-color CW Ti3+:Al2O3 laser, IEEE J. Quantum.Electron., 35, 1731-1736 (1999)

7. Y.Louyer, J.-P. Wallerand, M. Himbert, M. Deneva, M. Nenchev. Two-wavelength passive self injection controlled opertion of a diode pumped cw Yb:doped crystal lasers, Appl. Optics, 42, № 27, 5463-5476 (2003)

8. Y. Louyer, J.-.P Wallerand, M. Himbert, M.Deneva, M. Nenchev. Diode pumped cw Yb:YAG lasers for application in spectroscopy and metrology: single-mode and two-wavelength operation, Proc. SPIE, Vol. 5226 54-58, 2003; J.- P.Wallerand,

Y. Louyer, M. Himbert, M.Deneva, M. Nenchev. "Development of diode-pumped Yb:doped crystal lasers for applications in spectroscopy and metrology", invited article (invited report) Proc. SPIE, Vol.5449, 85 – 97 (2004) USA

Page 21: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 21 -

9. S. Topcu, Y. Alayli, M.Deneva, N. Kaymakanova, M. Nenchev. Two-wavelength cw red diode laser: application of original and simple two-channel resonators for independent tuning, Proc. of LTL Plovdiv'05, Intern. Symposium, ISSN 1312-0638, 339-345, (2005)

10. O.Svelto. Principles of lasers, 5nd ed., Plenum Press, New York, 2000

11. M.Nenchev, M.Nenchev. IIR Bulg. Patent No32703/1981, Tunable two-wavelength laser

12. E.Stoykova, M.Nenchev. Transmission and reflection properties of the interference wedge for illuminating with limited diameter beam relevant to laser applications, Opt.

Quant. Electronics, 25, 789-799, West Eur. (1993)

13. E.Stoykova, M.Nenchev. Fizeau wedge with unequal mirrors for spectral control and coupling in a linear laser, Appl.Optics, 40, 5402-5412 (2001) and the literature therein

14. D.Slavov, M.Nenchev. A comparative study of approaches for spectral control of Ti:Sapphire lasers, Opt. Communications, 200/1-6, 283-301 (2001) West Eur.

15. H.Lotem, Z.Pan, M.Dagenais. Tunable dual-wavelength continuous-wave diode laser operated at 830 nm, Appl.Optics, 32, 5270-5273 (1993)

Department of Optoelectronics and Laser engineering Technical University–Sofia, Plovdiv Branch 25, Tsanko Dystabanov Str. 4000 Plovdiv BULGARIA E-mail: [email protected]

Page 22: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 22 -

Page 23: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 23 -

©Journal of the Technical University at Plovdiv “Fundamental Sciences and Applications”, Vol. 13(3), 2006 Anniversary Scientific Conference’ 2006 BULGARIA

EFFECTIVE TWO COAXIAL ACTIVE MEDIA Nd3+:YAG – DYE FLASH LAMP PUMPED LASER

PLAMENA TODOROVA, MARGARITA DENEVA, MARIN NENCHEV

Abstract: On the base of our earlier proposal we have developed an original two coaxial active media lasers, that oscillate at two different colors (fundamental - 1.06 mμ and 0.59 mμ Nd :YAG – Rh6G, possibility for frequency doubling). The two coaxially disposed media with different absorption spectrum are pumped by a single flash – lamp, exploiting in this manner better the pump radiation. The laser construction is essentially simpler and chipper that this one of two separate lasers due to the use a common pumping source, mechanics and optical elements. The laser action is modeled by a rate differential equations system. The conditions for simultaneous operation of both media and for comparable output characteristics are carried out.

+3

Key words: solid state and dye active media, flash-lamp pumping, simultaneous two-wavelength operation.

ЕФЕКТИВEН Nd:YAG-БАГРИЛО ЛАМПОВО ВЪЗБУЖДАН ЛАЗЕР С

КОАКСИАЛНИ АКТИВНИ СРЕДИ

1. Introduction Many practical applications of the lasers need a laser light at two independently

tunable wavelengths, in particular, in two different colors – e.g. in differential absorption spectroscopy – in LIDAR aerosols and pollutants monitoring, in holography, in metrology, in generation of sum or difference frequencies in nonlinear optics [1]. The trivial and traditional way to solve this question is to use two separate lasers each producing one of needed emission. However such realization, evidently, is expensive and complicated, especially in the case of flash-lamp pumped lasers. It is necessary also to synchronize the two excitation subsystems. When the needed spectral range of the two wavelength emission is limited (single color case) there are possibilities to operate one laser with a single active medium at needed two wavelengths [2-4]. However this limitation of the operating spectral range of the two wavelengths is not convenient in many cases of the noted above applications. In this work we develop our previously patented [5] technique for nontraditional and effective solution of the question to produce in simple and chipper manner two laser emissions with different colors and with independent spectral and temporal control of each of them. Also, our original laser presents the basic advantages to use in very effective manner the pumping energy and to have an increased efficiency comparing with two-separated lasers combination. We describe here a variant of realization of the laser proposed using as active medium a combination [5] of

Copyright © 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

Page 24: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 24 -

Nd:YAG and Rhodamine 6G day ethanol solution and modeled the action of this nontraditional laser on the base of rate differential equations system. On the base of the numerical investigations we have studied the action of the systems and have carried out the conditions for effective operation, including the simultaneous emissions at the two wavelengths with comparable output energy.

2. An original coaxial two active media laser architecture and evaluation of the pumping parameters

The principle of our original system relates to use an active element composed by two closely coaxially disposed active medium with different absorption spectra that overlap the emission spectrum of the pumping lamp. In this manner we realized three main advantages:

- the construction of the two-color laser system is essentially simplified in comparison with the use of two separate lasers

- the laser system efficiency is high due to the more effective use of the pump lamp

radiation

- the two color emissions are synchronized and superimposed in time in a simplest manner and propagate in a single beam. After focusing in the space under illumination of both radiations the two color laser light can be completely specially superimposed

A very convenient combination for such complex active element is the pair of Nd:YAG crystal or Nd: glass and Rhodamine 6G dye solution (e.g. in ethanol). As the illustration of the proposal in Fig. 1 are shown the typical three spectra - these ones of the Nd:YAG absorption (top), of the Rhodamine 6G (10-4 M/l) ethanol solution (middle) and of the emission of the pump flash lamp (bottom), taken from the Reference 6 and 7. The noted convenience is evident.

Fig.1

In Fig.2 is shown schematically one variant of the realization of the coaxial laser active elements that use the noted combination. The shown variant is very convenient when the tuning of the dye laser emission is needed. The construction is clear from the examples in the figure. The Nd:YAG standard cylindrical laser crystal with external diameter 2A (of 8 mm in the example, variable in the investigations) and length l (of 100 mm in the example) is used. Axially, in the crystal is bored a cylindrical hole with a bore diameter 2a (of 3 mm in the example, variable in the

Page 25: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 25 -

investigation). The two 1-mm bore diameter holes that are perpendicular to the laser axes connect the two ends of the axial bore with the laser solution pump circuitry. In the shown in Fig. 2 variant of realization, the crystal is cut in one end at Bruster’s angle and in the other – perpendicular to the crystal axes in the shown variant. The axial hole is closed at the two ends by two Bruster’s angle cut in the external side stoppers, made of the YAG crystal material.

The described coaxial laser element is disposed in the two channel resonator, shown in Fig. 2. The common output reflector for the two media is a flat dielectric mirror M1 with 80 % reflectivity for he two generated wavelengths (1.06 mμ and 0.57 – 0.61 mμ ). At the other end of the complex element the beams from the two active medium are separated and closed each with corresponding high reflectivity flat mirror – M2 and M3. Thus, at each channel can be separately introduced a spectral selector (for the dye laser channel) or Q-switching element for the Nd:YAG laser channel. This is shown schematically by the dashed sub-pictures in Fig. 2.

IW

Dye active medium

circulation system

Electricalsupply

Q-switchM2

M3

M1

pump flash lamp

λ = 1,06 μm

λ = 0,59 μm

Nd:YAG rod

Laser output

Dye

λ = 1,06 μm

λ = 0,59 μm

Output beam cross-section

tuneIW

Dye active medium

circulation system

Electricalsupply

Q-switchM2

M3

M1

pump flash lamp

λ = 1,06 μm

λ = 0,59 μm

Nd:YAG rod

Laser output

Dye

λ = 1,06 μm

λ = 0,59 μm

Output beam cross-section

λ = 1,06 μm

λ = 0,59 μm

Output beam cross-section

tune

Fig. 2

We accept that approximately 80% of the pump light energy, emitted by the pump flash- lamp, is concentrated in the complex laser element (use of high quality elliptical light concentrator). The pumping is by the typical pulse parameters for Dye laser pumping – steep leading front of a few microseconds and microsecond duration pulse with few Joules optical pump energy. We considered the pump pulse with trapezium shape with leading front of 5 sμ , plateau of 3 sμ and trailing front of 15 sμ . This can be obtained effectively by operating the flash lamp in the co-called electrical circuitry “double pulse” [8 -10]. Following Ref. 6, we must take into account the focusing effect in our complex cylindrical laser element. From this Reference we can take the pump energy density ρ(r) as a function of distance r from the active

Copyright © 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

Page 26: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 26 -

road axis. In our case the external Nd:YAG part of the element does not absorb the pump light for the Rhodamine 6G dye solution. In this condition, the dependence ρ(r) that is valuable for the considered by us element, taken from Ref. 6, are shown in Fig. 3. We have found that the shown curve can be very closely approximated by the exponential curve. The found by us approximation curve can be analytically described as:

( )3ln

2

2

3⋅−

= A

r

o er ρρ (1) Using this approximation we can calculate the parts of

the absorbed energy in the dye volume and in the Nd:YAG volume respectively. The absorbed in the dye volume part ED is given by the volume integral (φ is the angle in the active medium cross section):

Fig. 3 (from Ref. 6)

( ) dlddrrrEa

0

2

0

l

0D ϕρ

π∫ ∫ ∫= (2)

The total incident pump energy Etot in the laser element is given by:

( ) dlddrrrEA

0

2

0

l

0tot ϕρ

π∫ ∫ ∫= (3)

Using (1), (2) and (3) we can found the ratio totD EE :

( )[ ]22

tot

D Aa3lnexp123

EE −⋅⋅−−= (4)

From (4) we can found the energy in the dye volume:

( )[ ]22totD Aa3lnexp1

23.EE −⋅⋅−−= (5)

In similar manner we can calculate the fraction of the total energy at the absorption wavelength in Nd:YAG volume.

DtotNd EEE −= (6)

We accept that the pump pulse shape is the same at the pump wavelengths for the Dye and for the Nd:YAG and the total pump energies are equal for the Nd:YAG and the Dye medium . Thus from the ED and ENd and the given above pump pulse shape we can calculate the pumping rates [7] and ( )tR D

p ( )tR Ndp as a function of the time t. Thus we are able to

analyze the proposed laser system threshold and output parameters as a function of the pump

Page 27: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 27 -

parameters, active laser element and laser resonator parameters. This is very important to optimize the mode of operation and to evaluate the real advantages of the laser system proposed.

3. Investigation of the laser system proposed We modeled the action of the proposed complex laser by adapting the set of the

differential rate equations for describing the laser operation [7]. Both lasers are treated as a four-level system with essentially different parameters. We can write:

( )D,Nd

D,Nd2D,NdD,Nd

2D,NdD,Nd

p

D,NdN

qNBtRdt

dN2

τ−⋅⋅−= (7)

D.Ndc

D,NdD,Nd

aD,NdD,Nd

2D,Nd

D,Nd qVqNBdt

dqτ

−⋅⋅⋅= (8)

Here the exponent indices Nd and D are for Nd:YAG active medium and for the Dye medium, respectively (in practice (7)-(8) describe two identical in form systems for four-level laser scheme with different values of the parameters). In the systems with is noted the population of the upper laser level per unit volume in both media,

DNdN ,2

( ) ( )DNdDNda

DNdDNd LVclB ,,,, '...σ= [s-1], where is the emission cross-section with maximum value of and , - the working volume, is variable depending on the diameter of the dye active medium; c=3x10

DNd ,σ219

21 103 cmNd −⋅=σ 21621 108,1 cmD −⋅=σ DNd

aV ,

10 cm/s is the light velocity; - the optical length of the cavity, l=10 cm is the length of the active media, are the optical lengths of the resonator for the Nd:YAG and the Dye and are 32 cm and 30 cm, respectively. The time members =0,23 ms and 3 ns, respectively are the lifetimes of the upper laser level for Nd:YAG and the Rh6G Dye. The dumping time of a photon in the resonator is

)1n(L'L −+=DNdL ,'

DNd ,τ

( )D,NdD,NdD,Ndc .c'L γτ = , where γNd,D [7] describes the

loss in the resonator for the two media, respectively (the value of γ is variable in the investigations, depending on the reflectivity of the resonator mirrors).

Except the four level parameters for the Dye and Nd:YAG active medium, we must taken into account the difference between the laser operating volumes. This lasing volumes for the Dye and Nd:YAG are and laV 2D

a ⋅⋅= π ( ) laAV 22Da ⋅−⋅= π .We solve

numerically the system (7)-(8) using the calculated function [7] for each active medium as it was described above.

( )tR p

Firstly, for the chosen from practical reasons bore diameters of the axial hole a of 3 mm (and 4 mm) for the Dye and for Nd:YAG crystal A of 8 mm (and 10 mm) and the length l of 100 mm we have obtained the dependence of the energetic and temporal output parameters for the Nd:YAG and for the Dye lasers as a function of the pump energy. In this first series of calculation the output mirror M1 reflectivity is 0.9 .The reflectivity of the end mirrors M2 (effective) and M3 for the Nd:YAG and for the Dye lasers are 0.84 and 1 (~ 0.99, approximately 1), respectively. We use the effective reflectivity of the end mirror for the Nd:YAG that is obtained for the real reflectivity of the end mirror of 1, corrected with the two reflections from the end facet of the Nd:YAG crystal (in sum ~ 16 %).

Copyright © 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

Page 28: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 28 -

One of the principal aim of the calculations is to show the condition for operation in the two type of laser active media and with overlapped in time pulses and to study the character of the emissions for typical realized in practice flash-lamp pumping for the Dye laser – very short pump pulse with described above temporal shape and length.. As an example of the calculations in Fig.4 are given typical computed curves of the Nd:YAG (top) and the Dye (bottom) lasers emissions for the total pump energy Etot of 2 Joules (that corresponds to electrical pump energy of ~ 25 Joules). From the figures it can be seen the simultaneous operation of both media in these conditions, however with different output peak powers (a different “y”-scale in the figure).

The obtained curves of the dependence of both outputs energies as a function of the total optical pump energy Etot,opt are plotted in Fig. 5. The calculations are for two cases of the active element dimensions (the diameters of Nd:YAG rod 2A is 10 mm and 8 mm with a diameter of the Dye active medium 2a of 4 mm and 3 mm, respectively as it is given in the figure).

In Fig.6 are plotted the curves of the dependence of the output energies as a function

of the bore diameter of the hole for the dye for the diameter of the Nd:YAG crystal of 8 mm The dashed curves are gives the same dependence only for the diameter of the Nd:YAG crystal of 10 mm. The last series of calculated curves, given in Fig. 7, are of interest because they show the possibility to obtain simultaneous operation with near equal output energy from both media for conveniently chosen set of pump active medium and resonator parameters. The plotted curves in the last figure gives the dependence of the Nd:YAG output energy from the reflectivity of the end mirror in the Nd:YAG channel of the resonator. The output coupler reflectivity (for M1) is constant of 90 % and equal for both channels. The dye hole bore diameter is 3 mm, the diameter of the Nd:YAG crystal is 8 mm. The Dye end mirror reflectivity is constant of 1 and thus

0.0E+00

6.0E+04

1.2E+05

1.8E+05

0 5 10 15 20 25t, μs

Pout, W

Nd:YAG

0.0E+00

1.0E+04

2.0E+04

3.0E+04

0 5 10 15 20 25t, μs

Pout, W

Dye

`

Fig. 4

0

0.2

0.4

0.6

0.8

1

1.2

0 1 2 3 4 5Etot, opt, J

Eout, J

Nd:YAG

(2A=8 mm, 2a = 3 mm)

(2A=10 mm, 2a = 4 mm)

Dye

Fig. 5

Page 29: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 29 -

the Dye laser output energy is also constant, shown in the figure with the dashed line.

From the presented results of the numerical investigations in Fig. 5, Fig. 6 and Fig.7 we can conclude that the proposed laser system is realizable for variety of combination Nd:YAG rod and axial Dye- hole diameters. It is easy to realize the conditions for overlapped in time emissions of both media –i.e. to produces the two color laser light simultaneously and also to realize conditions to produces laser outputs for the two color with near equal energies.

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0 1 2 3 4 5

Etot,opt = 5 J

2a, mm

Eout, J

Nd:YAG

Dye

2A=8 mm

2A=10 mm

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0.5 0.6 0.7 0.8 0.9R2

Eout, J

Nd:YAG (2A=8 mm, 2a = 3 mm)

(2A=10 mm, 2a = 4 mm)

Dye

Fig. 6 Fig.7

4. Discussion of the two-color operation in condition of Q-switching of the

Nd:YAG laser part The possibility to place the Q-switching element for the Nd:YAG laser is shown in

Fig.2. In the scheme proposed the Q-switch do not disturb the operation of the Dye laser part. Both parts can operate independently. To study the character of both operation we can change the general system (7) – (8) for the part of Nd:YAG laser operation. Before the Q-switch opening, in absence of the laser operation, the photon number in the laser cavity can be completely neglected, i.e. we accept q = 0. The accumulation of the inversion population – i.e. the population of the upper laser level N2

Nd taken into account that for the Nd:YAG four level generation scheme the population of the low laser level can be accepted equal of 0 [7], is given by :

( ) Nd

Nd2Nd

p

Nd2 N

tRdt

dN

τ−= (9)

To obtain N2

Nd we solve this equation for the three parts of our pomp pulse, approximated with trapezium with the leading front of 5 μs, plateau of 3 μs and tailing front of 15 μs. The general solution of (9) is:

Copyright © 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

Page 30: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 30 -

t ( ) t tτ τ τekdtetReN p ⋅⋅−= +∫2

− (10)

From the Etot it easy to obtain the temporal dependence of the pump power Ppump (t) for

the noted three parts of the pump pulse and respectively the corresponding Rp(t). From (9) for our pulse shape we can obtain the dependences N2 (t) for any moment before Q-switch opening.

For the leading front:

⎥⎥⎥

⎦⎢⎢⎢

⎣⎟⎟⎟

⎠⎜⎜⎜

−−⋅⋅⋅

= τν

etVht

tNai

p 1)(2

⎤⎡ ⎞⎛ −⋅ ττ tPmax

(11)

For the plateau:

⎟⎟⎟⎟

⎜⎜⎜⎜

⎛ −−⋅⋅

⋅+−⋅= ττ

ντ

te

Vh

PteNtNa

pi

1)(max

22 (12)

where N2i is the population at the end te of the leading front and the time is taken from the end of the leading front – e.g. from te (in this case te is accepted to be 0).

For the tailing fronts we have:

⎥⎥⎥⎥

⎤⎢⎡

⎟⎟⎟⎟

⎞⎜⎛

⎟⎞

⎜⎛

−−

−⋅+−

−⋅⋅+−⋅= tt

et

etPteNtN p

pττττ 11)(

ma

⎢⎢⎢

⎣⎜⎜⎜

⎝⎟⎟⎟

⎠⎜⎜⎜

⎝⋅⋅ eetVh aν

x

(13) 22

where N2p is the population of the upper laser level at the end tp of the plateau and the time is taken from the end of the plateau – e.g. from tp (in this case tp is accepted to be 0).

The generation in the Nd:YAG part is given by the simplified system, in which we have accepted Rp=0 ( [7] – the effect of pumping for the formation of the giant pulse is neglected):

Nd

Nd2NdNd

2Nd

Nd2 N

qNBdt

dN

τ−⋅⋅−= (14)

Ndc

NdNd

aNdNd

2Nd

Nd qVqNBdt

dqτ

−⋅⋅⋅= (15)

Solving this system with initial N2, given by the (11)-(13) for the moment t of the Q-

switch opening by respect the pump pulse starting, we can obtain the temporal curve of the Nd:YAG generation. From the system (14)-(15) we can obtain the curve of the dye laser emission. In fig.8 are shown typical curves of both emissions that demonstrate the possibility

Page 31: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 31 -

of simultaneous operation of the two active media. The more detailed investigation of the combined laser operation for the case of Q-switched Nd:YAG laser part are in progress.

5. Conclusion

0.0E+00

5.0E+06

1.0E+07

1.5E+07

2.0E+07

2.5E+07

0 5 10 15 20 25t, μs

Pout, W In the work we have reported an original laser system that is able to produces simultaneously and in coaxial beams the laser light at two colors. The system has as basic advantages to be essentially simpler in construction and chipper in comparison with two separate lasers and to use effectively the pump radiation of the standard flash-lamps for dye-laser pumping. From the theoretical modeling and numerical investigations we have shown that the system is realizable for variety of the parameters of a coaxially combined laser active element and for the typical flash-lamp pulses for Dye laser pumping, including the case of Nd:YAG part Q-switching. The proposed technique can be realized also in combination of Dye-Dye coaxial medium for flash-lamp pumped dye lasers. The system developed can be useful tool for variety of applications in differential absorption spectroscopy, in atmosphere pollution monitoring by LIDAR technique, in h dye-laser pump holography and in scientific works.

Nd:YAG

0.0E+00

2.0E+04

4.0E+04

6.0E+04

0 5 10 15 20 25t, μs

Pout, W

Dye

`

Fig.8

6. Acknowledgments

The work is supported by NSF – Bulgaria, (contr. VUPh 12 and Ph 1305).

REFERENCES

1. W. Demtreder. Laser spectroscopy: Basic concepts and instrumentation, New York, 1982.

2. M. Nenchev and M. Meyer. Two-wavelength Dye-Laser Operation Using a Reflecting Fizeau Interferometer. Appl. Phys. 24, 1981, 7-9.

3. M. Gorris-Neveux, M. Nenchev, R. Barbe and J. Keller. A two-wavelength, passively self-injection locked, CW Ti:Al2O3 laser. IEEE J. Quantum Electron, USA Vol.31, 7, 1995, 1253-1260.

4. Y. Louyer, J. P. Wallerand, M. Himbert, M. Deneva and M. Nenchev. Two-wavelength passive self-injection-controlled operation of diode-dumped cw Yb-doped crystal lasers. Appl. Optics, USA Vol. 42, Sept. Issue, 2003, 4301-4315.

5. M.Nenchev. Multicolor laser, Bulg. Patent No 25954/1978. 6. A. Mikaelian, M. Ter-Mikaelian and K. Turkov. Solid-State Optical Generators,

Moscow, ed.Sovetskoe Radio, 1967.

Copyright © 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

Page 32: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 32 -

7. O. Svelto. Principles of lasers, 4-th ed., Plenum Press, New York and London, 2002. 8. M. Nenchev and V. Stefanov, IIR Bulg. Patent No 2738/1975, System for producing of

discharge with high intensity preliminary discharge in flash lamps. 9. M.Nenchev and V. Stefanov. Induced losses in the Active Medium of Flash lamp

Pumped Rhodamine 6G Laser. Bulg. J. Phys., 7 (1978) 310-321 (in eng.). 10. I. G. Bassov. Optical schemes of two-wavelength dye lasers, Sov. J .Appl. Spectroscopy,

vol. 49, No 6, 1988, 887-918. Department of Optoelectronics and Laser Engineering Technical University–Sofia, Plovdiv Branch 25, Tsanko Dystabanov Str. 4000 Plovdiv BULGARIA E-mail: [email protected]

Page 33: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright © 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 33 -

©Journal of the Technical University at Plovdiv “Fundamental Sciences and Applications”, Vol. 13(3), 2006 Anniversary Scientific Conference’ 2006 BULGARIA

OPERATOR IDENTIFICATION USING UNIVERSAL MODULE

ANTONIY PETROV, BOYKO PETROV

Abstract. This paper describes iButton application used for identification of software operator. This module realizes bidirectional connection with PC using RS-232 and/or USB interface for change of unique code word for each operator. The unique code word give operator’s personal rights, duties and responsibilities to applied software with high security conduct. The table containing unique code words for each operator is encrypted information stored in EEPROM memory which is not available without authorization and is available for request from encrypted interface. The module work in 3 modes: education mode, service mode and work mode and support up to 50 unique code words. Each work mode can be changed by code stored in separate EEPROM memory. Key words: iButton, RS-232, USB, security

УНИВЕРСАЛЕН МОДУЛ ЗА ИДЕНТИФИКАЦИЯ НА ОПЕРАТОР

1. Въведение Универсалният модулът осъществява двупосочна връзка с персонален компютър

по интерфейс RS-232 и/ или USB с цел обмяна на уникална за всеки оператор кодова дума предоставяща му персонални права, задължения и отговорности пред приложен софтуер с повишена сигурност на поведението. Таблицата с уникалните за обслужваните оператори кодове се съхранява в криптиран формат в енергонезависима памет, недостъпна за неоторизиран достъп и е достъпна при заявка по криптиран интерфейс. Универсалният модул работи в режим на обучение, сервиз и работа . Поддържа до 50 уникални кода. Превключването между работните режими се извършва посредством записан в отделна енергонезависима памет код.

2. Описание на блоковата схема на устройството и алгоритъма на работа Модулът може да се реализира със следната блокова схема:

С PC е означен персонален компютър, където е софтуера пред който оператора трябва да се идентифицира. Интерфейсния блок реализира връзката между микроконтролера и персоналния компютър, която може да е по RS-232 или по USB. RS-232 интерфейса може да се реализира със буфери които съгласуват нивата на RS232C протокола с TTL нива. Един от най-често използваните буфери за RS-232 е интегрална схема MAX232. USB интерфейса може да се реализира с помощта на интегрална схема конвертор от USB към сериен интерфейс или към паралелен интерфейс, ако микроконтролера няма вграден USB модул. Фирмата FTDI предлага такива схеми конвертори.

Page 34: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 34 -

Микроконтролера може да е всеки микроконтролер който работи достатъчно бързо за да може да поддържа комуникация по интерфейс 1-wire(обмен на данни по една жица) и има достатъчно програмна памет за реализация на програмното осигуряване. Приемника за iButton установява връзката между микроконтролера и iButton. iButton обменя информация с микроконтролера по една жица. Този интерфейс още се нарича 1-wire. Предимството на iButton е че съдържанието на неговата ROM памет е уникално и трудно може да бъде фалшифицирано. Захранващия блок има за цел да осигури необходимите нива на захранващото напрежение за всички блокове на модула. Външния EEPROM съдържа кода за режимите на работа на модула.

Режимите на работа на модула са 3:

− Обучение - при този режим на работа в модула се записват идентификационните номера на iButton устройствата за съответния оператор.

− Сервиз – при този режим може да се правят промени на правата и задълженията

на операторите, да се изтрива информацията за даден оператор или да се променя идентификационни номер на iButton за даден оператор.

− Работа – в този режим на работа модула идентифицира оператор и изпраща

необходимата за софтуера информация по криптирания интерфейс.

В микроконтролера е реализирано програмно осигуряване със следния алгоритъм, който е показан на фигурата по-долу. От алгоритъма се вижда, че в по-голямата част от времето микроконтролера ще работи в режим на изчакване на прекъсване по интерфейса и тогава да извърши някакво действие, в зависимост от избрания режим. Това позволява микроконтролера да работи в икономичен режим (“sleep mode”) при който модула има минимална консумация. От този режим може да се излезе при прекъсване по интерфейса с персоналния компютър.

При стартиране на работа на модула най-напред се изчаква захранващото напрежение да достигне до необходимите за микроконтролера стойности. След това се извършва инициализация на модулите вградени в микроконтролера и портовете му.

Page 35: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 35 -

Избира се режима на работа на модула като се чете кода от външния EEPROM. Във всеки от трите режима на работа се прави проверка за смяна на режима. След това се чака за прекъсване по интерфейса и се изпълнява останалата част от алгоритъма за съответния режим на работа на модула.

Page 36: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 36 -

3. Резултати

Като резултати ще приложа няколко снимки които показват комуникация през виртуален сериен порт между модул и персонален компютър и обмена на информация между тях. На първата снимка се вижда работата на модула в режим на обучение. При този режим на работа се добавят идентификационни номера на iButton за различните оператори и нивото им на достъп.

На втората снимка се вижда работата на модула в режим на работа. При този режим се сравняват записите от EEPROM паметта с прочетения идентификационен номер на iButton.

Page 37: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 37 -

На третата снимка е показана информацията която се обменя между модула и персоналния компютър при отказване на достъп.

На последната снимка е показан режим на работа “сервиз”. На снимката се вижда изтриването на данните за оператор от списъка със записите.

Page 38: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 38 -

4.Заключение

Като заключение искам да кажа че модулът за идентификация на оператор пред софтуер може да работи със всеки персонален компютър, който има интерфейс RS232 и/или USB. Съгласуването на начина по който ще се криптират данните, както и протокола за обмен на информацията със софтуера който е на персоналния компютър не трудно.

ЛИТЕРАТУРА

1. MAXIM. iButton overview.pdf, www.maxim-ic.com 2. ATMEL. ATmega8.pdf, www.atmel.com, 2003год. 3. FTDI. DS_FT232R_v104.pdf, www.ftdichip.com, 2005год. Department of Electrical Engineering Technical University–Sofia, Plovdiv Branch 25, Tsanko Dystabanov Str. 4000 Plovdiv BULGARIA E-mail: [email protected]

Page 39: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 39 -

Journal of the Technical University at Plovdiv “Fundamental Sciences and Applications”, Vol. 13(3), 2006 Anniversary Scientific Conference’ 2006 BULGARIA

PULSE CHARGE OF RECHARGEABLE

BATTERIES WITH THE ENERGY HEAPED

IN A PARALLEL RESONANCE SET

SVETOSLAV IVANOV, ELENA KOSTOVA, MINCHO PEEV

Abstract. In this paper we report method for converting DC energy of photovoltaic in AC energy and after this in DC energy again, with possibility for dosage amount of electrical charge for charging batteries. The first converted of DC energy in AC is performed in series included generator of unipolarity square pulses, wide band filter and parallel resonance circle. The next converted in DC energy is performed from half wave rectifier. There are worked out basic equations describing processes in charging devise. There are submitted results of experimental research of wide band filter and parallel resonance circle, and parameters of current pulses for batteries charging. Possibility regulation of amplitude value of charge current pulses is researched.

ИМПУЛСНО ЗАРЕЖДАНЕ НА АКУМУЛАТОР С ЕНЕРГИЯТА НАТРУПАНА В ПАРАЛЕЛЕН РЕЗОНАНСЕН КРЪГ

1. Въведение

Съвременните компютъризирани системи за зареждане на акумулатори (CCS) от енергията получена от фото-волтаични панели, се развива бързо през последните години. Създадени са специализирани контролери за управление на процеса на зареждане на акумулаторите. Най-често използваният метод на управление е методът на широчинно-импулсната модулация [1,2]. Слънчевите инсталации (слънчеви системи, соларни системи, фотоволтаични системи) за добиване на електрическа енергия съдържат най-често слънчев панел, преобразуващ слънчевата енергия в електрическа, акумулаторна батерия (акумулатор) и преобразувател на постоянно напрежение в променливо с промишлена честота. Повишаването на ефективността и дълготрайността на акумулаторната батерия се определя от режима й на експлоатация, като особено голямо значение има процеса на зареждане [4,5,6]. Прилагането на правоъгълни токови импулси за зареждане на акумулатора, в сравнение със зареждането с постоянен ток, допринася за увеличаване на броя на зарядно-разрядните цикли (т.е. живота на батерията) [4,5], възможност за зареждане с по-голям ток (значително надвишаващ 10% от капацитета на акумулатора, както е при постояннотоковия режим) [5], което е предпоставка за намаляване времето на зареждане. Една система за импулсно зареждане на акумулатор, захранвана от слънчев панел има структура показана на

Page 40: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 40 -

фиг.1. Базовата структура се състои от слънчев панел, контролер за управление на процеса на зареждане и разреждане, акумулатор и активен товар [3].

Фиг.1

В тази статия се предлага метод на преобразуване на постояннотоковата енергия на слънчевия панел в променливотокова и след това отново в постояннотокова, с възможност за дозиране на количеството електрически заряд за зареждане на акумулатора. Началното преобразуване на постояннотоковата енергия в променливотокова се извършва от последователно включени генератор на правоъгълни импулси, лентов филтър и паралелен резонансен кръг. А следващото преобразуване в постояннотокова енергия се извършва от еднополупериоден токоизправител с капацитивен товар.

2. Широколентов филтър и паралелен резонансен кръг. Основните параметри на използваният слънчев панел са следните: изходно

напрежение 40V, максимален изходен ток - 500mA и изходна мощност – 20W. Реализиран е генератор на правоъгълни импулси с честота 1kHz, с коефициент на запълване – 50% и амплитудна стойност на импулсите – 40V. На фиг.2 е показана схемата на изследвания широколентов филтър, който се състои от последователно свързани Г-образни НЧ и ВЧ звено, а за товар на схемата служи паралелен резонансен кръг. Изследваният резонансен кръг има висок качествен фактор и резонансна честота

kHz1f рез = . На входа на филтъра постъпват еднополярните правоъгълни импулси от

изхода на генератора – G1. Лентата на пропускане на филтъра е определена от хармоничният състав на импулсите постъпващи на входа му. Нейната широчина е съобразена с хармониците от 1 до 7, тъй като само тяхната амплитудна стойност е по-голяма от 5% от стойността на основният хармоник. Коефициентът на предаване по

Page 41: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 41 -

Фиг.2 напрежение на реализираният филтъра е 0.89 в честотния интервал от 185Hz до 10kHz, които са граничните честоти съответно на високочестотния филтър (ВЧФ) и нискочестотния филтър (НЧФ).

Граничната честота на НЧФ във функция от предварително избраният коефициент на предаване - kНЧФ може да се определи от уравнението:

11

2

НЧФ

2

НЧФ

1CR2

k

k1

= (1)

Граничната честота на ВЧФ може да се определи от следното уравнение

2

ВЧФ

2

ВЧФ

2CR2

k1

k

−= (2)

където p

R е съпротивлението при резонанс на паралелния трептящ кръг, включващ

елементите C3 , L1 и R2. Съпротивлението R2 е вътрешното съпротивление на бобината. Предназначението на филтъра е да осигури синусоидално напрежение за захранване на паралелният резонансен кръг, а благодарение на този резонансен кръг тази честота е равна на основният хармоник. При проектиране на филтъра съпротивлението на кръга се приема за константно и равно на съпротивлението при резонанс – Rp.

При определена честота of НЧ звено предизвиква изоставане на изходното

напрежение спрямо входното по фаза, а ВЧ звено изпреварване със същата стойност. В резултат на това фазовата разлика между изходното и входното напрежение при тази

честота of е равна на о0=ϕ . Честотата of се определя от следния израз:

21

o2

1f

ττπ= (3)

Времеконстантите имат следните стойности: 111 C.R=τ и 2р2 CR=τ .

3. Резултати

Експериментално изследване на импулсното зарядно устройство

Принципната схема на реализираното импулсно зарядно устройство е показана на фиг.3, то е предназначено за зареждане на литиево -йонен акумулатор с напрежение – 7,4V и с капацитет 1,7A/h. В паралел на резонансният кръг е включен еднополупериоден изправител работещ с капацитивен товар (VD1 и C4). Зареждането на акумулатора е импулсно, като включването се осъществява с полевият транзистор VT1. Управляващите импулси се генерират от генератора G2 и са с честота 10Hz и

продължителност ms10tи = . Резисторът R3, ограничава големината на зарядния ток за

акумулатора и определя времеконстантата на разрядната верига (C4, R3). Този резистор е с малка стойност и може да се използва за сензор за големината на зарядния ток.

Page 42: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 42 -

Фиг.3

Заместващата схема на акумулатора включва елементите: R4 (омическо съпротивление на електролита и електродите), R5 (поляризационно съпротивление, което се получава при протичане на ток във веригата) C5(капацитет на двойния електрически слой на електродите) и ACC+6,4V. На фиг.4 е показана формата на напрежението на резонансният кръг по време на положителния полупериод, както и на зарядния ток за

кондензатора C4. Напрежението на кръга достига стойност mU , а заряден ток през диода VD1 преминава в интервала от време 1t - 2t , означен с зарt . Ток протича когато

напрежението на кръга стане по-голямо от напрежението на кондензатора и пада от напрежение върху диода VD1.

Фиг.4

Средната стойност на зарядният ток за един период на синусоидалното напрежение се определя от уравнението

1VD

VD4Cmср

R.2

)2)(UU(cosU2I

π

απα −+−= , (6)

където mU е амплитудната стойност на напрежението; 4CU е напрежението на

кондензатора С4; 1VDU е пада от напрежение върху диода в права посока; RVD1 е

съпротивлението на диода, α е ъгълът на закъснение при включване на зарядния ток и

се определя от израза

4arcsin C VD

m

U U

+= (7)

Page 43: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 43 -

На фиг.5 е показана формата на зарядните импулси за акумулатора с честота 10Hz и с продължителност 10mS. Токовите импулси имат максимална стойност – Imax

= 1,8A и достигат минимална стойност Imin = 423mA. Амплитудната стойност на тока е 10 пъти по-голяма от номиналната стойност на зарядния ток (170mA) в постояннотоков режим на зареждане. Напрежението на акумулатора нараства по време на зарядния ток и достига максимална стойност за време равно на – 800µS. Максималната стойност на зарядното напрежение е Umax =11V.

Time

1.090s 1.095s 1.100s 1.105s 1.110s 1.115s 1.120s 1.125s

V(M1:s)

0V

5V

10V

U akumulator

I(R4)

1.0A

2.0A

-0.1A

SEL>>

I akumulator

Фиг.5 Схемата позволява регулирането на максималната и на минималната стойност на зарядният ток по време на импулса – Imax и Imin. Консумираният ток от слънчевият панел има средна стойност равна на 170 mA, а консумираната мощност от панела е равна на 6,8W (фиг. 1,6, тока I(R1)). Тока в резонансния кръг достига амплитудна стойност равна на 2,7 А, и е показан като ток през бобината – I(L1). На фиг. 6 двете графики на тока са синхронизирани със зарядния токов импулс за акумулатора – I(R4).

Time

1.098s 1.100s 1.102s 1.104s 1.106s 1.108s 1.110s 1.112s

I(R4)

1.0A

2.0A

SEL>>

I(L1)

-2.0A

0A

2.0A

I(R1)

0A

100mA

200mA

Фиг.6

Page 44: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 44 -

Регулирането на зарядния ток може да се осъществи, чрез промяна на коефициента на запълване на импулсите, на генератора захранващ лентовият филтър – G1 (фиг.7). Регулирането на стойността на зарядният ток може да се реализира, чрез промяна на коефициента на запълване на импулсите в интервала от 0.1 до 0.5. Връзката между продължителността на генерираните импулси от втория генератор – G2 и максималната стойност на зарядния ток за акумулатора е показана на фиг. 8.

300

975

14701740 1800

70232 343 407 423

0400800

120016002000

0.1 0.2 0.3 0.4 0.5 0.6

Коефициент на запълване

Imax

,Im

in [

mA

]

Imax

Imin

Фиг.7

29302460 2168 1950 1800

1460831 610 495 423

0

1000

2000

3000

4000

2 4 6 8 10 12

Продължителност на импулсите [mS]

Imax,Im

in [

mA

]

Imax

Imin

Фиг.8

4. Заключение.

Реализираната и изследвана схема за импулсно зареждане на акумулатор позволява нейното включване в затворена система за автоматично управление на зарядния процес. Големината на зарядния ток може да се регулира, посредством коефициента на запълване на входните импулси за филтъра (G1), или чрез промяна в продължителността на управляващите импулси на генератора (G2). Максималната стойност на зарядният ток може да бъде ограничена, чрез избора на стойността на резистора – R3. Продължителността на зарядните импулси може да се регулира, чрез промяна в коефициента на запълване на импулсите, на вторият генератор – G2. Това дава възможност за пряко цифрово управление на схемата от микропроцесор. По време на паузата между зарядните токови импулси е възможно включването на акумулатора към разрядна верига, което ще подобри качеството на процеса на зареждане. Консумирана електрическа мощност от схемата за импулсно зареждане е 6,8W, при максимална изходна мощност на слънчевият панела 20W . Това осигурява надеждната работа на зарядното устройство, дори и при слаба слънчева радиация, и при висока

Page 45: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 45 -

температура на околната среда. Енергията на слънчевият панел може да се ползва и за други цели по време на зарядния процес за акумулатора.

ЛИТЕРАТУРА

1. Joseph R. Woodworth, Michael G. Thomas, John W. Stevens: Sandia National

Laboratories, “Evaluation of the batteries and charge controllers in small stand-alone photovoltaic systems”, Presented at the 24th IEEE Photovoltaic Specialists Conference, 1994.U.S. Government work not protected by U.S. copyright 1 First WCPEC; Dec. 5-9, 1994; Hawaii

2. Lam, L. T., et al, ‘Pulsed-current charging of lead/acid batteries are possible means for overcoming premature capacity loss?,CSIRO, Australia, Journal of Power Sources 53, 1995.

3. Hund, Tom, ‘Battery Testing for Photovoltaic Applications,’ Sandia National Laboratories, Albuquerque, NM, presented at 14th NREL Program Review, Nov. 1996.

4. Гишин С., Акумулатори, София, 2003г. 5. http:// www.bticcs.com – A new technology for the automatic charge control of

rechargeable batteries. 6. Venkat Srinivasan, G. Q. Wang, C. Y. Wang, Mathematical Modeling of Current –

Interrupt and Pulse Operation of Valve-Regulated Lead Acid Cell, Journal of the Electrochemical Society, 150 (3), 2003, pp A316 – A325

Department of Electronics Technical University – Sofia, Plovdiv Branch 61, Sankt Peterburg, Blvd. 4000 Plovdiv BULGARIA E-mail: [email protected]

Page 46: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 46 -

Page 47: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 47 -

Journal of the Technical University at Plovdiv “Fundamental Sciences and Applications”, Vol. 13(3), 2006 Anniversary Scientific Conference’ 2006 BULGARIA

FLYBACK DC-DC CONVERTER WITH IMPROVED

PERFORMANCES

IVAN NEYCHEV, PETAR GYOSHEV, RUMEN KAROV

Abstract. The paper presents simulation analysis and experimental oscillograms of improved topologyes flyback dc-dc converters. The conventional circuit of this kind converters draw pulse current from input sourse and aslo can be sayed that the power factor is not high. In such cases is needed a bulk input filter. The areas of applications of this circuits are more restrictive regarding high power applications. In the presents topologyes this problems are eliminated and the performances of the flyback converters are improved. Key words: flyback, improved, performances.

ОБРАТЕН КЛЮЧОВ ПРЕОБРАЗУВАТЕЛ С ПОДОБРЕНИ ХАРАКТЕРИСТИКИ

1. Въведение

Основен недостатък на голяма част от DC-DC преобразувателите е прекъснатият импулсен ток консумиран от захранващия източник. Това е характерно за понижаващия (buck), инвертиращия (buck-boost) и обратния (flyback) преобразувател [4-6]. Импулсният характер на входният ток води до влошаване на фактора на мощността на съответната схема [1-3] и до наложителното използване на големи входни филтри. Когато въпросните преобразуватели се свържат посредством изправител към променливотоковата мрежа (фиг.1) и консумират прекъснат импулсен ток се получава влошаване на хармоничния състав на мрежата, вследствие на което се увеличават смущенията и претоварванията на съоръженията [1].

Фиг.1.

Поради тази причина преобразувателните устройства свързани към

променливотоковата мрежа е желателно да консумират синусойдален ток от нея. По

Page 48: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 48 -

този начин се поддържа добър хармоничен състав на мрежата, редуцират се

смущенията и се подобрява факторът на мощността на самите преобразуватели.

2. Подобрен обратен преобразувател на постоянно напрежение реализиран с

един транзистор

Модифицираната схема на обратния DC-DC преобразувател, при която се постига непрекъснат характер на входния ток е показана на фиг.2. Той е с галванично разделен вход и изход, като използваният трансформатор трябва да е с въздушна междина. Принципът на работа на този конвертор се състой в запасяване на енергия в магнитното поле на трансформатора през времето в което транзисторът Т е отпушен (правия ход) и отдаване на запасената енергия към товара, когато ключа Т не работи (обратен ход).

Фиг.2.

Прав ход на работа на схемата (отпушен транзистор): при отпушване на

транзистора Т започва да тече ток по веригата: +Uin, L, D2, T, -Uin, като по този начин

се запасява енергия в индуктивността L. Диодът D1 е запушен и токът през първичната

намотка на трансформатора TR се осигурява от разреждането на кондензатора С. Токът

през транзистора Т се получава като сума от консумирания входен ток (течащ през

индуктивността L и диода D2) и разрядния ток на кондензатора С (течащ през

първичната намотка на трансформатора TR). През този интервал диодът D3 е запушен и

товарния ток се осигурява от изходния филтров кондензатор Ct.

Обратен ход на работа на схемата (запушен транзистор): при запушване на

транзистора Т спира да тече ток през първичната страна на трансформатора TR и

диодът D2 също спира да провежда. Отпушва се диодът D3 и запасената енергия в

трансформатора от правия ход на работа на схемата се отдава в товара. По време на

обратния ход се отпушва диодът D1 и по веригата L, D1 се осигурява зареждане на

кондензатора С.

Принципът на действие на тази схема се изяснява с помощта на

експерименталните и симулационните резултати дадени на фиг.3.

Page 49: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 49 -

Фиг.3 а) входен ток през индуктивността L;

Фиг.3 б) ток през диода D1;

Page 50: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 50 -

Фиг.3 в) ток през диода D2;

Фиг.3 г) ток през транзистора Т;

Page 51: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 51 -

Фиг.3 д) ток през трансформатора TR;

Фиг.3 е) напрежение върху кондензатора С;

Page 52: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 52 -

Фиг.3 ж) управляващите импулси (Uge) и напрежението върху транзистора Т (Uce);

От горните пояснения става ясно, че ток от входния захранващ източник се

консумира постоянно (по време на правия и на обратния ход) за разлика от

традиционната схема дадена на фиг.4а, където ток от входния източник се консумира

само през времето на работа на транзисторът , поради което той се получава прекъснат

с импулсен характер(фиг.4б), което определя неравномерното натоварване на входния

източник [4-6]. Този съществен недостатък се отстранява при подобрената схема от

фиг.2.

Фиг.4а. Фиг.4б.

Модифицираната схема (фиг.2.) може да се разглежда като една комбинация

между типичните повишаващ и обратен импулсен преобразувател. Предимството на

повишаващия конвертор е непрекъснатият входен ток, който може да бъде с малки

пулсации при достатъчно голяма стойност на входната индуктивност. Като недостатък

може да се спомене трудното галванично разделяне, тъй като включеният

трансформатор трябва да има допълнителна размагнитваща намотка, през която да се

Page 53: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 53 -

размагнитва магнитопровода на трансформатора, когато транзисторът е изключен.

Предимство на традиционния обратен преобразувател (фиг.4а) е лесното галванично

разделяне, тъй като през правия ход (отпушен транзистор) се запасява енергия в

магнитопровода на трансформатора, която през обратния ход (запушен транзистор) се

предава към товара. Поради това не е необходима допълнителна разматнитваща

намотка както при повишаващия преобразувател. Като недостатък на схемата може да

се отбележи импулсният входен ток, поради което нарастват използваните филтри и се

понижава факторът на мощността. В резултат може да се каже, че подобрената схема

обединява предимствата на типичните повишаващ и обратен преобразувател, а именно

непрекъснат входен ток и лесно галванично разделяне.

На фиг.5 е дадена формата на входния ток при различни стойности на

индуктивността L, за които е изпълнено отношението L1 < L2 < L3.

Фиг.5а. Iвх при L1. Фиг.5б. Iвх при L2. Фиг.5в. Iвх при L3.

3. Двутранзисторна схема на подобрен ключов преобразувател (схема с

натрупване на енергията)

На фиг.6 е дадена схемна разновидност на разглеждания “flyback” конвертор от

фиг.2. Разликата се състой в добавянето на втори транзистор, като долният край на

първичната намотка на трансформатора вместо към колектора на първия транзистор Т1

се свързва към колектора на втория Т2. Следователно отпада необходимостта от диод

свързан последователно към колектора на първия ключ, като той се свързва директно

към входната индуктивност. Чрез използването на два транзистора вместо един се

постига разделяне на зарядната и разрядната верига на кондензатора С. Ключът Т1

работи с честота по-висока от тази на Т2, като постепенно капацитетът С натрупва

енергия, (напрежението върху него расте) докато настъпи моментът на включване на

втория транзистор (работещ при ниска честота) при което следва разреждане на

кондензатора през първичната намотка на трансформатора. След като транзисторът Т2

се запуши процесите се повтарят, т.е. отново започва стъпално зареждане на

кондензатора С с по-високата честота определена от Т1.

Page 54: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 54 -

Фиг.6.

Чрез разделянето на зарядната и разрядната верига на кондензатора и

зареждането му с по-висока честота от тази на разреждане се осигурява голям товарен

ток при сравнително малък консумиран ток от захранващия източник, което е

предимството на тази схема.

4. Заключение При подобрения обратен DC-DC преобразувател даден на фиг.2 е отстранен един от съществените недостатъци на основната схема на този тип преобразуватели (фиг.4а) а имено импулсния характер на тока консумиран от захранващия източник. Това прави схемата удобна при свързването й към променливотоковата мрежа (посредством изправител) – фиг.1, тъй като не са необходими големи филтри поради непрекъснатия характер на входния ток. В този случай не се влошава хармоничния състав на променливотоковата мрежа и се подобрява фактора на мощността на самия преобразувател. При необходимост от осигуряване на по-голяма товарна мощност е целесъобразно използването на схемата с натрупване на енергията дадена на фиг.6.

ЛИТЕРАТУРА

1. Steigerwald, R. Power electronic converter technology, IEEE transaction on power

electronics, June 2001. 2. Kheraluwala, M., R.Steigerwald. A fast response high power factor converter with a single

power stage, IEEE transaction on power electronics, April 1992. 3. Schutten, M., R.Steigerwald. Characteristics of load resonant converters operated in a high

power factor mode, IEEE transaction on power electronics, April 1992. 4. Каров, Р. Преобразувателна техника, “Техника” , София, 1994. 5. Бобчева, М., Н.Градинаров. Силова електроника, ТУ, София, 1998. 6. Стефанов, Н. Токозахранващи устройства, “Техника” , София, 1985.

Department of Electrical Engineering Technical University–Sofia, Plovdiv Branch 25, Tsanko Dystabanov Str. 4000 Plovdiv BULGARIA

Page 55: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 55 -

Journal of the Technical University at Plovdiv “Fundamental Sciences and Applications”, Vol. 13(3), 2006 Anniversary Scientific Conference’ 2006 BULGARIA

CONNECTION OF INVERTERS TO JOINT LOAD

RUMEN KAROV, PENKA LENTCHEVA, DANIEL KAROV

Abstract. Joint load circuits are proposed and investigated for the three converter groups: voltage inverters, current inverters and resonant inverters. Multi-step inverter, compound of three voltage inverters, is shown for improvement of output voltage form. Improved circuit decision for resonant inverter with parallel-operating sections is proposed and an analysis is made based on the equivalent circuit. For the analysis of parallel operation of joint load current inverters is suggested an investigation of separate inverters as dependent ones. Keywords: compound multi-step inverter, joint load inverters operation

СВЬРЗВАНЕ НА ИНВЕРТОРИ КЬМ ОБЩ ТОВАР

1. Въведение

В съвременните силови електронни преобразуватели в случаите, когато има изисквания за подобрена форма на изходното напрежение или допълнително повишаване на изходната честота и мощност, освен наличието на силови ключове с подобрени параметри, се използва свързване на инвертори към общ товар.

Известни са различни начини за подобряване на формата на изходното напрежение, понижаване на изкривяванията и нивото на хармониците – векторен метод, сумиране на напрежения с различни честоти, Широчинно-Импулсна Модулация(ШИМ), филтри и др.[Л3,4].

В настоящата статия са изследвани схеми на свьрзване кьм общ товар за трите групи преобразуватели: инвертори на напрежение, инвертори на ток и резонансни инвертори, които намират приложение при екологични източници и при електротехнологии. Предложен е многостепенен инвертор за подобряване на формата на изходното напрежение, сьставен от три инвертора на напрежение, свьрзани кьм общ товар. Изследвано е оригинално схемно решение на резонансен инвертор с паралелно работещи звена и е направен анализ вьз основа на еквивалентната схема. При паралелна работа на инвертори на ток на общ товар за анализа се предлага разглеждане на отделните инвертори като зависими.

2. Описание

Единият от начините, който обобщава първите два метода, е чрез многостепенен преобразувател с паралелно свързване на инвертори, захранвани от отделни източници с напрежения, изчислени от гледна точка за формиране на синусоидата на изходното напрежение [Л1]. Благодарение на такива схеми се получава синусоидално напрежение

Page 56: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 56 -

с много добро качество, при което липсват 3,5,7 и кратните на тях хармоници. На фиг.1 е показана такава схема, състояща се от три инвертора на напрежение, свързани в паралел на изхода – фиг.1. На фиг.2 е представена диаграмата на напрежението. Изходното напрежение е формирано от трите напрежения като при показаните съотношения липсват 3,5,7,9,15,17,19,21,27..... хармоници. На едната двойка транзистори за първия инвертор се подава управляващ импулс от 0 до π. На втория инвертор управляващите импулси на една двойка ключове се подава от θ1 до θ2 , където θ1 = 30о, а θ2 = 150о. За третия инвертор тези импулси са от θ3 до θ4 , θ3 = 60о, θ4 = 120о за

едната двойка транзистори. Първият инвертор работи от 0 до θ1 и от θ2 до π. Вторият инвертор работи от θ1 до θ3 и от θ4 до θ2, като поради по-високото напрежение на входа му запушва първия инвертор и работи последователно след него. Третият инвертор работи от θ3 до θ4, като аналогично поради по-високото напрежение на входа му запушва втория инвертор и работи последователно във времето след него. След третия инвертор се включва отново втория и след това първия. Така се получава кривата на изходното напрежение, в която липсват изброените по-горе хармоници.

Фиг.1 Фиг.2

+ E3

-

+ E1

-

+ E2

-

~220V

U1 U2

U3

Uизх

θ1 θ2

θ3 θ4

θ

θ

θ

θ

Е 3

Е 2

Page 57: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 57 -

В сила са равенствата:

(1) Е2 = E1 + 3 E1

(2) E3 = 2E1 + 3 E1,

Кривата на изходното напрежение, разложена в ред на Фурие има вида:

(3) [ )( ]∑±

θ±+θ

π=

π=θ

112к

112кsinsinE

12E33

4)(Uизх , където k = 1,2,3..........

Когато има изисквания и за допълнително повишаване на изходната честота и мощност може да се използва свързване на инвертори към общ товар.

3. Резултати На фиг.3 е показана възможност за подобрено схемно решение при резонансните

инверторни с паралелно работещи звена за големи мощности [Л2]. Въпреки наличието на качествени мощни тиристори се оказва, че схемата с паралелно работещи звена има интересни предимства. Особеност на такава схема е свързването на звената към общ товар с отделни комутиращи кондензатори, но с обща комутираща индуктивност. Предимството на схемите преди всичко във възможността за реализиране на мощни

инвертори с проста конфигурация чрез паралелно свързване на групите тиристор - обратен диод посредством комутиращи кондензатори, които кондензатори изпълняват ролята на изравняващи елементи за токовете на тиристорите и диодите. Нещо повече - дори за мощности около 50 - 100кW използуването на паралелно работещи групи позволява разделянето на общия ток и намаляването стойността на такива важни параметри като скорост на нарастване на анодния ток и времето за възстановяване на тиристорите при това не за сметка на увеличаване на анодните индуктивности, а напротив - намаляване на две малки по стойност индуктивности.

Фиг.3

Page 58: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 58 -

Фиг.4

Фиг.5

Фиг.6

Page 59: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 59 -

Изнасянето на анодните индуктивности извън отделните звена и получената конфигурация на схемата с общи точки на анодите на горната група тиристори и съответно на катодите на долната група тиристори позволява използуването на една обща RC група (за намаляване на пиковите напрежения) за всички тиристори.

От друга страна намаляването на анодните индуктивности чрез включването им в общата част, освен че има големи конструкторски предимства, но и позволява да се увеличи многократно бързодействието на защитата чрез самоликвидация на късото съединение при едновременно включване на противофазно работещи тиристори.

Еквивалентната схема при паралелно работещи звена е показана на фиг.4. При еднакви стойности на капацитета на комутиращите кондензатори и началните стойности на напрежението на кондензаторите също са еднакви. В такъв случай в съответствие с теоремата на Тевенен може да се премине към еквивалентна схема на фиг.5. Диаграмите на тока и напрежението са показани на фиг.6.

Изразът за общия ток в съответствие с еквивалентната схема и известния коефициент на периодичност е:

L0 =Lт, R0 =Rт при некомпенсиран товар

L0 =Lek в, R0 =Rek при товарен кръг с индуктивна разстройка

ωизх=ωу изходна (управляваща) честота Изразът за тока през тиристора е:

(4)

(5)

(6)

(7)

Page 60: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 60 -

Очевидно скоростта на нарастване на тока през един тиристор е n пъти по-малка от скоростта на нарастване на общия ток, което е особено важно при схемите с обратни диоди.

Вследствие намаляването на тока през един тиристор и намаляване на скоростта му нарастване има възможност да се намали и времето за възстановяване, т. е. да се повиши работната честота, което е второто важно предимство на схемата.

При паралелна работа на инвертори на ток на общ товар за големи мощности освен паралелна работа на инверторите се налага да се използват и n отделни токоизправители защото входният ток, ако е един токоизправител ще бъде n –пъти входният ток на всеки инвертор. Такава схема е показана на фиг.7.

Фиг.7

При такъв вариант на паралелна работа не са необходими разделителни последователни кондензатори и това опростява изключително инверторната схема. Автономните постояннотоковите източници фактически определят режим на зависими инвертори в инверторната част, тъй като напрежението на изходните шини е синусоидално. Въз основа на изведените изводи за външната характеристика, описваща работата на инвертор-токоизправител [Л2] може да се определи входният ток I0 , ако се използва един общ токоизправител и един инвертор:

(9) 0S

qm

0L

E2

cosU

π−β

=

(8)

Page 61: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 61 -

При “n” токоизправители и “ n“ инвертори , работещи паралелно на общи изходни шини, входния ток на всеки инвертор ще бъде ”n” пъти по-малък, откъдето индуктивностите Ls ще бъдат n пъти по-големи:

(10) 0S

qm0

n00201nL

E2

cosU

n

IIII

ω

π−β

====

(11) 0SS nLL =

От тези равенства следва, че за получаване на едни и същи характеристики е по-изгодно от гледна точка на комутационните процеси схеми с паралелна работа на инвертори на общ товар в сравнение с един инвертор с по-мощни тиристори.

4. Заключение 1.Предложен е многостепенен инвертор за подобряване на формата на изходното

напрежение, сьставен от три инвертора на напрежение, свьрзани кьм общ товар и е направен анализ на хармоничния състав на изходното напрежение.

2.Предложено е оригинално схемно решение на резонансен инвертор с паралелно работещи звена и е направен анализ вьз основа на еквивалентната схема.

3.При паралелна работа на инвертори на ток на общ товар за анализа са разгледани отделните инвертори като зависими.

ЛИТЕРАТУРА

1. Д. Каров. Преобразуване на постоянно напрежение от децентрализирани източници с формиране на синусоидално изходно напрежение, доклад на Трета научна конференция по електротехника, електроника и комуникации “Сливен-2004”, ISSN 1312-3920, Известия на ТУ-Сливен, Том 4, 2004.

2. Р. Каров. Теоретични обобщения, дуални съотношения и схемни подобрения на мощни инвертори и звена, Дисертация за дтн., 2003.

3. N. Hingorani, L. Gyugyi. Understanding FACTS, by the Institute of Electrical and Electronic Engineers, 2000.

4. CIGRE TF 38.01.08. Modeling of Power Electronics Equipment in Load Flow and Stability Programs, August 1999

Department of Electrоnics Technical University–Sofia, Plovdiv Branch 61, Sankt Peterbourgh Str. 4000 Plovdiv BULGARIA

Page 62: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 62 -

Page 63: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 63 -

Journal of the Technical University at Plovdiv “Fundamental Sciences and Applications”, Vol. 13(3), 2006 Anniversary Scientific Conference’ 2006 BULGARIA

STAND-ALONE MODULE FOR MULTICHANEL EEG

SIGNAL DIGITAL FILTTERING USING ADSP-21065L

NELI VARDEVA, GALIDIA PETROVA

Abstract. The paper describes a DSP application for EEG digital filtering using Floating point АDSP-21065L. This stand-alone module is designed for real-time digital filtering using Floating point implementation of IIR structure. The coefficients of digital filters are calculated and the program algorithm for separation of four basics components in EEG record - δ (Delta), θ (Theta), α (Alpha) and β (Beta) waves, is designed. The program is simulated with data taken from one channel real EEG record and the clock cycles number for calculating four difference equations for one sample period is estimated. Then the filter system is designed and a program algorithm for processing (filtering) of 16 channels EEG records simultaneously through one digital signal processor is designed. Key words: EEG signal, ADSP-21065L

УСТРОЙСТВО ЗА ЦИФРОВА ФИЛТРАЦИЯ НА МНОГОКАНАЛЕН ЕЕГ ЗАПИС ПОСРЕДСТВОМ ADSP-21065L

1. Въведение

Електроенцефалографията (ЕЕГ) е метод за изследване на биопотенциалите, генерирани от мозъчната кора, при който се снемат, усилват, регистрират и обработват променливотоковите биоелектрични сигнали снети от повърхността на главата. Това са напрежения (сигнали) с непостоянен характер и с малки амплитуди от няколко µV до 100-200µV при нормални сигнали и до 500 µV при сигнали съпътстващи нарушения в централната нервна система. ЕЕГ сигналът е нискочестотен, като честотата на разпространение най-често е в честотния обхват е от 0,5 до 25 Hz и по-редки случаи до 100Hz. Реалният ЕЕГ сигнал е комплексен сигнал, съставен от четири основни типа вълни, които носят най-съществена информация за състоянието на ЦНС и множество шумове и артефакти. Основните съставни са δ (Delta) вълните с честота на разпространение 0,5÷4 Hz, θ (Theta) - 4÷8 Hz , α (Alpha) - 8÷13 Hz и β (Beta) - 13÷25 Hz. Разчитането на ЕЕГ сигнала представлява оценка на количественото отношение на δ, α, θ и β вълните за единица време [1].

Отделянето на четирите основни съставни се осъществява посредством

нискочестотни лентови филтри, поставени преди регистриращото устройство. Поради

посочените особеността на ЕЕГ сигнала (припокриващите се гранични честоти на отделните ритми и тесния честотен обхват на разпространение) (фиг.1) тези филтри трябва да са

Page 64: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 64 -

лентови, от 3-4 ред (със стръмност на преходната област 60-80 dB) и много тясна лента на пропускане. Всички тези изисквания определят в съвременните ЕЕГ апарати да се предпочита отделянето на специфичните вълни да се извършва с цифрови филтри.

Фиг. 1. АЧХ на четирите филтъра за отделяне на съставните вълни на ЕЕГ сигнала

2. Проектиране цифров филтър за ЕЕГ сигнали

Проектирането на цифров филтър се осъществява на няколко етапа. Първата задача е да се определи реда на филтъра и избере типа на апроксимацията за извеждане предавателните функции на четирите аналогови филтъра (аналогови прототипи). Съобразено с изискванията и с избрания ред на филтъра – 4 (стръмност 80dB/dec), най-подходяща за случая е апроксимация по метода Чебишев-1. АЧХ на филтъра на Чебишев-1 има висока стръмност на преходната област (ПО) и липсват пулсации в лентата на задържане. Предавателните функции на четирите аналогови филтъра имат следния общ вид (1):

1( )

( )k k

H pp p

=−

∏ (1)

където pi са полюсите на предавателната функция. За улеснение при проектиране на аналоговия прототип в началото се работи с нискочестотен филтър с честота на сряза 1rad/sec, стръмност на ПО 80dB/dec и неравномерност в областта на пропускане 3dB. Затова след намиране полюсите е необходимо да се развие предавателната характеристика на нискочестотен лентов филтър с предварително зададените стойности за граничните честоти. Този преход НЧФ→НЧЛФ се прави посредством полагане

2 2

0pp

Bp

ω+→ , където (2)

Където 2 1

2

0 1 2

B ω ω

ω ω ω

= −

= (3), (4)

а ω1 и ω2 са съответно долна и горна гранична честота. След заместване на (2) в (1) получаваме предавателната функция на НЧЛФ (5):

Page 65: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 65 -

2 2

0

1( )

( )ЛФ

k

k

H pp

pBp

ω=

+−

∏ (5)

След извеждане на равенство (5) следва намиране предавателната характеристика на цифровите лентови филтри във функция от променливата z, намиране коефициентите на филтрите и общото разликово уравнение. Преобразуването на H(p) в H(z-1) се осъществява със съответните полагания, определени от метода на билинейната трансформация (6).

1

1

2 1

11( ) ( )

zp

Td zH p H z

−=

−+→ (6)

където Тd е периода на дискретизация. След полагане (6) и развиване на уравнението, коефициентите пред

променливата z-1 в числителя и знаменателя на H(z-1) са търсените коефициенти a и b на цифровите филтри. На базата на намерените коефициенти се разписва и съответното разликово уравнение, описващо основния принцип на работа на четирите цифрови филтъра. То има следния общ вид (7)

1 1

0 1

[ ] [( ) ] [( ) ]N M

изх i вх i изх

i j

U nTd a x n i Td b y n j Td− −

= =

= − − −∑ ∑ (7)

където xвх е стойността на съответния дискрет на входния сигнал, yизх е стойността на съответният дискрет на изходния сигнал.

Уравнение (7) описва рекурсивен филтър. Текущата изходна стойност зависи от текущата входна, умножена по коефициент „a” и предходната изходна, умножена по коефициент „b” [2],[4].

3. Реализация на филтрово устройство посредством ADSP-21065L След намиране коефициентите и разликовото уравнение е извършено схемното и

програмно проектиране на филтровото устройство. Реализирането на цифровите филтри за ЕЕГ сигнали е осъществено посредством специализиран сигнален процесор на Analog Device ADSP-21065L с 32-битова дума за обработка на АЛУ, с възможност за работа с данни с плаваща запетая и работна честота на ядрото – 60MHz [5],[7]. Проектираната филтрова система е представена чрез блоковата схема, показана на фиг.2:

Фиг. 2. Блокова схема на устройство на цифров филтър за ЕЕГ сигнали

Блокът ADC е реализиран с два аналогови-цифрови преобразуватели с

разрядност на квантоване -12 бита, всеки с по 8 входни канала. Подаването на дискретните данни към процесора се осъществява посредством сериен SPI интерфейс. ADSP-21065L има два серийни порта - SPORT0 и SPORT1, всеки с по два RXD и TXD извода. Изборът на канала, от който ще бъдат прочетени данните се осъществява програмно. Преди прочитане на данните, процесорът подава по SPI шината към Control

Page 66: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 66 -

Register на АЦП адреса на избрания канал. След прочитането на дискретните стойности на сигнала в момент nTd, те биват обработени от сигналния процесор. Това се осъществява чрез предварително зададен програмен алгоритъм. След филтриране, изходните данни се подават отново чрез SPI интерфейс към блока DAC. DAC е реализиран с 12-битов ЦАП, чийто изход е свързан с демултиплексор с един вход и 16 изхода. По този начин е възможно да се отделят ЕЕГ сигналите за всеки един от каналите. Честотата на дискретизация на входния сигнал е 300Hz.

4. Програмно осигуряване

Фиг. 3. Диаграма на програмния алгоритъм

Page 67: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 67 -

Програмният алгоритъм за филтрация на ЕЕГ сигналите е създаден на

програмния език С (Фиг. 3). На базата на високата работна честота на ядрото, програмния алгоритъм осигурява за един период на дискретизация обработването на данните от 16 отвеждания. За период с продължителност 3,33333 msec процесорът подава clock сигнал с честота 5кHz към CS извода на АЦП (Фиг. 4). По спадащ фронт на CS се генерира прекъсване и към 32-битов RX буфер в процесора започва подаване на данните за дискретизирания сигнал. Запълването на буфера започва от старшия бит. В STCTLx регистъра на процесора предварително се задава дължината на думата, която може да бъде от 3 до 32 бита.

Филтрирането се реализира чрез алгоритъм, който осигурява при всяко едно DMA прекъсване да бъдат прочетени текущите данни за сигнала и последователно да бъдат подадени за решаване на четирите разликови уравнения. Всяко уравнение представлява съответния лентов филтър, отделящ определена вълна от комплексния сигнал. След решаване на уравненията изходните данни се подават към един общ DAC регистър и от там към ЦАП. При всяко едно прекъсване се задава и съответната комбинация за управляващите входове (А0÷А3) на демултиплексора. Целта е на изхода да има комплексен ЕЕГ сигнал, съставен от четирите основни вълни. Възможно е и извеждането на всяка една съставна по отделно.

Фиг. 4. Времедиаграми на дискретизиране на ЕЕГ сигналите от 16 канала с честота 300 Hz.

3. Резултати

За входни данни при симулиране работата на проектирания филтър са взети стойности от реален ЕЕГ запис с продължителност 10 секунди. От Фурие анализа на входния сигнал се вижда, че освен основните съставни (преобладаващи алфа вълни) се наблюдава и допълнително внесен шум от захранващата мрежа с честота 50 Hz (Фиг. 5). След прилагане на филтъра за отделяне на α-вълните се получават резултатите, които могат да се видят на фиг. 6. По аналогичен начин е приложена и филтрация за отделяне на останалите ЕЕГ съставни.

Page 68: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 68 -

Фиг. 5. Фурие анализ на ЕЕГ запис от първи канал

Фиг. 6. Фурие анализ на изходния сигнал, след филтриране на входния сигнал с цифров

филтър за отделяне на α-вълните

Page 69: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

69

Copyright © 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

4. Заключение От направените опити може да се заключи, че при реализирането на цифрови

филтри за ЕЕГ сигнали, посредством специализирани сигнални процесори (DSP) се постигат високо-стойностни резултати от диагностична гледна точка [6]. Работната честота на ядрото на ADSP-21065L - 60 MHz позволява реализирането на четирите филтъра за отделянето на основните съставни (δ, α, θ и β) с един единствен процесор. Времето за което се филтрира входен сигнал от един канал (с период на дискретизация 300Hz) е 9,4μsec. От тук може да се направи заключението, че с използването на DSP е възможно обработването на ЕЕГ сигнали, снети от 16 канала, в рамките на един преиод на дискретизация, който е 3,3 msec

ЛИТЕРАТУРА 1. Иван Даскалов, Илион Стамболиев. Електромедицинска диагностична техника,

София 1978. 2. Р. Иванов. Цифрова обработка на едномерни сигнали, Габрово 1999. 3. Е. Шойкова. Синтез на активни филтри, ТУ София 2000. 4. Г. Лем АналоговьIе и цифровьIе филтрьI 5. Руководство пользователя по сигнальным процессорам семейства SHARC, Санкт-

Петербург 2002. 6. Дан Кинг, Грег Гирлинг, Кен Воурин, Ноам Левин, Джесс Моррис, Уолт

Кестер.Аппаратура цифровых сигнальных процессоров 7. Analog Devices. ADSP-21065L SHARC DSP User’s Manual Department of Electrical Engineering Technical University–Sofia, Plovdiv Branch 25, Tsanko Dystabanov Str. 4000 Plovdiv BULGARIA E-mail: : [email protected]

Page 70: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 70 -

Page 71: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 71 -

Journal of the Technical University at Plovdiv “Fundamental Sciences and Applications”, Vol. 13(3), 2006 Anniversary Scientific Conference’ 2006 BULGARIA

SOFTWARE DEVELOPMENT OF SUN TRACKER REAL

TIME CONTROL ALGORITHM

ALEKSANDAR GEORGIEV, RUMEN POPOV, MARIN PAMUKCHIEV, KLIMENT PETKOV

Abstract: The software realization in C – Assembler environment of real time control algorithm for two axes sun tracker is presented. HC11 microcontroller based platform is used. The software provides service of peripherals, such as real time clock, communication port, digital inputs – outputs, keyboard and indicators, as well as implementation of online sun position calculation procedures and the current manipulator axes position evaluation.

ПРОГРАМНА РЕАЛИЗАЦИЯ НА АЛГОРИТЪМ ЗА УПРАВЛЕНИЕ В РЕАЛНО ВРЕМЕ НА УСТРОЙСТВО ЗА СЛЕДЕНЕ НА СЛЪНЦЕТО

1. Въведение

Реализацията на устройства за следене на слънцето (УСС) се базира основно на три принципа. При единият от тях пасивни системи следят слънцето без никакъв електронен контрол или мотори (например Zomeworks [1]) При електро- оптичните устройства се търси максимума на излъчването на слънцето. Тези устройства дават много добри резултати при добри метеорологични условия. Примерни реализации, илюстриращи този принцип са описани в Rumala [2], Lynch [3] и Poulek и Libra [4]. Работата на тези УСС има сериозни проблеми при облачни дни. При третия принцип микропроцесорите контролират следящите системи използвайки математични формули за предсказване позицията на слънцето и не е нужно да го следят. За определяне на позицията му те използват стъпкови мотори или оптични енкодери. Те често се използват в големи системи, където един контролер управлява много съоръжения [5]. Много УСС, контролирани с микропроцесори. използват електро- оптичните сензори за собствена калибровка [6]. Съществуват системи, които комбинират втория и третия принципи [7]. Допълнителен обзор на съществуващите системи е публикуван от Myers [8]. Достатъчно висока точност на следене осигуряват електро- оптичните и предсказващите системи, като последните работят при всякакви метеорологични условия. Затова в нашия проект за реализация на устройство за следене не слънцето сме се спрели именно на този вариант [9].

Ключов момент при създаването на УСС е изборът на алгоритъм за предсказване на позицията на слънцето в координатната система на наблюдателя.

Page 72: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 72 -

За пример, изчисленията на Михалски [10] са ограничени в периода 1940 до 2050 и са с неточност по-голяма от ±0.01˚, изчисленията на Бланко - Мюриел [11] са ограничени в периода от 1999 до 2015 с неточност по - голяма от ±0.01˚.

Пример, показващ важността от повишаването на точността в измерването на слънчевата позиция до грешка по-малка от ±0.01˚, е калибровката на пиранометрите, които измерват глобалното слънчево излъчване. По време на калибровката, чувствителността на пиранометъра се изчислява при зенитни ъгли от 0˚ до 90˚ с делене на изходното му напрежение на опорната глобална слънчева радиация (G), която е функция на косинуса от зенитния ъгъл (cosθ). Големина на грешката, със стойност 0.01˚ може да доведе до неточност при пресмятането на cosθ, G и оттам на чувствителността. Неточността в cosθ експоненциално нараства (Фигура 1.) ако θ достигне 90˚ (при θ = 87˚, неточността в cosθ е 0.7% което може да се отрази като неточност 0.35% в пресмятането на G; понеже при такива големи зенитни ъгли нормалното моментно лъчение е приблизително равно на половината от стойността на G). При това нараства нуждата от използване на алгоритъм за определяне позицията на слънцето (АОПС) с по – малка неточност за потребителите, които са заинтересувани да измерят глобалното слънчево лъчение с малки неточности в пълния зенитен диапазон от 0˚ до 90˚.

Фигура 1. Неточност на косинуса на слънчевия зенитен ъгъл, получен при 0.01° и 0.0003° неточност в изчисляването на ъгъла.

В този проект използваме АОПС за пресмятане на слънчевите зенитен и

азимутен ъгли с неточност по- малка от ±0.0003˚ в периода от -2000 до 6000 година, реализиран на програмен език СИ и публикуван в [12]. Този алгоритъм в оригиналния си вариант е компилиран за платформата на персонален компютър. В нашия случай, имайки предвид, че системата ни е с автономно захранване, значително по- удачно е базирането и върху едночипов микроконтролер. Това предполага редица ограничения, свързани с обема на програмната памет, паметта за данни, бързодействието и организацията на изпълнението на процедурите в режим на реално време. Непрекъснатото обслужване на процедурите на операторския интерфейс, предсказващите слънчевата позиция изчисления, управлението на осите на

Page 73: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 73 -

манипулатора и комуникацията с външни устройства налагат организацията на многозадачно операционно ядро, работещо в режим на времеделене.

Изчислителният блок е изграден на базата на едночиповия микроконтролер на Motorola 68HC11A1FN и е описан подробно в [13] и е показан на фиг. 2.

Фигура 2. Изчислителен блок

2. Описание на програмната реализация

Структура на многозадачното ядро

Използваната многозадачна система позволява превключване между отделни задачи за малки интервали от време (около 2 милисекунди) с цел повишаване устойчивостта на системата към всякакъв вид въздействие от страна на потребителя и прекъсванията, които могат да възникнат. Освен това то предоставя възможност за „паралелно” изпълнение на различни действия без опасност от загуба на информация, както и осигурява свободата на потребителя да има пълен достъп до системата в реално време.

При организацията на многозадачното ядро на системата за реално време с микроконтролер е изпълнено следното:

− организиран е виртуален стек, разположен в RAM памет, в който всяка задача разполага използваните от нея локални и глобални променливи, както и системният контекст, позволяващ й след смяна на следваща задача и повторно извикване да продължи своето изпълнение без загуба на каквито и да е данни.

− създаден е диспечер на задачите, който прави превключването от приключила задача към задача, която предстои да се стартира.

− реализиран е виртуален процесор който се грижи за правилното управление на задачите и следи за тяхното изпълнение.

− организирани са два режима на работа на операционната система – системен и потребителски.

Основни и допълнителни задачи към многозадачното ядро Основните задачи в системата за реално време са:

− Потребителки интерфейс – извършва обработка на данните, постъпващи от клавиатурата – обработва натиснатите клавиши, избор на опция от зададеното меню и обръщение към съответните системни функции свързани с менютата.

Page 74: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 74 -

− Обслужване на АЦП – обработва всички постъпили аналогови данни и при необходимост извиква функции по управление на периферни устройства, или предава/записва част от данните.

− Пресмятане на текуща позиция на слънцето – пресмята текущата позиция и при необходимост подготвя новите данни за преместване на двата двигателя, за установяване на новата позиция. Второстепенните задачи (извикват се от някоя от основните задачи), които се

изпълняват са:

− Обслужване на сериен интерфейс – предава данни по сериен интерфейс без загуба на информация и позволява изпълнението на „паралелни” процеси докато трае предаването. Функционалната схема на алгоритъма на работа на многозадачното ядро е

показана на фиг. 3.

Фигура 3. Функционална схема на алгоритъма на работа на многозадачното ядро.

Page 75: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 75 -

Разпределение на паметта.

Реализацията на контролера на базата на вградена система налага оптимално използване на нейните ресурси с цел осигуряване на нормална и надеждна работа. В Таблица 1. е представена карта на разпределението на паметта в адресното пространство на 68HC11A1. Таблица 1. Разпределение на паметта в адресното пространство на 68HC11A1

Начален адрес

Краен адрес

Описание Забележка

0000 00FF Памет за данни (RAM) Вътрешна

0100 0FFF Памет за Програми (ROM) Външна

1000 103F Регистри обслужващи системните функции на HC11

Могат да бъдат преместени !

1040 7FFF Памет за програми (ROM) Външна

8000 9FFF Памет за данни (RAM) Външна

A000 B5FF Памет за програми (ROM) Външна

B600 B7FF EEPROM на HC11

B800 FFD5 Памет за програми (ROM) Външна

FFD6 FFFE Вектори на прекъсвания обслужващи различни системни функции

Запазено адресно пространство !

В гореописаните адресни пространства сумарно се получава следният обем

памет:

− RAM – 8448 байта

− ROM – 56511 байта

− EEPROM – 512 байта

Основни модули в проекта на програмното осигуряване:

− eeprog – осигурява функции за запис и изтриване на данни в EEPROM. Функциите в модула са: ee_save_byte – подготвя байт за запис; ee_save_word – подготвя два байта за запис; ee_save_float – подготвя четири байта за запис; Erase_Byte – изтрива байт от паметта; Prog_Byte – програмира байт в паметта.

− init_c – съдържа функцията init_c, която извършва началното установяване на подсистемите в HC11.

− int_oc2 – съдържа функцията int_TOC2, която служи за установяване основния времеви интервал в системата за реално време. В тази функция диспечера на задачите сменя старата с нова задача.

− spa – изчислява текущите предсказани стойности на слънчевия зенитен и азимутен ъгли на базата на GPS координатите на точката на наблюдение и местното време.

− stepp – управлява движението на стъпковите двигатели на манипулатора по време на първоначалното установяване и в режим на следене.

− irq_ – модула прави обработка на маскируемото прекъсване IRQ в HC11, което в настоящия проект се управлява от часовник за реално време (RTC). Съдържа

Page 76: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 76 -

функцията irq_int, която записва данните за текущото време и при необходимост подготвя за установяване новите дата и час към RTC.

− keyb__ – съдържа функцията key_fil, която определя има ли натиснат, прочетен или приет клавиш и установява флага за натиснат клавиш.

− pr_int – съдържа функциите за печат на цели и дробни числа print_int_ и print_float.

− read_adc – модула се използва за четене на стойности от АЦП. Функцията read_adc – чете данни от АЦП. Функция send – изпраща подаден брой байтове четейки от буфер с подаден начален адрес. Функция resive – приема определен брой байтове и ги записва в буфер започвайки от установен начален адрес

− read_key – съдържа функцията read_key която определя кой бутон от клавиатурата е бил натиснат

− read_rtc – чете/записва информация от/в RTC. Функция w1rtc записва един байт на посочен адрес. Функция r1rtc чете един байт от посочен адрес. Функция rburst чете поредица от байтове от посочен адрес. Функция wburst записва поредица от байтова на посочен адрес.

− read_int – модула прочита въведени числа от клавиатурата. Функцията read_int_ чете въведено цяло число. Функция read_lim_int чете цяло число което трябва да бъде в зададен интервал. Функция read_float чете фробно число от клавиатурата.

− ref_ind – опреснява клавиатурата и индикацията Съдържа функцията ref_ind, която опреснява данните на поредния индикатор от динамичната индикация и прочита една колона от клавиатурата.

− ref_izh – съдържа функцията ref_izh, която опреснява данните на цифровите входове и изходи на всеки 2 mS .

− sci_fun – съдържа основни функции за работа със сериен комуникационен интерфейс (СКИ). Функцията sci_interrupt обработва прекъсването на СКИ. Функцията write_sci – изпраща данни по СКИ. Функцията read_sci чете данни от СКИ. Функцията print_string_sci изпраща поредица байтове по СКИ. Функцията init_sci извършва начално установяване на СКИ.

− sci_task – обслужва второстепенната задача по СКИ. Съдържа функцията smuk, която потготвя байтовете за изпращане и извиква функцията print_string_sci.

− set_tym – показва текущото време и ако потребителя потвърди го запазва, в противен случай потребителя въвежда текущ час и дата посредством функцията set_tyme_fun.

3. Резултати

При провеждането на първоначалните тестове бяха проверявани резултатите от изчисляването на Юлиянския ден, представени в Таблица 2. и времето на изгрева и залеза, представени в Таблица 3.

Таблица 2. Примери за тестване при изчисляване на JD

Дата UT JD Дата UT JD

1Януари, 2000 12:00:00 2451545.0 31 Декември, 1600 00:00:00 2305812.5

1Януари, 1999 00:00:00 2451179.5 10 Април, 837 07:12:00 2026871.8

27 Януари, 1987 00:00:00 2446822.5 31 Декември, -123 00:00:00 1676496.5

19 Юни, 1987 12:00:00 2446966.0 1Януари, -122 00:00:00 1676497.5

Page 77: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 77 -

27 Януари,1988 00:00:00 2447187.5 12 Юли, -1000 12:00:00 1356001.0

19 Юни, 1988 12:00:00 2447332.0 29 Февруари, -1000 00:00:00 1355866.5

1Януари, 1900 00:00:00 2415020.5 17 Август, -1001 21:36:00 1355671.4

1Януари, 1600 00:00:00 2305447.5 1Януари, -4712 12:00:00 0.0

Таблица 3. Резултати от сравнението на Астрономическия Аалманах [14] и АОПС за изгрев и залез от Гринуич в 0 – UT

Изгрев Залез Дата Географска ширина на наблюдателя АА АОПС АА АОПС

2 Януари, 1994 35° 7:08 7:08:12.8 17:00 16:59:55.9

5 Юли, 1996 -35° 7:08 7:08:15.4 17:00 17:01:04.5

4 Декември, 2004 -35° 4:39 4:38:57.1 19:02 19:02:2.5

Забелязва се достатъчно добра сходимост на резултата, която напълно

съответства на резултатите, публикувани в [12]. 4. Заключение

Програмната реализация на АОПС върху платформа, базирана на микроконтролер и функционираща в режими на многозадачност и реално време дава възможност за създаване на високо- точни устройства за следене на слънцето, подходящи за целите на измерването на нивото на слънчева радиация.

Модификацията на програмното осигуряване чрез добавяне на алгоритми за предсказване на позицията на спътници би дала възможност разработения контролер да се ползва и за решаване на комуникационни задачи.

ЛИТЕРАТУРА

1. Zomeworks Corporation, Passive solar TRACK RACK, Albuquerque, NM, US Pat. No.

4275712. 2. Rumala SSN. A shadow method for automatic tracking. Solar Energy 1986;37(3):245–7. 3. Lynch WA, Salameh ZM. Simple electro-optically controlled dual-axis sun tracker. Solar

Energy 1990;45(2):65–9. 4. Poulek V, Libra M. New solar tracker. Solar Energy Mater Solar Cells 1998;51(2):113–

20. 5. Gay CF, Yerkes JW, Wilson JH. Performance advantages of two-axis tracking for large

flat-plate photovoltaic energy system. In: Proceedings of 16th IEEE Photovoltaic Specialists Conference, 1982. p. 1368–71.

6. Edwards BP. Computer based sun following system. Solar Energy 1978;21(6):491–8. 7. Georgiev, A., Roth, P., Olivares, A. Sun Following System Adjustment in UTFSM.

"Energy Conversion and Management", 2004, V. 45 (11-12) p. 1795-1806. 8. Myers DR, Stoffel TL, Reda I, Wilcox SM, Andreas AM. Recent progress in reducing

the uncertainty in and improving pyranometer calibrations. J Sol Energy Eng 2002;124:44–50.

Page 78: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 78 -

9. Кинова, Т. К. Контролер за управление на слънчев тракер. Дипломна работа. ТУ София, филиал Пловдив, фак. ФЕА, кат. ОЛТ, 2005.

10. Michalsky, J. J. “The Astronomical Almanac’s Algorithm for Approximate Solar Position (1950-2050)”. Solar Energy. Vol. 40, No. 3, 1988; pp. 227-235.

11. Blanco-Muriel, M., et al. “Computing the Solar Vector”. Solar Energy. Vol. 70, No. 5, 2001; pp. 431-441, 2001.

12. Ibrahim R. Afshin A., Solar Position Algorithm for Solar Radiation Applications National Renewable Energy Laboratory, Technical Report № NREL/TP-560-34302 Revised November 2005.

13. R. Popov, A. Georgiev, M. Pamukchiev, Y. Skerletov, T. Kinova. Controller for direction of a sun-tracker. Accepted for publication in “Journal of the Technical Univ.at Plovdiv” 2005.

14. The Astronomical Almanac. Norwich: 2004. Departments of “Mechanics” and “Optoelectronics and Laser Engineering” Technical University–Sofia, Plovdiv Branch 25, Tsanko Dystabanov Str. 4000 Plovdiv BULGARIA E-mail: [email protected]

Page 79: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 79 -

Journal of the Technical University at Plovdiv “Fundamental Sciences and Applications”, Vol. 13(3), 2006 Anniversary Scientific Conference’ 2006 BULGARIA

A UNIFIED ANALYSIS OF A SERIES RESONANT

CONVERTER

TSVETANA GRIGOROVA – SHTARBEVA

Abstract: The paper presents a unified analysis of a series resonant converter, operated above or below the resonant frequency. On the basis of the proposed analysis, the currents and voltages universal and normalized graph are given as a function of the

voltage ratio in' gU/U0 under the different values of the parameter ν. The criteria for the

optimal choice of the using design coefficients are proposed. The given analytically calculations are confirmed by computer simulation, using OrCad PSpice. Key words: Series resonant converter, simulation OrCAD PSpice

ОБОБЩЕН АНАЛИЗ НА ПОСЛЕДОВАТЕЛЕН DC/DC ПРЕОБРАЗУВАТЕЛ

1. Въведение

Последователните DC/DC транзисторни преобразуватели, работещи на честоти по-високи от резонансната, намериха широко приложение в различни индустриални области. Това се дължи основно на предимствата, които осигурява този режим на работа, а именно възможността за превключване при нулев ток (ZCS) и нулево напрежение (ZVS) [1],[2],[3]. За този клас преобразуватели са предлагани различни методи за анализ и проектиране [4],[5].

Целта на настоящата статия е да се предложи oбобщен анализ и метод за проектиране на последователен DC/DC преобразувател, работещ на честоти по-високи или по-ниски от резонансната, с помощта на който се получават удобни зависимости за инженерни пресмятания. Предложени са критерии за оптимален избор на използваните коефициенти.

2. Основни съотношения в преобразувателя

На Фиг.1 е показан полумостов DC/DC последователен преобразувател. В процеса на анализa са направени следните допускания: транзисторните ключове,

диодите и трансформаторите са идеални; капацитетът 0С в изходния филтър е

безкрайно голям, следователно, изходното напрежение 0U е константа.

Page 80: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 80 -

Фиг.1 Полумостов последователен DC/DC преобразувател

В зависимост от режима на работа на преобразувателя, паралелно на

транзисторите се включват подходящи демпферни групи. В анализа тяхното влияние е пренебрегнато.

Въведени са и следните означения: - Sω - управляваща честота;

-

rrCL

10

=ω - собствената резонансна честота; - tQ 0�=λ - ъгъл на провеждане на

транзисторите; - tD 0�=λ ъгъл на провеждане на диодите; -inU - входно напрежение;

- ν ( 0ωων /S= ) - честотен коефициент; - 00 kUU' = e приведеното изходно напрежение, а

k е коефициентът на трансформация; - g е топологична константа (g =1 за мостова и g =0.5 за полумостова схема ).

Диаграмите, описващи процесите в схемата за честоти 0ωω >S са илюстрирани

на Фиг.2а, а тези описващи процесите в схемата за честоти 02

1ωω >S са илюстрирани

на Фиг.2б.

Фиг.2а Времедиаграми за честоти 0ωω >S Фиг.2б Времедиаграми за честоти

02

1ωω >S

Page 81: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 81 -

Принципът на работа на преобразувателя е детайлно разгледан в литературата.

Въз основа на анализа, направен в [7] се получават следните обобщени изрази, поместени в таблица 1, за основните параметри на силовата схема.

Таблица1. Основни съотношения в преобразувателя

ν

π+

ν

π

22

0 sinsiningU

'U

arcsin (1)

Dλ Qλ−ν

π (2)

N

λ

λ

22

1

DQtantan

(3)

'QavI

( )

+

−π

ν=

in

'Qav

gU

U

NI

gI01

1m (4)

'DavI

( )

−π

ν=

in

'Dav

gU

U

NI

gI 011

m (5)

avin'

I

( ) in

'inav

gU

U

N

g

I

gI 0

1

4

−π

ν= m (6)

'QmI

( )( )

+

−λ−=

in

'

Q

Qm

gU

U

Ncos

g

I

gI01

11

2m (7)

'DmI

( )( )

−λ−=

in

'

D

Dm

gU

U

Ncos

g

I

gI 0111

2m (8)

'CoU

( ) in

'

gU

U

N

g 0

1

2

−m (9)

Cm'U

1

2

−N

gm (10)

'inP

( ) in

'in

gU

U

N

g

P

gP 0

1

4

−π

ν= m (11)

Знакът минус в изразите се отнася за режим на работа при честоти 02

1ωω >S .

Решаването на уравненията от табл. 1 позволява да се построят характеристиките на

Page 82: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 82 -

0.0 0.2 0.4 0.6 0.8 1.0

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

2.4

2.6

2.8

3.0

νννν=1.3

νννν=1.3νννν=1.25

νννν=1.2

νννν=1.15

νννν=1.1λλλλ

Q, rad

U'0/gU

in

0.0 0.2 0.4 0.6 0.8

0.0

0.5

1.0

1.5

2.0

2.5

3.0

I'Qav

ν=1.35

ν=1.3

ν=1.25

ν=1.2

ν=1.15

ν=1.1

U'0/gU

in0.0 0.2 0.4 0.6 0.8 1.0

0

1

2

3

4

5

6

ν=0.7

ν=0.55ν=0.6

ν=0.8

ν=0.9

ν=0.95I'QAV

U0'/gU

in

преобразувателя в зависимост от отношението in'

gU/U0 за различни честотни

коефициенти ν. Графичното представяне на зависимостите ъгъл на провеждане на

транзисторите ( )ν=λ ,gV/Vf in'

Q 0 и средна стойност на тока през транзисторите

( )ν= ,gU/UfI in''

Qav 0 , за двата режима на работа, е показано на Фиг.3 и Фиг.4.

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

1.6

1.8

2.0

2.2

2.4

2.6

2.8

3.0

U0'/gU

in

ν=0.55

ν=0.6

ν=0.7

ν=0.8

ν=0.9

ν=0.95

λλλλQ,rad

Фиг. 3 Зависимост на ъгъла на провеждане на транзисторитеQλ от

in' gU/U0

а). за честоти 0ωω >S б). за честоти 0

2

1ωω >S

Фиг. 4. Зависимост на средната стойност на тока през транзисторите 'QavI от in

' gU/U0

а). за честоти 0ωω >S б). за честоти 0

2

1ωω >S

Методика за проектиране на DC-DC последователен преобразувател, работещ при честоти по-високи от резонансната честота е разгледан в [7]. Въз осова на предложения алгоритъм за работа и резултатите, получени в таблица 1, може да бъде проектиран последователен DC/DC преобразувател при известни входно напрежение Uin, изходно напрежение U0, изходен ток I0, управляваща честота fs.

Page 83: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 83 -

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

0.0

0.5

1.0

1.5

2.0

P'in

ν=1.35

ν=1.3

ν=1.25

ν=1.2

ν=1.15

ν=1.1

U'0/gU

in

От уравнение (11) може да бъде построена зависимостта на входната,

респективно изходната постояннотокова мощност, от отношението in' gU/U0 при

параметър ν, т.е. ( )ν,gU/UfP in''

in 0= . Резултатите са илюстрирани на Фиг. 5.

Препоръчва се съгласно [6] товарът да се избира, така че номиналната работна

точка да съответства на максимума на характеристиката ( )ν,gU/UfP in''

in 0= .

0.0 0.2 0.4 0.6 0.8 1.0

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

ν=0.6

ν=0.7

ν=0.8

ν=0.9

ν=0.55

ν=0.95P'

in

U'0/gU

in

Фиг.5 Зависимост на входната (изходната) постояннотокова мощност ( )ν,gU/UfP in''

in 0=

а). за честоти 0ωω >S б). за честоти 0

2

1ωω >S

3. Симулационни резултати

За проверка на предлаганата методика, с помощта на OrCAD Pspice, е симулирана схема на полумостов последователен DC/DC преобразувател (g=0.5) при

следните входни данни: VU in 300= , VU 1000 = , kHzfS 80= . Като ключови елементи са

избрани MOSFET транзистори тип IRF730. Стойностите на последователните елементи

са HLr µ200= и nFCr 33= . Използвани са беззагубни демпферни групи С1÷С4,

паралелно свързани на транзисторите със стойност 1nF. Избран е ν=1.3. Резултатите от симулацията са показани на фиг.6.

Фиг.6. Симулационни резултати: ток през резонансния кръг ri ; напрежение върху

последователния кондензатор Cu ; напрежение DSu върху транзистор Q1

Page 84: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 84 -

Преобразувателят е проектиран, с помощта на предложената в [7] методика. Сравнителните резултати между теоретичните изчисления и тези, получени от компютърната симулация са показани в таблица 2.

Таблица 2

Qλ ,rad Dλ ,rad QmI ,A CoU ,V CmU ,V

Изчислено 1,886 0,538 3 122,3 183,5

От PSpice 1,895 0,529 3,09 122,7 191,6

Вижда се изключително добро съвпадение между изчислени и симулирани резултати. Грешката е в рамките на 4%, което е приемлив резултат за инженерни изчисления.

4. Изводи

В статията е разгледан обобщен анализ на транзисторен резонансен DC/DC преобразувател, работещ на честоти, по-високи или по-ниски от резонансната, с помощта на който се получават изрази, удобни за инженерни пресмятания. Представени са графични зависимости на токовете и напреженията в схемата от

отношението in'

gU/U0 при различни стойности на параметъра ν, което

благоприятства избора на работния режим. Предложени са критерии за оптимален избор на използваните коефициенти. Определени са ъглите на провеждане на

транзисторите и диодите, като функция на отношението in' gU/U0 и коефициента ν.

Чрез компютърна симулация с OrCad PSpice са потвърдени получените аналитични пресмятания, като точността е в рамките на 4%.

ЛИТЕРАТУРА

1. M.J. Schutten, R. L., Steigerwald. Characteristics of load resonant converter operated in high power-factor mode, IEEE Trans. Power Electron, vol.7, pp.304-314, 1992.

2. A.C Lippincott, R. M. Nelms. A capacitor-charging power supply using a series resonant topology, constant time/variable frequency control, and zero-current switching” IEEE Trans. Ind. Electron, vol.38, pp.438-447, 1991.

3. L. Malesani, P.Mattavelli and other. Electronic welder with high frequency resonant inverter, IEEE Trans. Ind. Application vol.31, pp.273-279, 1995.

4. K. Alhaddad, K., Y. Cheron, H.Foch, V.Rajagopalan. Static and dynamic analysis of a series resonant converter operating above its resonant frequency, SATECH'86 Proceedings, Boston, pp. 55-68, 1986.

5. G. Ivensky, I.Zeltser, A. Kats, S. Ben-Yaakov. Reducing IGBT losses in ZCS series resonant converters, IEEE Transaction on Ind. Application, vol.33, pp. 68-74, 1996.

6. N. Bankov, V.A. Kostova. Load Characteristics of a Transistor Inverter Functioning with a Frequency Higher than the Resonance One, Electronics ET’2001,Sozopol, pp. 161- 165, 2001.

Page 85: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 85 -

7. Цв. Григорова. Модифициран анализ и проектиране на последователен DC/DC преобразувател, работещ на честоти по-високи от резонансната, сп. “Е+Е”, бр.11/12, стр.42 – 48, 2005.

Department of Electrical Engineering Technical University–Sofia, Plovdiv Branch 25, Tsanko Dystabanov Str. 4000 Plovdiv BULGARIA E-mail: [email protected]

Page 86: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 86 -

Page 87: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 87 -

Journal of the Technical University at Plovdiv “Fundamental Sciences and Applications”, Vol. 13(3), 2006 Anniversary Scientific Conference’ 2006 BULGARIA

SOFT SWITCHING CAPACITOR VOLTAGE CLAMPED

SERIES RESONANT DC/DC CONVERTER

TSVETANA GRIGOROVA – SHTARBEVA

Abstract: The paper presents variation of the capacitor voltage clamped series resonant converter with improved performances. The switching transistors control algorithm of the shown circuit ensures soft switching conditions. In addition this converter has an inherent power-limiting property. Converter regulation is carried out by frequency modulation. These operation conditions allow the design without any voltage or current overshoot. Steady state and transient equations for different operating stages are presented. The converter output characteristics are drawn. The analysis results are confirmed through the computer simulation and experimental investigation. Key words: Soft switching converters, simulation OrCAD PSpice

DC/DC ПОСЛЕДОВАТЕЛЕН ПРЕОБРАЗУВАТЕЛ С ОГРАНИЧАВАНЕ НА НАПРЕЖЕНИЯТА, ОСИГУРЯВАЩ УСЛОВИЯ ЗА МЕКА

КОМУТАЦИЯ

1. Въведение

С цел предотвратяване възникването на недопустими режими в работата на силовите схеми, се реализират преобразуватели с подобрени товарни характеристики, като стабилност в работа им се постига чрез ограничаване на схемните напрежения [1-2]. Осигуряването на условия за мека комутация е надежден способ за намаляване загубите в преобразувателите на електрическа енергия. В силови схеми, използващи MOSFET и IGBT транзистори като ключови елементи, комутацията при нулево напрежение е ефективен метод за намаляване на загубите, възникнали вследствие на ефекта на Милер [3].

Преобразувателят представен в [3,4] е реализиран на база полумостов резонансен DC/DC преобразувател с разделен последователен кондензатор. Освен това в схемата са добавени диоди, които ограничават напреженията на последователните кондензатори, така че те се зареждат винаги до напрежението на захранващия източник и се разреждат до нула. При такъв режим на работа, напрежението върху кондензаторите има постоянна съставка, което е предпоставка за увеличаване на инсталираната мощност в схемата.

Въз основа схемната реализация [5], е предложен резонансен DC/DC преобразувател с подобрени товарни характеристики [6], при който се избягва

Page 88: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 88 -

посочения по-горе недостатък. В настоящата статия изследванията се разширяват и задълбочават, като са получени изходните характеристики на преобразувателя. Предложена е и алгоритъм за проектиране на този клас силови схеми.

2. Принцип на действие на преобразувателя

Принципната електрическа схемата на полумостов резонансен DC/DC преобразувател с ограничаване на напреженията [6] е показана на Фиг.1, а на Фиг.2-

диаграми илюстриращи нейното действието. Последователният кондензатор С е включен в диагонала на моста, като неговото напрежение се ограничава от диодите D3 и D4.

Фиг.1 Полумостов резонансен DC/DC преобразувател с ограничаване на напреженията

Транзисторите М1 и М2 се управляват чрез асиметрични управляващи импулси

с 50% коефициент на запълване с изключение на преходния процес, когато и двата транзистора са запушени, за да се осигури време за комутация при нулево напрежение.

Фиг.2. Време-диаграми, илюстриращи действието на преобразувателя

Page 89: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 89 -

Анализът на преобразувателя е направен при следните условия: всички елементи в схемата са идеални; резонансната индуктивност L е много по-малка от индуктивността на трансформатора; изходният индуктивен филтър е с достатъчно

голяма стойност за осигуряване на постоянен товарен ток 0I . Регулирането на

изходната мощност се осъществява чрез промяна на управляващата честота [4]. На Фиг.3 е показана принципната електрическа схема при провеждане на

транзистор М1.

M2

M1

LC

D1 Cs1

D2 Cs2

Cf1 D3

D4Cf2

Фиг.3. Принципна схема при провеждане на транзистор М1

Резонансният ток зарежда последователния кондензатор С до достигане на захранващото напрежение, при което се отпушва ограничителният диод D3 (Фиг.4) и се прекратява консумацията на енергия от захранващия източник. Транзистор М1 продължава да е отпушен, като токът през последователната индуктивност L формира контурa D1, M1 и първичната страна на трансформатора. Напрежението върху

трансформатора 1Tu и индуктивността L спада приблизително до нула (Фиг.2).

M2

M1

LC

D1 Cs1

D2 Cs2

Cf1 D3

D4Cf2

Фиг.4. Принципна схема при едновременно провеждане на транзистор М1 и ограничителен диод D3

След запушване на М1 (М2 е все още запушен) се осигурява необходимото

време за мека комутация на транзисторите. Токът, който протича в първичната страна разрежда демпферния кондензатор CS2, след което се комутира към обратния диод D2. По този начин отпушването на М2 ще се осъществи при нулево напрежение. В следващия полупериод започва да провежда М2 и процесите в схемата се повтарят.

Времето, през което и двата транзистора са запушени е много малко в сравнение с управляващия период, поради което може да се приеме, че изходното напрежение е равно на средната стойност на напрежението в първичната страна на трансформатора, приведено към вторичната. Този вид преобразуватели притежават вътрешното свойство

Page 90: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 90 -

да фиксират мощността предавана към товара. Изходната мощност зависи от енергията, натрупана в последователния кондензатор по следния начин [3]

maxinmax fCUP2= . (1)

За осигуряване на комутация при нулево напрежение, е необходима достатъчна

енергия за зареждане на демпферните кондензатори до стойността на захранващото напрежение. Следователно минималната стойност на последователната индуктивност L се определя съгласно условието

2

0

22

'

inSmin

I

UCL = , (2)

където 'I0 е приведеният към първичната страна изходен ток, SSS CCC == 21 .

3. Преходни процеси, осигуряващи условията за мека комутация

Преходният процес на комутация между двата транзистора условно се разделя на три времеви интервала. Резултатите от анализа в [6] са представени в таблица 1.

Таблица1

Интервал- 0÷t1

ω=

'

in

IZ

Uarcsint

0111

1 (1)

Интервал- 21 tt ÷

in

'

in'

U

IZ

UarcsincosLI

t

=01

0

2

(2)

Интервал- 32 tt ÷

ω=

in

'

U

IZarcsint 02

23

1 (3)

където sC

LZ =1 - характеристичен импеданс на трептящия кръг;

sLC

11 =ω -

кръгова резонансна честота за първи интервал; C

LZ =2 - характеристичен импеданс

на трептящия кръг в трети интервал; LC

12 =ω кръгова резонансна честота за трети

интервал .

Page 91: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 91 -

Третият интервал приключва в момента t3, когато резонансният ток достигне

стойността 'I0 . Едва след неговото завършване започва прехвърляне на мощност към

товара. Резонансният ток е

( ) 'L Iti 0= . (4)

Моментните стойности на напрежението на първичната намотка на

трансформатора uT1 и напрежението Bu се дефинират съгласно изрази

( )

−+=ω+=

in

'

in

'

in

'

TU

IZarcsincosUt

C

ItcosUt

C

Itu 020

320

1 , (5)

( ) ( )

−−+=ω−+−=

in

'

in

'

in

'

BU

IZarcsincosUt

C

ItcosUt

C

Itu 020

320 11 . (6)

При inB Uu = за 4tt = се получава

'

in

'

in

I

U

IZarcsincosCU

t

0

02

4

= . (7)

Това е моментът, в който напрежението на кондензатор С достига стойността на

захранващото напрежение, в резултат на което се отпушва ограничителният диод D3. В този интервал напрежението върху трансформатора има триъгълна форма – Фиг. 2. Процесите в схемата продължават отново със съвместното провеждане на М1 и D3.

Ако транзисторите М1 или М2 се запушат в момент, когато напрежението, приложено върху първичната намотка на трансформатора не е станало нула или

напрежението Bu не е достигнало нула или захранващото напрежение в края на

полупериода Т, то тогава [4]

( )3214 tttTt −−−= , (8)

а напрежението върху първичната намотка на трансформатора се определя чрез израза

( ) ( ) tC

I

C

ITUtuUtu

''in

Bin'T

001

42−

⋅+=−= . (9)

Този режим е възможен при малка стойност на изходния ток. Фиксирането на мощността, предавана към товара зависи от процесите на заряд

и разряд на последователния кондензатор.

Ако допуснем, че к.п.д. на преобразувателя 1=η , то енергията, консумирана от

захранващия източник е равна на енергията в изхода

Page 92: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 92 -

( )∫ ==2

0

0001

2

T

S

'T

f

IkUdtItuW , (10)

където Sf - управляваща честота; 21 wwk = - коефициент на трансформация.

След заместване на (5) и (6) в (10) и извършване на несложни преобразувания за изходното напрежение се получава

⋅−

⋅⋅=

in

'inS

U

IZarcsincos

Ik

UCfU 02

0

2

0 . (11)

При малки стойности на изходния ток, изходното напрежение се изчислява чрез

(9).

( )k

Udttu

TkU in

T'T

'

2

2 2

010 =∫

⋅= (12)

Уравнение (12) показва, че изходното напрежение при този режим, не зависи от

честотата и промените в товара. Зависимостите (11) и (12) са представени в графичен вид на Фиг.5.

Uo, V

40 Io

0 1 2 3 40

20

40

60

80

fs = 80kHzfs = 111kHzfs = 150kHzConst.output voltage mode

Фиг.5 Изходни характеристики на преобразувателя

Решаването на уравнение (10) спрямо мощността има вида

⋅−⋅⋅=

in

'

inSU

IZarcsincosUCfP 022

0 , (13)

а в режим на малък изходен ток

000 IUP'' = . (14)

Page 93: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 93 -

Зависимостите (13) и (14) са представени в графичен вид на Фиг.6.

Po, W

40 Io

0 1 2 3 40

50

100

fs = 80kHzfs = 111kHzfs = 150kHzConst.output voltage mode

Фиг.6 Изходната мощност P0 в зависимост от изходния ток Io

при различни управляващи честоти fS

Като се има пред вид направения анализ, може да се препоръча следната

последователност при проектиране на този вид преобразуватели:

− Дефинира се максималната изходна мощност;

− Определят се управляващата честота fS и стойността на последователния кондензатор C;

− Изчислява се стойността на резонансната индуктивност L;

− Необходимото изходно напрежение се определя от (11) или (12);

− Стойността на резонансната индуктивност L се проверява в условието за осигуряване на мека комутация [6] и при необходимост се извършва корекция на стойността.

4. Симулационни и екпериментални резултати

Въз основа на направените разглеждания беше създаден и изпитан DC/DC преобразувателя от фиг.1 при следните изходни данни: захранващо напрежение U1=200V; управляваща честота f=111kHz; изходна мощност Р=60W; елементи на резонансния кръг L= 102µH и C= 20nF; преводно отношение на трансформатора 10:4; демпферни кондензатори CS=330pF. Резултатите от симулация с PSpice са дадени на фиг.7, а на фиг.8 – получените осцилограми на напрежението върху един от транзисторите и последователния кондензатор при номинален товар.

Page 94: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 94 -

Фиг. 7 Резултати от симулация с OrCAD PSpice

Фиг.8. Горе -напрежение дрейн – сорс на транзистора -

uds(M2) 50V/div; Долу – напрежение върху последователния кондензатор С - uC 20V/div

5. Изводи

В статията е предложена разновидност на резонансен DC/DC преобразувател с подобрени товарни характеристики. Алгоритъмът за управление на силовите транзистори в разглежданата схема, осигурява условия за мека комутация. Този вид преобразуватели притежават свойство да фиксират мощността предавана към товара. Въз основа на направения анализ са получени уравненията, описващи работата на силовата схема в преходен и установен режим. Построени са изходните характеристики на преобразувателя. Предложена е и алгоритъм за проектиране на този клас силови схеми. Резултатите от анализа са потвърдени чрез компютърна симулация и експериментално изследване.

Изследването на предлагания преобразувател е по НИД договор № I-447/12.06.06 на тема “Изследване и разработка на силови електронни устройства с подобрени енергетични показатели”.

ЛИТЕРАТУРА

1. J. G. Cho, J. A. Sabate, F. C. Lee. Zero-voltage and zerocurrent switching full bridge

PWM converter for high power applications. IEEE Trans. Power Electron., vol. 11, July 1996, pp. 622–628.

Page 95: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 95 -

2. W. J. Gu, K. Harada. A novel self-excited forward DC-DC converter with zero-voltage-switched resonant transitions using a saturable core. IEEE Trans. Power Electron., vol. 10, Mar. 1995, pp. 131–141.

3. N. K. Poon, M. H. Pong. Computer aided design of a crossing current resonant converter (XCRC). Proc. Ind. Electron. Contr. Inst. Conf., 1994, pp. 135–140.

4. N. K. Poon, Bryan M., H. Pong, Chi K. Tse. A constant-power battery charger with inherent soft switching and power factor correction. IEEE Trans. on Power Electronics, vol. 18, No. 6, 2003, pp.1262-1269.

5. Цв. Григорова, Изследване и симулации на автономни инвертори с ограничителни диоди. Дисертация, Пловдив 2000.

6. Tsv. Grigorova, Soft switching capacitor voltage clamped series resonant DC/DC

converter, сп. “Инженерни науки”, кн.2, стр.35-43, 2006, ISSN 1312-5702. Department of Electrical Engineering Technical University–Sofia, Plovdiv Branch 25, Tsanko Dystabanov Str. 4000 Plovdiv BULGARIA E-mail: [email protected]

Page 96: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 96 -

Page 97: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 97 -

Journal of the Technical University at Plovdiv “Fundamental Sciences and Applications”, Vol. 13(3), 2006 Anniversary Scientific Conference’ 2006 BULGARIA

MATLAB/SIMULINK IGBT MODEL

KATYA ASPARUHOVA, TSVETANA GRIGOROVA, SNEJANA TERZIEVA

Abstract: The paper presents implementation in MATLAB/SIMULINK of IGBT behavioral macromodel. The model describes IGBT transistor in static mode. The required parameters setup in the model is shown too. This is easy for the users and makes the model applicable for different type of IGBTs using only data sheets. The results are compared with the simulations of the same model, implemented in PSpice simulator. Key words: Behavioral model, IGBT, MATLAB, Simulink, simulation.

МОДЕЛ НА IGBT ТРАНЗИСТОР В СРЕДАТА НА MATLAB/SIMULINK

1. Introduction

Recently a simulation method for power electronic devices has emerged, which is able to account for details of the device behaviour in a similar fashion to the finite element method, but with a significantly shorter run-time. MATLAB [4] is a general-purpose mathematical computation package widely used in engineering, science and mathematics. Simulink is a graphical front end for MATLAB, primarily intended for simulation of dynamic systems in classic block-diagram form. The MATLAB/Simulink model realization gives possibilities for optimization and simple parameter extraction.

Only physics-based models have implemented [3] until now in Simulink. This paper presents DC part of the IGBT behavioral Simulink model, which is built using the configuration of the Hammerstein model, consisting of a nonlinear static block followed by a linear dynamic block. The nonlinear DC equations and correction functions are precisely represented using wide computational possibilities of the MATLAB package, which is resulting in a good accuracy.

The parameters and characteristics from datasheets are the starting points for determination the parameters for Simulink simulations. Most of the model parameters are obtained directly from the output characteristics. The verification is accomplished with comparison together the Simulink and the PSpice implementations of the same model.

2. DC Model of IGBT

The DC model is based on the empirical formulas for the IGBT collector current detailed described in [1]-[2]. It combines the equations that circumscribe the MOSFET in cutoff, the linear and saturation regions with the equations of a bipolar junction transistor

Page 98: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 98 -

operating in the active mode. The model accounts for high-level injection and the voltage drop in the extrinsic part of the IGBT. The next expressions give the collector current in all regions:

( )( ) ( )

( )

−+>−

−+<

−−−−

<≤

=

thDGECEthGE

thDGECE

DCEDCEthGE

DCEthGE

C

UUUUifUU

fk

UUUUif

UUfUUfUUfk

UUorUUif

I

,2

.

,2

.

,0

2

2

21

12

(1)

The IGBT output characteristics are different from these of the MOSFET. The output

current in the cutoff region is zero. Typical value of the output voltage is between 0,7V and 1V. The linear region of the IGBT and MOSFET can be assumed to be the same. Only the transition from the linear into the saturation region must be corrected. In order to make the saturation voltage of MOSFET equal to the corresponding IGBT saturation voltage the correction function f1, dependent on gate-emitter voltage, is introduced:

thDGECEsat UUUUf −+=1 , (2)

where thU is the threshold voltage of the MOSFET and UD is the voltage drop across the

emitter-base junction. The correction function f1 is approximated with a third order polynomial of UGE. The

degree of this polynomial depends on the number of the accounted points and required accuracy.

The output current in the saturation region are fitting using the correction function f2:

( )

2

2

2

thGECsat UUk

f

I −= , (3)

where k is the process transconductance parameter and f2 is a third order polynomial of UGE.

3. Simulink Model Implementation

The proposed behavioral model uses the hierarchical visualisation offered by Simulink [4] to simplify its development and appearance. Figure 1 shows the contents of the IGBT model. The implementation requires setting up the relevant physical processes in block diagram form for graphical entry into Simulink. The inputs for the subsystem are VC, VG and VE and IC is considered to be output. Particular blocks of the model are also combined into different subsystems to facilitate the users.

The nonlinear equations in the model are performed using math operation blocks like add, abs, math function, gain, product. The conditions for different operation regions (cut-off, linear and saturation) are realized with logic and bit operations blocks. The constants are setting by constant sources. The correction functions f1 and f2 are performed with polynomial

Page 99: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 99 -

blocks. The polynomial’s coefficients are calculated with MATLAB program for each value of the gate voltage.

Fig.1 Simulink IGBT DC Behavioral Model

Page 100: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

- 100 -

The MATLAB’s workspace format files are provided the set of input DC voltage values and the output results for the current. In this way the further visualizations and calculations using MATLAB are easy to perform.

The model is simple to use and easy to change for different types of IGBTs.

4. Simulation Results

The type CM600-24H IGBT is chosen for the verification of the implemented model Fig. 2 shows the static I-V characteristics offered by the behavioral PSpice macromodel [2] and Fig. 3 these received from the Simulink device model. In the calculation it was assumed that Uth = 6V, UD = 0.6V. The simulations are made under following initiation conditions: UCE

= 0÷12V, UGE = 9÷12V.

Fig.2. Behavioral PSpice Macromodel Simulated Static I-V characteristics for the device

CM600-24H

IC, [A]

0 2 4 6 8 10 120

200

400

600

800

1000

1200

VGE=9V

VGE=10V

VGE=11V

VGE=12V

VCE, [V]

IC,

[A]

Fig.3 Simulink Model Simulated Static I-V characteristics for the device CM600-24H

Page 101: JOURNAL - tu-plovdiv.bg- 6 - and studied such mode of operation for the two modern lasers - CW diode-pumped Yb-doped crystal lasers and the CW red diode lasers [7-9]

Copyright 2006 by Technical University at Plovdiv, Plovdiv, BULGARIA. ISSN 1310 - 8271

- 101 -

The analyses results are summarized in Table I.

TABLE I

IC, A Behavioral PSpice

macromodel

IC, A Simulink Model

UGE,V IC, A Data Sheet

Value Error, % Value Error, %

9 236.269 247.052 4.56 236.7 0.18

10 484.855 478.886 1.23 484.8 0.01

11 795.855 792.413 0.43 795.0 0.11

12 1181.35 1179 0.2 1180.8 0.047

5. Conclusions

The behavioral IGBT Simulink model is presented. The nonlinear DC equations and correction functions are precisely represented, which is resulting in a good accuracy. The model parameters are easy to change for different types of the IGBTs. The static I-V characteristics offered by the model are shown. The good agreement between the simulation results and data sheets can be observed in the paper, as the average error is less than 1%.

A future research direction could be the extension of the model with the blocks describing the dynamic behavior of the device.

REFERENCES

1. K. Asparuhova, T. Grigorova. IGBT Spice Behavioral Model Using the Hammerstein Configuration, 14th International Scientific and Applied Science Conference Electronics ET’2005, Sozopol, Bulgaria, b.1, pp 71-76, 2005.

2. . Asparuhova, T. Grigorova. IGBT Behavioral PSpice Model, 25th International Conference on microelectronics MIEL 2006, Belgrade, Serbia, vol.1, pp 215-218.

3. P.R. Palmer, A.T. Bryant, J. Huigins, E. Santi. Simulation and Optimization of Diode and IGBT Interaction in a Chopper Cell Using MATLAB and Simulink, IAS Conf. Rec., Pittsburgh, October 2002.

4. http://www.mathworks.com

Department of Electrical Engineering Technical University–Sofia, Plovdiv Branch 25, Tsanko Dystabanov Str. 4000 Plovdiv BULGARIA E-mail: [email protected]