MMIC PACKAGE FOR MILLIMETER WAVE FREQUENCY · PDF filenasa-or-205037 final technical report mmic package for millimeter wave frequency contract no: nas 3-27813 prepared for nasa-lewis

NASA-OR-205037

FINAL TECHNICAL REPORT

MMIC PACKAGE FOR MILLIMETER WAVE FREQUENCY

CONTRACT NO: NAS 3-27813

PREPARED FOR

NASA-LEWIS RESEARCH CENTER

21000 BROOKPARK ROAD

CLEVELAND, OH 44135

Sarjit S BharjSteve Yuan

21st June 1997

PREPARED BY

PRINCETON MICROWAVE TECHNOLOGY INC.

3 NAMI LANE, UNIT C-10

MERCERVILLE. NJ 08619

Distribution limited to U.S. Government Agencies only. Other requests for this document

should be referred to NASA-LeRC, Cleveland, OH 44135

https://ntrs.nasa.gov/search.jsp?R=19970023908 2018-04-28T14:41:23+00:00Z

TABLE OF CONTENTS

SECTION

1.0 EXECUTIVE SUMMARY

2.0 INTRODUCTION

2.1

2.2

2.3

2.4

2.5

2.6

2.7

2.8

Substrate Material Measurements

Package Fan out Design

Fabrication

Mechanical Measurements

Electrical Measurements

MMIC insertion into package

Package Layout

Package Data Sheet

3.0 CONCLUSION

References

2

3

5

7

12

14

18

21

22

23

24

i/i

i i ¸

i' i ._!i_i:,

1.0

EXECUTIVE SUMMARY

Under a previous program sponsored by NASA Lewis Research Center (NASA/LeRC) in Cleveland, a

reliable, high performance, low cost package for high frequency communications electronics was

successfully developed. The package has many unique features which include:

• Broadband capability covering 18 to 36 GHz,

• Hermetically sealable

• Adaptability to accept MMIC's of different sizes• Low cost manufacturability for commercial and military applications.

Taking the next step, NASA Lewis Research Center initiated a program by making use of the existing

18-36 GHz package design as the baseline from which manufacturing, mass production, reliability and

quality assurance issues are addressed and perfected. Princeton Microwave Technology, assembled an

experienced team in microwave technology, monolithic microwave integrated circuit technology, and

ceramics processing technology for the successful transition of the existing package to a commercial

product which was manufactured by one of the leading microwave package manufacturing companies

in the United States.

The MMIC packages developed under this packaging technology were designed to operate over the

frequency range of 18 to 44 GHz. The package design was based on 92% alumina ceramic material

using a high temperature co-fired system. Extensive efforts were used to replicate the dielectric

constant and dielectric loss properties of the original design material. To this end a dielectrometer was

used to verify the properties of materials obtained from major suppliers of HTCC material. It was

quickly determined that the availability of the ceramic material with the desired properties was not

available and that the data supplied by the suppliers were inconsistent and did not take into account the

frequency dispersion of the dielectric constant. The characterization eventually led to the selection of

a tape at Microcircuit Packaging of America Inc., and manufacturing of the package followed soon

thereafter.

The performance of the packages were measured in an Intercontinental Microwave test fixture up to

40 GHz. The package demonstrated repeatable performance from DC to 40 GHz, with an insertion

loss of 1 dB per transition and a return loss of 15 dB at 30 GHz.

The packaging effort performed under this contract was intended to determine the manufacturing

issues associated with the packages. Demonstration of the package performance using single stage

MMIC amplifiers was accomplished from DC to 30 GHz.

'! .

NASA MMIC PACKAGE

•, 'i, _ •

" 'i, _ '

H i

2.0

INTRODUCTION

The objective of the MMIC packaging technology program is to develop a generic packaging scheme

that permits low cost and timely insertion of MMIC devices into NASA communications systems.

The development and design of the original package was well documented. A complete design was

licensed from Hughes Aircraft. PmT visited Hughes Aircraft and talked to the original designers of the

package in order to determine the critical parameters. Their input was to obtain the substrate with a

dielectric constant and dielectric loss consistent with the original design. The objective for the

manufacturing of the package was to replicate the material characteristics. The ceramic substrates to

be used for this program is one of the key elements for the MMIC package. The package designed

under the original R & D program used high temperature co-fired ceramic substrates with a dielectric

constant of 9.5 provided by Kyocera. PInT contacted many US ceramic substrate manufacturers

including MPA, Mistier, Coors, etc. None of the companies had ceramic substrates for co-firing

applications with a dielectric constant of 9.5. For example, Coors Ceramic Company, for their

production packages, has tape cast substrates with a dielectric constant of 9.1 measured at 1 MHz.

The frequency dispersion of dielectric constant was not specified or known.

The limited availability of a substrate with a dielectric constant of 9.5 forced PInT to look at the

redesign of the package feed through transition to adapt to a dielectric constant of 9.1. The transition

was analyzed using a simple model in Touchstone and confirmed [1], by 3D electromagnetic simulator.

The simulations indicated that a dielectric constant of 9.1 for the ceramic substrate will provide a good

electrical performance which was very close to the original design. The change in the dimensions of

the transition due to the change of the dielectric constant was within the manufacturing tolerance and

therefore required no modifications. The only other issue left was to determine if the manufacturer

supplied dielectric constant was accurate fbr our application when frequency dispersion is taken into

account.

To determine the accuracy of the value of dielectric constant and its dispersion with frequency we

contacted GDK Products Inc., to conduct the appropriate measurements. The GDK measurements

Princeton Microwave Technology Inc. 2

/

ii

NASA MMIC PACKAGE

system uses a non-destructive method of determining the dielectric constant and loss factor. All

samples received from the substrate manufacturers were subject to measurements and it was

determined that substrate specified at 1 MHz with a dielectric constant of 9.1 exhibited a dielectric

constant of 8.3 GHz at 15 GHz. The frequency of measurement is determined by the thickness of the

sample. This was consistent with the majority of the substrate vendors. For example Mistler procured

the raw material from Coors and used the ! MHz values for his formulation. The measurements of

Mistler substrate exhibited similar characteristics to that of Coors. The measured properties of the

various substrates are documented in section 2.1

To speed up the overall measurement procedure Princeton Microwave utilized a GDK products

dielectrometer. This enabled PInT to test a large array of ceramic substrates at a faster rate and

provide appropriate feedback to MPA.

Following the selection of the green tape material, MPA was asked to proceed with the manufacture

of the package with their standard high temperature co-fired ceramic process. The selected substrate

exhibited a dielectric constant of 9.063 at a frequency of 11.484 GHz with a loss tangent of 0.00189.

The only other change to the original package design was a change in the layout of the bias ports due

to a discrepancy in the earlier design. The modifications are detailed in section 2.2.

2.1 Substrate Material Measurements

A number of substrates were received from Microcircuit Packaging of America, MPA. All of the

substrates were processed internally by MPA. The substrates were sent to GDK Products in

Cazenovia, NY, by PInT for characterization. Following is the summary of the measurements on the

first batch of substrates.

:::::::: :'.'.'.:::-": ::: ::':::::::::" :::::":: ": :: ::: ::: :::: ::::: ":!:_:_:i:i:i""'::}_:i:!:?"!:i:i:i:!:i::'i:i:i:i¢i:i:i:!:_::"===================================================

::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::?::?:!:: :::::::::::::::::::::::::::::::::::::::: iiiiiii_!_!!iii!iiii_i_i!_!i!i!i!!i!iiiiiiii!_!iii!_!_*!*!*!*i:!*!*i*i*i*i:i*i:i:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::

1 8.664 .0015 13.0736

2

3

3*

3**

* Tight Coupling

8.657

8.775

8.830

8.889

** Loose coupling

.0022

.0031

.0026

.0028

13.0376

13.0417

13.0213

13.0194

during measurement in test cavity [2].

Princeton Microwave Technology Inc.

NASAMMICPACKAGE

Due to the low dielectric constant,MPA sentmoresubstratesfor characterization.The following

propertiesweremeasured.

ii!iiii!iiiiiiiiiiiiiiiiiii!i!iiiiiii! !!i!!i!i_i_i_ii!1(White)

2 (White)

3 (White)

1(Black)

2 (Black)

3 (Black)

8.690

7.875

7.450

8.928

9.063

8.932

iiiiiiiiiiii ! i!iiiiiiiiiiiiiiiiiiiii!!iiii_i!i_i_iiiiiiii!!!i!i!i!ii_i_!!i!i!_!_!_!i!!_!!_i_!_!i_!_ii!i!_!i!_!i!_!_ii!i

.0O69

.0061

.0034

.0022

.0019

.0018

::::::::::iiiil::::-i_iii::iii::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::

13.0968

13.5457

14.0669

11.5141

11.4840

11.5510

None of the substrates received from MPA achieved the desired dielectric constant of 9.5.

Subsequently a substrate from MPA with a dielectric constant of 9.063 (#2, Black) was selected.

Other vendors were not able to provide a substrates with the desired dielectric constant.

In order for our manufacturing to be successful and to meet RF performance requirement with the

new substrate a 3D electromagnetic simulation of the package was conducted [1], with a lower

dielectric constant. The results of this analysis is detailed in Figure 1. The change of the dielectric

constant did not cause a significant of the RF performance. Since the loss tangent was close to the

design value, it was decided to place an order with MPA for the package manufacture with the

.elected tap e.

'''1'''1'''|'''1'''1'''|'''|'''t'''1'''1'' '

0 _C_t_ "_ °'_'_'2"O_'_"_ _g_'_'6"&c_t3_'_

-10 _X_ " _, o%

000000 0 0

-20 o ,_5_

-30

......... 15211:Er= 9.5

-40 e ISll I:Et=9.1

t_ IS211:Er= 9.1

-50 ''['''l'''l'''l'''|'''l'''l'''l'''l'''l''.

18 20 22 24 26 28 30 32 34 36 38 40

Frequency [GHz]

Figure 1: 3D Analysis of the Ceramic Package

O

:Z


NASA MIVIIC PACKAGE

2.2 Package Fan-out Design

The mechanical layout of the original package was conducted by MPA before manufacture. A flaw

was discovered in the fan-out pattern associated with the bias lines at the corner. The flaw, detailed in

Figure 2, was due to the slant lines having an angle that interfered with the stress relieving circles at

the corners of the cavity. To alleviate this problem without affecting the RF performance of the

package, the following modifications were agreed upon with MPA. Basically, the dimension 0.015 has

been changed to 0.040 as shown in Figure 3. This change will not affect the location of the visa which

are critical to the tLF performance of the package.

There were some inconsistencies also with the dimensions of the transition as shown in Figure 4. The

first inconsistency was the length of the transmission lines at the inside edgeofthe substrate next to

the cavity. If we subtract 0.064 from 0.076, the length should be 0.012 instead of 0.016. The second

inconsistency was the length of the transmission lines at the outer edges of the substrate. This

dimension, shown in Figure 4, was 0.026 whereas in Figure 3 it was 0.030. It was ascertained that

0.026 was the correct dimension and therefore it was modified as shown in Figure 3.

am#

--0.of_!

lltoI'

_a,e

..i_ --------i

'i"

I tl:l

Figure 2: Package Corner Detail


NASA MMIC PACKAGE

L

W

--O01E] E] PLE

[_--Od:_b 30 PL

11[•-_------- b _ e 8 _

!!,i-

.rag e [ON[I_ F_LE:F

O01B

-_ .-_ I--o.ol_

Figure 3: ]Package Corner Detail Showing Modified Corner

/.

•: .r

0.2EIO

////

fl.12B

///

//

©

----(1.040

I

I

I 'IO.Q76 ,

I©

Figure 4: Transition Design

L


NASA MMIC PACKAGE

-i( _ _

<

), :,:, , •

i:

2.3 Fabrication

The existing baseline package designed by Hughes Aircraft Company (Hughes package) takes

advantage of the maturity of the co-firing process technology for hermetically sealable MMIC

packages. However, the co-fired ceramic process does have limitations on resolution associated with

printing process and co-firing process. With the Hughes package, such problems have already been

addressed and resolved satisfactorily.

The W/Ni/Au Metallization on system used for the Hughes package is generally accepted as a good

choice for this application. PInT instructed MPA to use the same Metallization or the conductors, the

ground planes and to fill the RF vias. The High Temperature Co-fired Ceramic process flow used by

MPA is detailed below. The process is very similar to that used by Kyocera for first package. Figure 5

is a typical process flow diagram for ceramic package fabrication. Understanding the co-fire ceramic

process enables the designer to appreciate the dynamics and complexities inherent in the technology.

The details below provides an Overview in terms of process which will be expanded to specific design

parameters.

Ceramic Slurry

The Co-fire multilayer process begins by milling precise amounts of raw materials into a homogenous

slurry. This mixture, which has the consistency of latex paint, is principally Alumina (aluminum oxide,

Al203), and fluxes with small amount of organic binders and solvents.

Ca iing

The shtrry is poured onto a mylar sheet and then passed under a doctor blade to produce a uniform

strip of specified thickness. When dried this strip becomes a ceramic filled plastic tape with the look

and feel of thick vinyl.

y

Cutting

The sheets of"green" (unfired) ceramic tapes are cut into cards. Alignment holes are punched in each

card to ensure accurate layer to layer alignment dtuiug subsequent operations. Exact registration

becomes more critical as circuits densities increase.


NASAMMICPACKAGE

i:i

i

• i

iii__ :

1Iiiiiiiiiiii!iiiiiiilii_!i::i_::i::!::i_ii::iii!ii!ii;iiiiiiiiliiiiiiiliiiiiii

tliiiii!!i

!•z. _ ]i .....

l

Figure 5: Flow Process for Ceramic Package Manufacture


NASA MMIC PACKAGE

Punching and Filling

Via holes, cavities and notches are filled in the cards as required. The vias are filled with tungsten

paste, a mixture of metal powder and organics that become conductive when fired. Although other

metals, such as copper, are better electrical conductors, their melting points are lower than +1600 °C

necessary to co-fire alumina ceramics. Tungsten combines a satisfactory level of electrical conductivity

with a sintering point and shrink rate compatible with alumina.

The standard filled via will carry the electrical signal from one metaUized layer to the next. When

densities dictate via diameters of 0.010" or less, process precision is essential. The bore-coated via, a

large open via with a metallized ID coating, can fimction either as a castellation or as a receptacle for

interconnecting wires or pins. These vias are punched in the same manner as standard vias but coated

by a proprietary process.

The diameter of via holes in this package will be 0.008_+0.001 although MPA has a capability Of 0.005

inch.

Screening

Conductive circuits are printed onto the ceramics tape by forcing tungsten paste through an open mesh

metal screen. The circuit pattern for each artwork layer is made into a screen by the same

photolithographic process used in thick film and other screen printing techniques. The artwork, is

adjusted by interamics to allow shrinkage during firing. MPA uses a 325 mesh/inch screen. The

straight line width is with in 0.125 rail after thick inks are filled. The ink is 0.005 inch away from any

edge of the substrate. The tolerance on printing linear lines is _+0.005 nch.

Assembly and Lamination

Screened layers are inspected, aligned and laminated under high pressure and low temperature into one

assembly. This assembly may be comprised of a single unit, or include multiple repetitions of the same

product. The multiple units are processed as a single assembly from punching through lamination.


NASAMh,flC PACKAGE

i _ _ /7

i. '¸ r •

, • (

Scoring and Side Screening

A high precision score tool is used to form single units or to separate multiples. This tool is also used

to partially score green ceramic structures to allow plating interconnects to be snapped off after firing.

When conductor pads are required on the side of the product, additional screening operations are

performed on the laminated, trimmed units.

Firing and Flattening

The ceramic structure is cofired at approximately 1600 °C in a carefully controlled reducing

atmosphere. During the firing process the product shrinks approximately 20 percent (linear) for a total

volume reduction of 40 percent. A subsequent high temperature operation may be used to achieve

.002 inches per inch flatness or better. MPA has a _+0.8% dimensional control using the HTCC.

Base Plating

Base finish material, nickel, was electrochemicallyl applied to all exposed tungsten Metallization to

improve performance, prepare the surface for subsequent braze operations and prevent oxidation of

conductors pads.

Brazing

Brazing is used to join metal components to the nickel plated ceramic structure. Carbon or ceramic

braze fixtures are utilized to properly position metal components in relation to the ceramic structure as

it is processed through a braze furnace. Copper-silver eutectic is the most widely used material for

brazing. It forms a strong hermetic joint. When thermal expansion mismatch between alumina ceramic

and the metal component (Kovar or Alloy 42) being significant, an indium-copper-silver eutectic was

used.

Final Plating

Electrolytic plating requires that all exposed circuits be connected electrically by means of a lead

frame, plating bus or a combination of the two. Gold was used as the final plating material.


NASAMMICPACKAGE

H ,i

2

Final Inspection

A product specification is used to define sample size and testing procedure and supplement inspection

parameters included in a product drawings. A certificate of compliance, test reports and first article

documents were provided.

All packages delivered are checked for hermeticity. The packages achieve typically 1 x 10 .8 cc/sec of

Helium.

Since the ceramic package has Kovar material for the package base, and the size is of the same order,

it will no change in thermal resistance significantly. However, the ceramic substrate offers a very close

match in thermal coefficient of expansion (TCE = 6.7 PPM per degree C). Therefore, the stress level

will be acceptable even for the larger size package. For high power dissipation (higher than 2 watts),

copper-tungsten material is perfectly matched with ceramic in term o£ thermal coefficient of expansion

and provides much lower thermal conductivity.

:i_ ,


:i(;_ ¸ _/,:

i _i _ f_'

,;i ¸

, • _ i_ _ " _/

.... . /

• . (', _: '_, i_ •

NASA MNffC PACKAGE

2.4 Mechanical Measurements

Two hundred packages were received from MPA. Incoming inspection of the packages were

performed. Each package has been given a serial number. Dimensional measurements were performed

on more than 10 % of the packages. The samples were randomly chosen. The following dimensions

were measured.

A - Package external width

B - Package external length

C - Package cavity length

D - Package cavity width

E - Cavity depth

F - Package height

Figure 6: Package Dimension Measurements

Princeton Microwave Technologic Inc.12

NASA MIvIIC PACKAGE

i_ :::i! :::ii: • _:

?

. _ii _ ,

i ¸

/

PAl( No. A

0006 0.285

0007 0.285

0008 0.285

0009 0.285

0011 0.284

0012 0.284

0013 0.283

0014 0.285

0015 0.285

0016 0.285

0017 0.286

0018 0.284

0019 0.285

0020 0.284

0031 0.286

0032 0.285

0041 0.285

0042 0.285

0051 0.285

0052 0.285

0061 0.285

0062 0.285

0071 0.285

0072 0.285

Max A 0.004

B C D E

0.284 0.168 0.107 0.032

0.285 0.169 0.107 0.031

0.285 0.162 0.106 0.030

0.285 0.169 0.110 0.031

0.285 0.165 0.110 0.032

0.285 0.169 0.108 0031

0.283 0.166 0.109 0.034

0.285 0.167 0.108 0.032

0.285 0.168 0.106 0.031

0.285 0.169 0.107 0.032

0.286 0.169 0.107 0.032

0.284 0.168 0.108 0.030

0.285 0.169 0.105 0.031

0.284 0.167 0.107 0.031

0.285 0.167 0.106 0.031

0.285 0.167 0.106 0.033

0.285 0.168 0.105 0.032

0.286 0..169 0.105 0.034

0.285 0.170 0.110 0.030

0.285 0.169 0.107 0.035

0.285 0.167 0.106 0.032

0.285 0.169 0.109 0.035

0.285 0.169 0.106 0.035

0.285 0.167 0.106 0.035

0.003 0.008 0.005 0.005

F

0.044

0.043

0.043

0.043

0.043

0.042

0.041

0.043

0.043

0.044

0.045

0.043

0.043

0.042

0.042

0.043

0.043

0.043

0.042

0.044

0.043

0.044

0.042

0.043

0.004

Based on the measurements, the mechanical dimensions for the package were within the full design

dimensions.


NASA MMIC PACKAGE

, • "7 •i?_H

r.

2.5 Electrical Measurements

An Intercontinental Microwave Test fixture was used to conduct RF measurements. Fifteen mil

alumina substrates with 50 Ohm lines were soldered in the packages. After full calibration of the

Network Analyzer with the Intercontinental test fixture, the insertion loss and return loss of the

packages were measured. The packages showed good performance characteristics. Basically, the

return loss was better than 15 dB from DC to 35 GHz, This means that the match provided by the

transitions is extremely Broadband. The insertion losses including both the input and output ports

averaged less than 1 dB from DC to 18 GHz, or less than 0.5 dB per transition. The insertion loss

increased above 32 GHz. The measurements were repeatable.

Princeton Microwave Technology has successfully demonstrated the manufacturability of the package,

which was the primary objective of this program. The measured performance of the package is detailed

in Figures 7 to Figures 12.

r

Figure 7 and 8 detail the insertion loss and return loss of package 001. The insertion loss includes the

loss associated with a 15 rail alumina, 0.110 inch long, bond wires and the two transitions. The

measurements were conducted in the Intercontinental Microwave test fixture.

Figures 9 and 10 detail the insertion loss and return loss of package 002.

Figures 11 and 12 detail the insertion loss and return loss of package 003.

.,_ L_¸ _, ,: .,_

i_ _,


NASA MMIC PACKAGE

:: ._ L- ¸ •

i

_! FORWABD_ISSI_|

•LOB F!AGNIllJOE ,_- o;ooo..,

J ) | -

_ 2.ooo__

. - .....

0.040000 BHz 40.000000

Figure 7: Forward Transmission of package 001

,f,."

Slt F0-_X_ IIEFLECTIOr(

LCG _ill_. F_'= 0,0_O00dB ............. _O'.L000 IB--/-P-i-_'Lr ........ 1........!...... l.... :! ! t l

, _ , o I ]( _ t ] I i ," i I ' ].......

L----_, -----_ ...... . .... I--.-'I" ! i J (I

j _ , ) _ t _ .... '-.... I..... -"

•, _ ) ) l "

_ , 1..... '..--........ .._---_--'t ....... '_.......... i_-- i i -

_ _, ! _ ! ! I i Ip:,.,,l

_--tl-/_-/-_---_L--q--L-/.--i...._......i......._.....-}• f " " _ (i/I/V _., -_.! }.-,

"-.... --!------r----_ _ , ! !

_, , : _, ] _] J1iz _11 1 L_.__.._.__:_._J .... ± .... -_.... --

o.o_oooo ., _._ ,40.000000

Figure 8: Forward Reflection of package 001

Princeton Microwave Technology Inc.15

NASA MMIC PACKAGE

i,, ,_,:,/,

:i _ ,

S2i FOR_=:_ TRAI_$HISSIO,II

LOG F.Af;,_ITL_F. RE,_- 0.000 fl;] ._.000 d3/DiV

.........r ........I....................i ..........I.........!........ _........ :---I ........]

, . . . ) " .:t i t i " . .; ..... - .......... _....._- ......................; _.......... i ....... ;---- -7 --_ i • "] -

;" ! I = _ t

:_-.._-

.... _-r-_- , _.. _., :_ ,.;_ -- -/x;X....... r- .'.

; 1.-', i _ i l! ! : i v it, ! !

; _ _ _ I _ ._ _ i I t ._

! ! ' ' i I _ _ .... __.___ ;:.:-.... -_...... i ........... -_---_ ........._............. - ,, --A_! ; _ ! , i ,, • i I_ll^t

i _ i i i _ _ ......_.V...'__0.040000 GHz _0,000000


S:li FOR_A_ FLEr'3..ECTIC_

l ..................i...........i........'.........._-'"..............I..........i..........I .......1! ! i ] i ! i !I

! I t ,-" t ; _ I" .,_ ; _ I ; _ ,

i i [ " _ I,t , , I I I .I .I i ' ! i , I i -I

I i i , , I '

-_ ; _ .; _ - _ ...... -t-,-- ....... | ........ r ....... "I_I i I I t , ! I"

l t t i

Y77-7:,,:i-i........]........i i /Ii---", - ,"'--i 7.... --r--......'--i .......' ! " •

.t.1.-/----!]I---Ii --tl....... r-:----'-i ...... 2.1 i . !,---/---71i_......... -r .... _"--_ ;

_ __ _ . _ " _ i i _ i ,_;/..... L....... I........ 7 ..L..:.._..__ ............. ..:,._i ..... 7--..I..... ,..... .,.,,-,

0.040000 , 6h_ 40.000000

Figure 10: Forward reflection of package 002


NASA MMIC PACKAGE

_21FORWARDTBANS_I5510N

2.000 dB/DIV

'i-7....L,_1i

_L_

!I

1

!

LOGNAGNITUDE REF=0.000 dB "

t I ,i i F-j• l . 1 1

• i! i., J ! iF-TI ! i I,i'

'•--- !_L l

, , i i ,, i l 1L__J ! k_. ! _ t '

0.040000 GHz 40.000000


LOG _AGXITUI)Z RT._=0.000dB !0.000dB/DIV

............7......."---7--_ .......{----_

, ! I• I

I ! j I ', ;

I _ i i I...................._ ..........."_...... _............................ .._....... _........ .-.........

I " I

..... -_ ......... _, ....... T ....... _

, , _IV'_...

• :..................V................i........! ' ] i ! i i ,

0.040000 61_ _0.000000

Figure 12: Forward reflection of package 003


NASA MMIC PACKAGE

2.6 MMIC Amplifier insertion into Package

Three different MMIC amplifiers were assembled into the MMIC package. The first NIMIC was a 2.4

dB noise figure, low noise amplifier made by Texas Instruments. The amplifier requires a single bias

for operation. The second amplifier was LMA422 fom Litton Industries, with a gain of 20 dB. Both

amplifiers are designed for the 28 GHz LMDS bands. The third amplifier measured was a HP

distributed amplifier which has a gain of 8 dB from 2 to 26.5 GHz.

All three MMICs were epoxy attached in the package and connected to the package input and output

using 0.8 nail gold wire.

Figure 13 shows the measured performance of the TI low noise amplifier.

Figure 14 shows the measured performance of the Litton MMIC amplifier.

Figure 15 shows the measured performance of the HP MMIC amplifier.

The measured performance of the amplifiers demonstrate the wideband performance of the MMIC

package.

The photograph below shows a MMIC mounted into a package.


6_ "ouI.,_'olotrtloo.L o^e...,c,oso._ uomou!..tcI

aob'TIdttrV [IBII, I[ oslou xto I IJ.JO oougm.tojaod poansgoIAI :1_I oan_hI

zH9 0002;0"8_

-zH9 O00EIO'Lg

_P _g_'_I

zH9 000_I0"9_

EP ,"0_"O_

ZV.S000II0'6_3V rc_h"dVd

oooo;o'o_ _ ooo_+o'-__.... '.---- _........ '.........-:..... - .... _...... '-...... .-T....... i"-....... q

.......:......i......+........+........._.........:......+-......!......., ........_,t........--.......... _..........;...........:...........i..........:...... ]..........;............... --]

................ _........ _......... ':.......

)i -"1 : V- ' I • .

: .... ........... _ .-....', ......... _ ............ ; ........... , ........" " : " + ; i i .

t-.........,+........+......{..........7 .......i.........._-- + i t _IL-7,._/_p'-+6U6+'...................... -_i_'-6_o'_ _£_ ' ._r,qu;,.'sv_,sol

aot.JlldulV DIIAII, kI ZZI_VIAFIuo_2,!'I jo o3ugm._oj_tod po.msgolA[ :£1_ o.tng_I

_p O6;'L;

zH9 O00£10"L_

BP _0_'6;zH9 000_i0"9_

_q9 000_;0"6_E_EfCA

I

i

II

I

O00_I0":;_

+ _4 + + _ I.__[ F i- t ! t

-',---LT-_+_.....L__t.._L+q;¢OISSI_hYSL OhygLllO-II_

_OV"ADVcI OIF_ VSVbt

0E •ouI,_o[out/oo+Lo^raxoIo.rrAuo_oou.t+M

',:, ¸.>,]5 :': L

DIIAIIAI d_zHo _'9_;-Z: jo oougm.mjJod po.ms_oIAI :_ o.m_!_I

000000'0_ _ ._ 09_o6";......... ..- .................................... I" ...........r....... _....................._ 1..... I r ! t 1 ! -]

, ° - : + - '. i - ,,L...... p.........._............._.......... t.......... T: ........T-- ..... -F _ t _.._..2,

........ ,: ........... tt .......... t ............ t ...... f-- t ' ;

,.o___et......... i ............ l _L_tN- _ .

O00000"Ot, z148 099966"T

! i\ i I x .....L ! # !...............

t ,. _ 't I ; _ l

| : _ t ; .--L ...... .,: ......... i ........... t ....... 7 ........... _-......... "_

, , - .......i ........ YT ., _ . , J I

L........ ;-+__-,+ ...... # ........ _T'-"'--- _+._-" ...... _ ii ; I i [ _ t ! ; f

I

L__.mt_,.'_p_....i....000,_g__,_....... L ............-iD-'_, _"ii_,_' ........... ...........k

GP 9ZL'GT-zi..EtOg6_Eg'B_

T t:EtY,_dVH

== 0000"0_NV_d _31'_- --r'_

TTS - E 140

,/

/

'i

HDV'ADVcI C)II_XI VSVN

NASA MMIC PACKAGE

2.7 t)ackage Layout

The drawing below details the package layout.

I

t

t

.

_o

_i_ ¸W_ _ Z

_ _Z

_ WMMM

_ v

!,!i-7"_i

.(

3

_e e

-o _ tl ::1

"1_io !13! +"_ 171

! !


NASAMMICPACKAGE

2.8 Package Data Sheet

Princeton Microwave Technology has begun the marketing of the MM]C package. The preliminary

data sheet is detailed below.

PRINCETON MICROWAVE TECI_NOLOGY, _NC.3Nami Lane, Unit C-10, Mercerville, NJ 08619

Tel: (609 586-8140)

FAX: (609) 586-1231

C [P,._flC WALL CC_R,_£O

LMDS PACKAGE _a_ w,_:c_Jg_T[

FEATURES: _c_PRCC_._E0

• Low Cost coN_croas

• DC-32.GHz Operation• Insertion Loss < 1.0 dB/Transition• Remm Loss > 15 dB

• 50-ohm Input/Output Ports• lODCI/OPorts

• Hermetically Sealable _p_c \\\\ _v'//_.• High Thermal Efficiency _wr_u _-_V/r// '\

_I_T._L• Cavity Size (125 x 70 mils)* a_

• Low RE Leakage/Radiation 0.044 o.o160.034• Gold Plated Copper Tungsten Base 4, 4 4,

• Cavity size can be customized -f--Ii," c.:_,._ ,I/

PMT-401125 Transmission Characteristics (Both I/O Transitions)

--7 j i7---i

0.040000

'.. _FEF_

OPTIONS:

• More Than Two RF Ports

• Locations of RF/dO Ports

• Multi-Cavity Package with High RF Isolations

• Multichip Module Configurations with High RF Isolations

• Thickness and the material of the cover can be specified


NASA MMIC PACKAGE

3.0

CONCLUSIONS

Princeton Microwave Technology has successfully demonstrated the transfer of technology for the

MMIC package developed under contract No. NAS3-25486. During this contract the package design

was licensed from Hughes Aircraft Company for manufacture within U.S. A low cost hermetically

sealable, high performance package has been manufactured with a typical insertion loss of less than 1

dB per transition up to 32 GHz and a return loss better than 15 dB. The low cost package can

encompass a large number of existing MMICs with a size of 100 mils by 50 mils. The performance of

the package has been demonstrated by assembling three different MMIC amplifiers into the package

and measuring their performance. Two of the MMIC amplifiers, designed for the Local Microwave

Distribution System, LMDS, at 26 GHz to 30 GHz exhibited measured data consistent with the device

data. The third MMIC amplifier was a broadband distributed amplifier for the DC to 26 GHz

bandwidth. The measured performance of the amplifiers showed good performance upto 27 GHz.

The measured performance of the package, when used with an alumina 50 Ohm line, does show

resonances within the band, as detailed in Figures 7 through 12. However in the broadband amplifier

measurements the resonances were not observed, and maybe due to the loading of the cavity with a

higher dielectric constant of CraAs.

Preliminary prices from the manufacturer for a higher quantity has been obtained. In quantities of

25,000 pieces the package price will be under $2.00. For a quantity of 1000 pieces the price of the

package will be under $5.00. In conclusion a low cost MMIC package has been manufactured. The

main effort in this program was focused on procuring the ceramic material with the right

characteristics. The continuing development and manufacturing of the MMIC package will involve this

type of interaction, between the package manufacturer and design center.

The MMIC package is presently being marketed in the US and will be made available to NASA and

other users.


NASA MMIC PACKAGE

• ::?

REFERENCES:

[1] Private communication between Mr. Steve Yuan and University of Michigan

[2] Non-Destructive Permittivity Measurement of Substratcs.

G. Kent. GDK Products .1995 Stanley Road, Cazenovia, NY 13035


_),Z_ 5/_ _

Z,: "ii _,

/ ,\

_i _''i I

H

, i"_ _ !

_i _i "_,

Form ApprovedREPORT DOCUMENTATION PAGE OMBNo.0704-0188

Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources,

gathering end maintaining the data needed, and completing and reviewing he collection of information. Send comments regarding this burden estimate or any other aspect of this

collection of information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson

Davis Highway, Suite 1204, Arlington, VA 22202-4302, end to the Office of Management end Budget, Paperwork Reduction Project (0704-0188), Washington, DC 20503.

1. AGENCY USE ONLY (Leave blank) 2. REPORT DATE

JULY 25, 1997

4. TITLE AND SUBTITLE

MMIC Package for Millimeter Wave Frequency

6. AUTHOR(S)

Sarjit Singh Bharj and Steve Yuan

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)

Princeton Microwave Technology, Inc.

3 Nami Lane

Unit (2-10

Mercerville, NJ 08619

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES)

National Aeronautics and Space Administration

Lewis Research Center

Cleveland, Ohio 44135-3191

3. REPORT TYPE AND DATES COVERED

Final Contractor Report5. FUNDING NUMBERS

WU-323-79-02

C-NAS3-27813

8. PERFORMING ORGANIZATION

REPORT NUMBER

E-10801

10. SPONSORING/MONITORING

AGENCY REPORT NUMBER

NASA CR-204131

11. SUPPLEMENTARYNOTES

Project Manager, Gerald J. Chomos, Communications Technology Division, NASA Lewis Research Center, organization

code 5640, (216) 433-3485.

12a. DISTRIBUTION/AVAILABILITY STATEMENT

Unclassified - Unlimited

Subject Category 33

This publication is available from the NASA Center for AeroSpace Information, (301) 621--0390.

12b. DISTRIBUTION CODE

13. ABSTRACT (Maximum 200 words)

Princeton Microwave Technology has successfully demonstrated the transfer of technology for the MMIC package

developed under contract No. NAS3-27813. During this contract the package design was licensed from Hughes Aircraft

Company for manufacture within the U.S. A major effort was directed towards characterization of the ceramic materialfor its dielectric constant and loss tangent properties. After selection of a ceramic tape, the high temperature co-fired

ceramic package was manufactured in the U.S. by Microcircuit Packaging of America, Inc. Microwave measurements of

the MMIC package were conducted by an intercontinental microwave test fixture. The package demonstrated a typicalinsertion loss of 0.5 dB per transition up to 32 Ghz and a return loss of better than 15 db. The performance of the package

has been demonstrated from 2 to 30 Ghz by assembling three different MMIC amplifiers. Two of the MMIC amplifiers

were designed for the 26 Ghz to 30 Ghz operation while the third MMIC was a distributed amplifier from 2 to 26.5 Ghz.

The measured gain of the amplifier is consistent with the device data. The package costs are substantially lower than

comparable packages available commercially. Typically the price difference is greater than a factor of three. The packagecost is well under $5.00 for a quantity of 10,000 pieces.

14. SUBJECT TERMS

MMIC package; Millimeter wave

17. SECURITY CLASSIFICATION 18. SECURITY CLASSIFICATION

OF REPORT OF THIS PAGE

Unclassified Unclassified

NSN 7540-01-280-5500

19. SECURITY CLASSIFICATION

OF ABSTRACT

Unclassified

15. NUMBER OF PAGES

25

16. PRICE CODE

AO3

20. LIMITATION OF ABSTRACT

Standard Form 298 (Rev. 2-89)

Prescribed by ANSI Std. Z39-18298-102

Documents

MMIC PACKAGE FOR MILLIMETER WAVE FREQUENCY · PDF filenasa-or-205037 final technical report mmic package for millimeter wave frequency contract no: nas 3-27813 prepared for nasa-lewis