An Earth's field magnometer that Utilizes the free precession of Protons

8/13/2019 An Earth's field magnometer that Utilizes the free precession of Protons

1/202

oAN EARTH'S FIELD MAGNETOMETER THAT

UllLIZES (HE FREE PRECESSION OF PROTONS

CHARLES H. BOWEN, IR.


2/202

LibraryU. S. Naval Postgiaduate SchoolMontore)', Caliiornia


3/202


4/202


5/202


6/202


7/202

AN EARTH'S FIELD MAGNETOMETERTHAT UTILIZES THE FREEPRECESSION OF PROTONS

Charles Ho BOIVEN, Jr.


8/202


9/202

AN EARTH'S FI^LD KAGNICTOMF.TSRTHAT UTILIZES THE FREEPRECESSION OF PROTONS

byCharles H. Bowen, Jr.

Lieutenant, United States Navy

Submitted in partial fulfillmentof the requirensntsfor the degree ofMASTER OF SCIENCE

INENGIN^.^RING EIECTHCNIC3

United States Naval Postp;raduate Schoolfionterey, California

1954


10/202

^>y


11/202


12/202


13/202

PH5FAC5This fHper describes the instrumentation of a new type of device for

measuring the absolute values of, as well as mall changes in, the earth'smagnetic fieldo

The author's work on this device was accomplished at the Varian Assoc-iates research Laboratory at Palo Alto, California, duxrlng the periodJanuary to March, 1954, while a student in the Electronics 3ngineeringcurriculum at the UoSo Naval Postgraduate School, Monterey, California.

The idea that the technique described in this paper would be a feas-ible method for measuring the earth's magnetic field was conceived byDro Russell Varian and a basic patent on use of this technique was grantedto hira in 1948,

The author wish-^s to express his anoreciation to Varian associates,to Messrs, Dolan Mansir and John Drake for their cooperation and especiallyto Dr. Martin Packard, the Director of Nuclear Spectroscopy at the labor-

atory, for his help and enttourageraent o

li


14/202


15/202

TABLE OF CONTENTS

PrefaceTable of ContentsList of IllustraticxisChapter 1Chapter II

Chapter III

Chapter IV

Chapter VChapter VI

IntroductionOther types of Magnetometers

1. Dip Needles2. Earth Inductor3. Schmidt-Type Magnetometer4. Magnetic Airborne Detector

a. Gulf Airborne Detectorb, NOL Magnetic Airborne Detector

5o Station MagnetometerMethod of Attack

1, Theoretical Background2, A New Type Magnetometer

Experimental Equipment1. Polarizing Coils2. Receiving Coil3. Water Sample4. Preamplifier5o Amplifier, Narrow Band6, Gating and Counting Circuits7o Analoging Circuit

Experimental ResultsConclusions

1. Possible Errors2o General Comments and Applications

PageiiiiiV1

222

333

516

2629313132374244

4646


16/202


17/202

TABLE OF CONTENTSPage

Appendix I 48Appendix II 57Appendix III 58Appendix IV 59Appendix V 86Bibliography 8?

lY


18/202


19/202

LIST OF ILLUSTRATIONSFif^ure Pagelo Precession of nuclear magnetic moment about 8

the polarizing field2o Position of polarizing and RoF. coils 83. Ho, Hi ind M vectors 94o Components of vector M 115o Proton resonance fcr ol molar solution of /^ r\ S04 12

in water6o 30 MoC. R.F. Head 14

7o Nuclear spectroscopy tape for ethyl alcohol 158o Direction of vector M after polarization 179. Orientation of receiving coil 19

lOo Block diagram of earth's magnetic field measuring 19equipment11. Input to first stage of preamplifier 2112 Exponential rise of sine wave signal impressed on 22

a high Q circuit13 o Waveform produced by signal/fC .4*^ t*^ impressed on a 23circuit with a Q of 20014. Timing wheel 2?.15. Polarizing coil circuit 28l6o Transient response of polarizing coil 2917. Preamplifier 3218 o Q multiplier circuit 3319 o Photographs showing result cf impressing a sine wave 35

on a hi^ Q circuit 3620o Narrow band amplifier 36


20/202


21/202

LIST OF ILLUSTRATIONSFigure Page21, One stage of binary circuit 3822. Binary gating counter 3923 o Timing wheel AG24 o Diurnal change in earth's magnetic field 45

vi


22/202


23/202

CHAPTER IINTRODUCTION

The measurerBrib of the earth's nagnetic field has Icaig bean of inter-est to various magnetic observatories throughout the world, to thoseinterested in iragnetic prospecting, arvl to the military in anti-submarinevork. This paper describes the instrumentation of a new type jnagneto-meter that measures the earth's field by measuring the free precessionfrequency of the protons in water

Results obtained indicate that with the present equipment configur-ati on the earth's magnetic field can be measured to within 2 y (out of50000 q) on an absolute basis and that changes of }; J can be detected.This technique probably represents the most accurate method presentlyavailable for measuring the absolute value of the earth's magnetic field.Further, changes of much less than h ^ could be detected by this methodif it were necessary or desirable siraply by increasing the frequency ofthe crystal controlled source in the counting system.

Possible applications include: use as a station magnetometer for amagnetic observatory, use for magnetic surveying or prospecting, arid usein harbor defense and air anti-submarine worko

^ 1 V = 10 ^ gausa


24/202


25/202

CHAPTER II0TH3R TYPES OF MAGRSTOMSTERS

1. m^ NaedleaThis is simply a compass needle free to move in a vertical plane

with an adjustable weight attached on one side of the pivot o A balanceis achieved between gravitational and magnetic torques. Changes in thevertical ccrnponent of the earth's magnetic field changes the magneticmoment and results in an unbalance in the forces.

This device measures changes in the earth's magnetic field and will

detect anomalies of approximately 300 Jo The earth's magnetic field,as mentioned in the introducticn, is about 50000 Y^2. Earth Inductor

This instrument is frequently used by magnetic observatories formeasuring the inclination of the earth's nagnetic field. It consistsof a high speed rotating coil which generates a voltage for all orien-tations exceot those for which its axis is parallel to that of the mag-netic Held being measured. It can also be used as a crude device formeasuring the magnitude of a magnetic field since the voltage generatedby the rotating coil is proportional to strength of the magnetic fieldwhose flux is being cut. When used to measure the absolute value ofthe earth's magnetic field it is accurate to within about lOCX) Y .3. 3chmidt-Type Magnetometer

This device is widely used in magnetic prospecting. It is similarto the dip needle in that there are ooposing gravitational and magnetictorques. It is much more sensitive tlmn the dip needle, however. Itconsists of a magnet that is pivotedbut not pivoted at its center o^mass, Hy careful adjustment it can be made to detect changes in the


26/202


27/202

earth's fiald of a few gammaeven one ganuna under favorable conditions,U. Magnetic Airborne Detector

ao Gulf Airborne Detector (9)This device is also known as the flux-gate magnetometer or satur-

able reactor. It utilizes a ferro magnetic element cf high permeabilitysuch that the earth's field can very nearly saturate it. If an alternat-ing field, obtained from a coil, is superimposed on the earth's field thecore will saturate each cycle. The phase cf each cycle at which satur-ation occurs gives a measure of the earth's field o Small changes of afraction of 1 ^ can be detected in this way. The instrument does notprovide an absolute measure,

b, NOL Magnetic Airborne Detector (6)The magnetic airborne detector developed by NOL (Naval Ordnance

Laboratory) and Bell Laboratories operates on a slightly different princ-iple. Here a single core is driven into saturation about 2000 times asecond. A back electromotive force is set up in the magnetizing coilT^rtiich has even harmonics when there is an external field in the directionof the axis but only odd harmonics when there is no such fieM , Theamplitude of the even harmonic content as recorded on a tape is propor-tional to th field strengtho This instrumsnt neasures changes of afraction of one ^ but does not measure the absolute value of the earth'sfield.5 Station iMagnetometer

5ven a brief description of a station magnetometer would be toolengthy for a paper of this sorto McCorab, (5) in a U.So Department ofCommerce Publication, describes in detail how a magnetic observatory


28/202


29/202

should be set uo and equipped. The Coast and Geodetic Survey maintainstwo magnetic observatories in this countryone at Tucson, Arizona andth^ other at Cheltenham, Massachusetts. A station typically has a triplewallad building with sawdust between one pair of walls and air betweenthe other pair to reduce temperature and humidity changes in sensitiveequipment. Standard magnets are used and complex optical equipaaent isnecessary. This is mentioned to illustrate ths difficulty previouslyencountered in making earth's field measurerre nts. Accuracy of betterthan 5 \ for an absolute earth's field measurement was very difficultto obtain.


30/202


31/202

CHAPTER IIIMETHOD OF ATTACK

1. Theoretical Back/;roundAn entirely new approach to the nagnetcmeter problepi was opened

by the field of nuclear induction or nuclear magnetic resonance. Bloch,Packard and Hansen (2) at Stanford and Purcell, Torey and Pound (8) atHarvairl pioneered in this work. A brief amount of background materialfollows. The notation used follows that of Block (2) and Packard (7).

The phenomenon of nuclear induction is possible because of twoinherent properties of the nuclei: their angular momentum and magneticmoment . The unit for magnetic moments is the Bohr magneton

where e is the charge on an electron, m represents its mass, c is thevelocity of light and ^equals lo05 x 10 ^^ erg-sec. {zITv^b Planck'sconstant )o The nuclear magneton is related to the Bohr magneton simplyby the ratio of the mass of the proton to that caf the electron. For thenuclear magneton

However, careful measurements have shown that fcr the proton, since itis not simply a spinning spherical shell of mass m and charge e, thatits magnetic moment is 2o7935 times the theoretical value. The measuredratio of magnetic mcaaent of proton to that of the electron is 1/660.

The second characteristic property is the angular momentum, a ^ ,and is given in units of j^ o

njJii


32/202


33/202

The angular momentum and magnetic moments are ralat-^d by the gyro-magnetic ratio gamma, \ , (This | is unrelated to ths unit of magneticfield )o

/^ -r5?The vectcrs /(Aand (2^re either parallel, in which case jis positive, oranti-parallel and ^ is then negative

We wish to have a relationship between magnetic field strength andfrequency. This can be obtained by considering the energy difference Ebetween the Zeeman levels for a material placed in magnetic field ofstrength Ho ^ ^ ^ ^ r: uO^X

The quantity is the parentheses in the gyrcxnagnetic ratio and wehave jOquals Y^* ^^^ frequency is termed the Larmor frequency andfor protOTis,

For example, if H equals 7000 gauss, the corresponding Larmor frequencyis approximately 30 mc. As indicated on the previous page the ratio ofthe nagnetic moment of the proton to that of electron is Z . The fre-quency of precession for the electron in a magnetic field is less bythis same factoro


34/202


35/202

Nuclear moments can be detected because they contribute to thetotal susceptibility o (A paramagnetic substance is one vd.th a positivesusceptibility while a diamagnetic material has a negative susceptibil-ity.) The average value of the magnetic moment for each nucleus is:^ ' Hi (^) ^ ^-Total magnetization per unit volume , '^/j'^ isi^tiraes number of nuclei perunit volume. Further/*^ equals%/Tmere TCis the susceptibility, for the

oproton at roan temperature: / -* ^ / / ^ x, X /yd =r 5 */ >^ /So that, if a magnetic field is Applied to a paramagnetic substance, there

is a total magnetization induced along the axis of the applied field.However this value is not reached instantly after applying a magnetic fieldbut approaches this value exponentially depending on a time constant 7 (For electron resonance the final value is approached in microseconds o)Ti for protons may vary from seconds to hourso During this time betweenapplication of a magnetic field and final alignment of the nuclear momenlja precession occurs about the external field at the Larraor frequency. Theprecessing nuclear magnetic moment vector, AJ, is shown in figure lo

Nuclear indue tioi was obtained at Stanford by the crossed coilsmethod. If Ho is approximately 7000 gauss then the frequency of precess-ion is approximately 30 mc as mentioned earlier o Ho will be termed thepolarizing field. The change to be observed is a voltage produced by thechcinge in orientation of the nucl-aar manentSo If ths field Ho was createdby two polarizing soils with their common axis the z axis (or by a magnetsimilarly oriented) and if to this is added two coils placed with their


36/202


37/202

axis alon-^ the x axis, then we would have the setup shown in figure 2,

i

Figure 1, Precession of nuclear magnetic mcment about polarizing field,

Fitnire 2a Position of polarizing and RoFo coils


38/202


39/202

If ths coils oriented along the >^ axis are excited at a frequencynear the Larmor frequency of 30 mc, it is possible for the nuclear momant,which has been processing about the polarizing field Ho, to absorb energyfrom the source of R.F. voltage. (The R.Fo field is termed the H^^ fieldand is small as compared with Ho) It is convenient to think of the RoFas composed of two vectors rotating in opposite directions. Thus, aslong as no sample was present to be polarized, the R.Fo vectors have ccnt-ponents alon^^ the x axis but c exponents along the y axis are cancelled tozero. However, the presence of a polarized sample provides a vector whichadds to one or the other of the two rotating RoF. vectors depending on thedirection of rotation of A\ . This vector, /ij , is initially processingabout the z axis and the angle , indicated in figure 3 is rather small.

Figure 3


40/202


41/202

Howevsr, as the vector attempts to precesa about the field that is theresultant of Ho and H^, M increases so that /^ , at resonance, is rotatingin the xy plane and is in phase with one of the two RcFo vectors previous-ly discussed. The R.Fo vectors are then no longer equal and so no longercancel to zero along the y axis.

fiil = fWhere M equals resultant nuclear magnetic moment per unit volume,

A - JL-^ ' J ^^ ->

The above equation describes the time variation to be expected fromthe polarizatiOTi vector /^ o It can be seen that the nsximum value, ioeo,resonance, will occur when /f is in the xy phase if Ho is along the z axis,For resonance u) equals T'Ho. If we assume that uJo ^ ^^ ^^ Uid

H;;


42/202


43/202

^ ccfi, i

r s:a{ Figure 4. Components of vector Mo

So that inrtien ^ equals ^ Ho equalstO^sW have resomnce, (f> equals 90and tgt equals^. The ill is used since Y'*'y ^ either positive ornegative in sign.

It is often easier to obtain resonance by havir^ a constant RoFofrequency and start with a value of polarizing field that is near thatrequired for resonance. Then by modulating the amplitude of the magneticfield along the z axis with sweep coilu at an audio rate, the nuclearmonent / can be made to pass thru the xy plane and, hence, resonance, atan audio rate^

If, in addition to these coils already described, there is addedstill another whose axis is along the J^ axis, termed the receiving coil,a small signal on the order of microvolts will be induced in the receiv-er coil by M as the vector/^ passes thru resonancey


44/202


45/202


46/202


47/202

One of the difficulties in realizing great accuracy vdth the magnet-omet-^r operating in the 30 megacycle region is the difficulty of reading-d thin a line width. The signal shown in figure 5 has a waveform givenby Bloch as: f \4 T -^

Where T, and T^ are relaxation tine constants and other factors havebeen previously given. The exact shape is difficult to determine, Furthei;the narrowest line (an electron resonance line) is approximately 20 milli-gauss or 2000 Y^o Therefore, for extremely accurate readings of fieldstrength, ths exact position of the resonant peak must be determined to1 part in 2000 if accuracy to within 1 gamma is desired. The problem ofreading within a line is quite complicated because of inhomogeneities thatmay exist in the field and because of the possibility of assymetry that mayexist in the wave shape due to a mixture of u and v modes as describedby Bloch. (1)

A common and very useful application of nuclear induction is in thefield of nuclear spectroscopy,, This is a tool used primarily by chemists.For example, the various proton lines of some substance such as ethyl al-cohol can be resolved into fine line structure as shown in figure 7


48/202


49/202

Figure 6. 30 MoC. RoF, Head


50/202


51/202

H -H ri1 1 c - c _- oH1 \

H ^

A

W a^^ci^^ /* 6* nf* //*' f^ i^^ S,


52/202


53/202

2. A New Type MagnetometerIn nuclear spectroscopy vork the relati onship 60 equals Ij H is made

use of in determining fine line structure. This same relationship canbe used in magnet caneter applications. For a given frequency the magneticfield H required to produce resonance in the protons of water is wellknovffio Thus we have the inherent possibility of being able to measureileld strength n by measuring frequency 60. This fact is made use ofin magnetometers working in high fields and at R.F. frequencies with littledifficulty since information to within d gauss is usually sufficientlyaccurate. A magnetometer that will operate in small fields with compareable accuracy is more difficult to come by. One approach to the problemwas to make use of electron resonance. For the fixed frequency of 30 Mc,proton resonance occurs at about 7000 gauss vrtiile electron resonanceoccurs in fields of only a few gauss for the same frequency. This possi-bility, the use of electrcai resonance for low field measurement, was in-vestigated by Levinthal and Rodgers, (4) They found that accuracy of t-.02 gauss or t, 2000 Y could be obtained.

The material that follows will describe a new technique that utilizesthe free precession of protons in water to measure tha small magnetic fieldof the earth. In this method a sample of water is polarized with a large

magnetic field (large as ccmpared with that of the earth). This polar-izing field is oriented approximately 90 degrees to the earth's nagneticfield. It is necessary to polarize the sample of water with a field Hoin order to orient the vector M approximately 90 degrees to the earth'sfield. The reason for this will become more apparent later.


54/202


55/202

TZTTK^i TTTU %

Figure 80 Direction of vector M after polarization.

The polarizing fLeld Ho is left on for a time long as compared withT^, where a time equal to 5 T^^ is required fcr alignment of the inducedmoment M approximately along HOo After 5 T, M would actually be alignedalOTig a line that is a resultant of Ho and the earth's field but if Hois much greater than the earth's field then the vector M would be po-sitioned approximately along Ho,

If this large polarizing field is now suddenly cut off, the vectorrepresenting the overall magnetic moment is left free to precess, withexponentially decreasing amplitude, about the only magnetic field remain-ing, that of the earth.

K will precess about the earth's field in the same way that it pre-

cessed about Ho immediately after Ho was impressed. Hovaver, the frequencyof precession is quite different in the tvo cases because Ho is much great-er than the earth's field and the frequency of pi*eces3ion is definitelyrelated to field strength as shown by the equation for (J equals | H onPage 7 which leads to a precession frequency of:


56/202


57/202

Thus, for the 30 Mc probe discussed earlier, the field associated withit was about 7000 gauss. New, however, with the earth's field of o5gauss, the frequency of precession is seen to be an audio frequency justover 2 Kc

If a receiving coil were placed 90 degrees to both the earth's fieldand the polarizing field, a snail signal of a few microvolts would be in-duced in it by the precession of the vector'Ul'about the earth's fieldafter Ho was cut off.

An important factor in practical application of this equipment shouldbe noted here. It is not necessary that Ho be exactly 90 degrees to theearth's field in order to determine the frequency of precession and thusthe earth's field strength. If . Ho is not 90 degrees to the earth's field,then the amplitude of the signal induced in the receiving coil is reducedbut its frequency is unchanged. Thus, if Ho is 45 degrees to the earth'sfisld, the ultimate signal is reduced to .7 of its previous amplitude butits frequency remains the same.

If the signal induced in the receiving coil after the nBgnstizingfield is turned off is amplified, and the exact frequency determined, itwould be possible to thus obtain a measure of the earth's magnetic field.A block diagram of the required equipment is sketched in figure 10,

It has been shown by Bloch (1) that the signal to noise ratio isproportional to the voluma of the sample. This is true because the great-er volume provides a larger number of dipoles. Therefore a fairly largesample should be used.

In nuclear spectroscopy the line resolution possible is limited byinhomogeneities in the big magnetic field, HOo For example, if the field

18


58/202


59/202

Y

1^

Figure 9. Orientation of receiving coil.

^JJo^:^ r

my^^

P/'C U*s\jp

^^f^'^h'cr

f^r c^ t/r H c y

1 rf/

Figure 10 Block diagram of Earth's magnetic field measuring equipment.

19


60/202


61/202

across the sample is homogeneous to within ,0C1 gauss, then spectrumlines separated by less than .Oul gauss cannot, in general, be resolved.Since it is difficult to create a magnetic field of 7000 gauss that ishomogeneous over a large area, the saraoles used in soectroscopy arenecessarily small, being on the order of one cubic centineter^ However,the earth's magnetic field is homogeneous over a very large area and thismakes it possible to use quite large samples and thus obtain better signalto noise ratios.

The polarizing coil should be capable of providing a magnetic field

of at l-aast 100 gauss in order to produce a magnetic field that is 200times that of the earth's magnetic field.

The open circuit voltage, vrtiich can be considered as being introducedin series with the coil has been shown by Packard (7) to be:Vooen c ircuit = ^fp'A^^^i^'C^SU^t A/O ^ i/c/^^N = number of receiver coil turnsA = area of cylindrical sample in cm~3M = 7^ ^

= susceptibilityjtM = 3.U x 10 '^^ //Ho = 100 gauss polarizing fieldN = 6o9 X 1022 ^^'^A^ = 1./* X 10~23 erg/gauss = magnetic moment for each nucleusK ~ 1.37 X 10 ^^ erg/degree = Boltman's constantT = 291 = absolute temperature in degrees Kelvin for room temperature

The voltage across the capacitor shown in figure 11 which is the vol-tage appli-^d to the grid of the first stage of the preamplifier is Q timesVoc. The rms value of the grid voltage is ^ -l^- where Q i th Q of the

20


62/202


63/202

receiv


64/202


65/202

The signal was centered about a frequency of 2182 cycles. A shift of1 cycle would correspond to a shift in the earth's field of approximately25 Q o Diurnal fluctuations based on data from the Coast and GeodeticSurvey's station at Tucson, Arizona, are about 40 j , corresponding toless than two cycles. Large magnetized material near the equipment couldcause somewhat greater fluctuations than this but is is apparent that avery narrow band amplifier is feasible. Noise^ equals ^/CTilwhere ^Lrepresents the bandwidth, A bandwidth of a few cycles seems advisable.However, a very narrow bandwidth brings in other considerations. If a

sine wave is impressed on a tuned circuit with a given Q, the ensuingoscillations do not commence at maximum amplitude instantly but build upexponentially to their maximum value in -^ t cycles of the impressed sinerwave.

Figure. 12. Exponential rise of sine wave signal impressed on a highQ circuit.


66/202


67/202

The signal occurring in the receiving coil is an exponentially d-creasing sine wave /f ^ t\l UJT If this signal were appliedto a low Q circuit it could be expected to build up very rapidly w.'.ileif it were applied to a very high Q circuit it would build up more slowly.Since the time constant of the exponential decay was approximately 1.6seconds, it was important that the signal not build up tor slowly for, ifit did, it would be decaying into noise before a satisfactory count couldbe obtained. This problem was investigated with LaPlace transforms andis included as appendix I, The results showed that for a C equal to 200,i.e., a bandwidth of approximately 10 cycles, the envelope of the waveformproduced would be as sketched below.

I.O

fi- ./ \i/>

\

/.o V

Feck

.z.

^V^

,i -t -y

^^Figure 13. Waveform produced by signal/^ ^ 4**^ U/T Impressed on a circuitwith a Q of 200

23


68/202


69/202

The waveform shows that the transient condition caused by the highQ circuit lasts only about one-tenth of a second which is not serioussince the total counting time expected to be available before the signaldecays into the noise is approximately two seconds.

The frequency of precession could be determined in various ways. Onedirect way thit inmadiately suggests itself is to amplify the signal tosufficient amplitude and count the frequency with some device such as aHewlett Packard 524A counter. The HoPc y^U\. will count a 2 Kc audio fre-quency by more than one method. It will count for 1 second of the signalfrequency with an uncertainty in the count of 1 cycle of the signal fre-quency which corresponds, for a one second count, to an uncertainty inthe earth's field of 25 J o Another nethod of counting possible withthe HoPo 52AA is to use the signal frequency, the 2 Kc, to gate on and

Ioff a 100 Kc crystal controlled frequency that is generated inside theHcP. 52/fAo In its normal use the gate would be kept open for 10 cyclesof the signal frequency during which 500 cycles of 100 Kc is counted witha possible error of 1 cycle out of 500 o Therefore, the count must con-tinue for much longer than 10 cycles of the signal frequency if accuracyon the order of 1 ^ is to be obtained. If the count were continued for4000 cycles of the signal frequency, 2 seconds of time, the number of100 Kc cycles occurring between the gating on and gating off pulses wouldbe approximately 200,000 and the uncertainty cf ona count in the 100 Kcwould correspond to a 5 Y^^'^^J^^i'^'^y ^^ ^^^ earth's field.

It might seem at first glance that the count should be continued aslong as any signals were available to count This is not true because ofnoise bursts that can occur and cause errors in the counting. The accuracy

24


70/202


71/202

will decrease by a factor l/e each tins the range is doubled after reach-ing some S/N ratio that first causes some small error to exist in thecounting. This is the critical S/N ratio. Thus there are two opposingeffectsone that produces a factor of two improvement in accuracy each

time the range is doubled; the other a l/e decrease in accuracy each timethe range is doubled after reaching the critical S/N ratio. The countshould always be continued to the critical value of S/N and to .567 of onetime constsmt thereafter This conclusion is justified in appendix II,


72/202


73/202

CHAPTBR IVEXPERIMENTAL EQUIPM3JT

The following system components will be described:1, Polarizing coils2, Receiving coil3, Sample of 'waterh. Preamplifier5. -\niplifier, Narrow Band6. Gating and counting circuits

7o Analoging circuit

lo Polarizing coilsThese two coils each consisted of 1200 turns of /^18 vdre. The D.C.

resistance of each coil was 25 ohmso They were wound on drum-like coilforms that could be moved relative to one another. The total inductancewhen the coils were placed as close together as possible was 1 henry. Thefield created by the coils vdth t>o amperes cf current flowing was 172gauss which is 244 times that of the earth's field. This field meets therequirement of being large as compared with that of the earth. The insidediameter was 8 which permitted use of a quite large water sajTiple, It wasnecessary to polarize the coil by supplying current for a period of severalseconds in order to allow the sample sufficient Xlms to reach its steadystate of maximum magnetization. Following this the current must be cut offlThis switching was accomplisted by a microswitch actuated by a slowly ro-tating wheel containing two notches. ;

The wheel rotated at 2 rpm which permitted one cycle of polarize andthen count during each 15 second period. A schematic diagram of the


74/202


75/202

t'cfO ^' ^(^

(D

Figure 14. Timing wheel

polarizing coil circuit is shown on page 2S. Terminals 1 and 2 are alsoindicated in the figure.

The sequence of action was as follows: when the microswitch was outof the notch, S]^ was closed o This in turn closed 82 and So. These relayscaused si and S5 to close., In this condition, which lasted for approxi-mately 10 seconds, the power supply provided two amperes of current to thepolarizing coils and the water sample was thereby polarized. As the ro-tation cf the wheel continued, the microswitch arm dropped into the notch

causing s-^ to open. Switches S2> S3, and s^i^ ooen immediately leaving thepolarizing coils in series with a resistor R]_ (220 ohms) and a capacitorC]^ (80,iH). This causes an overdampad pulse of current to flow in thepolarizing coil whose duration is approximately 20 milliseconds.

There is a delay of several milliseconds between the opening of 32,33, and s^, and that of sc. C2 was ^r. v^| and R2 was approximately 31 Ko


76/202


77/202

Fe h- f > < i ^^ 1tc/

Cvt kl< \

\\ X->

YT d^i\^

Co .1

~yMyI \j I

,

v^ i> i

jrr\yyy'\

\\-

J

Figure 15. Polirizirif; coil circuit.

28


78/202


79/202

Pc/a y^'r t yt s Cc ^ / ^ (^ ^/^

)-se- if-//^A / recU c ^i c a

Figure 16, Transient response of polarizing coil

\

The opening of Sr leaves the polarizing coils free to ring at the highfrequency but low amplitude deteniiined by the inductance of the coils andtheir distributed capacitance. See figure 16. The voltage rises initi-ally then decays for 2A raillisec(xids before the high frequency ringingccanmenceso This two-step cutoff was devised by Varian and greatly simp-lifies the transient problem. This is true because large transients inthe polarizing coils induce large, unwanted signals in the receiving coilwhich block the amplifying stages and cause ringing of the high Q circuits.2. Receiver coil

This coil was 4 long and had a 3.5 inside diameter. It consisted


80/202


81/202

of 3500 turns of )^26 ^dr^ and had an inductance of .235 henryc The D,C.resistance was 95 ohms. The A.C. resistance equals ths D.C, resistanceat the audio frequency of 2 Kc

Sufficient capacity to resonate the receiving coil was placed acrossitc

In the cutoff of the polarizing coils it is desirable to have a rapidcutoff so that the many small nuclear mom-^nts which combine to form thevector M will start precessing in phase and tJius produce a coherent signalExperience his shown that if cutoff is too rapid a smaller signal results.This is tentatively accounted for by a fanning out of the nucl3ar momentsthat combine to form Kwith a resultant destmctive phase interference.

On cage 20 the equation for the maximum Wns voltage at the grid ofthe first stage of the preamplifier was given. It depended on severalfactors such as the Q and number of turns of receiving coil, magnitude ofpolarizing field, area and susceptibility of sample and frequency of pre-cession. Usin/^ the vali^ s Just given in the description of the polarizingcoils and r'eceiving coil, the theoreticdj. value of maximum rms voltage tobe exoected at the grid of the first stage of the preamplifier is lo8millivolts. (The saii^ple area used was 213 sq. cm.)

Later, it was desired to check this value of loS millivolts at thegrid of the first stage of the preamplifier to see how it compared withthe amplitude of the signal at this ooinb obtained from an actual watersample. In order to accomplish this a water sample was polarized andthen allowed to pr


82/202


83/202

after pre.amplification was observed on an oscilloscope. An artificialsignal fron a signal generator was then introduced at the first grid ofthe preamplifier and its amplitude at that point adjusted until the pre-amplifier output as shown on the oscilloscope was the same as that obtain-ed with the actual signal. By this substitution method it was found thatthe signal from the actual sample of water produced a voltage of ,1 milli-volt at the grid of the first stage of the preampliiler rather than thecomputed value of 1.8 ndllivolts. The reason for this difference is notfully understood. Possible explanations include the fact that the equationused is for a long cylLndrical sample which did not in fact exist. Further,si-Tial redaction can be caused by coil orientations that are not optimum.3. Water sample

This was enclosed in a glass Jar that was fitted inside the receivingcoil, so that the coupling would be close. The actual sample area was 3or 7.62C^in diamieter and 3.5 or 8.9^vwlongo

A equals area of sample and was 213 sq. cm. Volume of sample was1620 err, . This can be compared with approximately 1 cm-^ samples that areused in most spectroscopy woric. Thus a great many more nuclei are beingpolarized here and the resulting signal is larger than that obtained inspec t rose ooy 'work,4. Preamplifier

This devi'ie was placed near the transmitting and receiving coil. Itwas built of n-dniature tubes and v/as powered with batteries on both fila-ments and 3-t- to reduce noise and 60 cycle hum. The schematic diagram isshown belowo All tubes were ck 628 Ratheon miniatures.


84/202


85/202

w

6-fJh- 'to\^

7* C (lyyt^l

Fisrure 1'', Preanplifier,,

This preamplifier provided amplification of the signal from .1 milli-volt to 7 millivolts. Thus, th'^ gain was 37 db,5. Narrow band Amplifier

Thsrs were no tuned circuits placed in the preamplifier so narrowbanding was required in the amplifier in addition to further ariplification.The required amplification could be'obtain-ed v.'ith one or two additionalstages. HowBTer, it was desirable to have an amplifier whose bandwidthcould be changed readily in order to observe the effect of narrow bandingon the signal. With this in mind a narrov-/ band, Q multiplier circuit (3) waconstructed.

The Q multiplier circuit consists of a cathode follower that has atuned circuit at the prid. It has an adjustable positive feedback thruresistor R.F. shown in the scherratic diagram, figure 15 =


86/202


87/202

(i^'

Z.0yo^

7. rAC x1}

-L_.

4AI


88/202


89/202

Oscillation will occur in a circuit with positive feedback only ifAk eauals onfi, io-?., if the gain tim^js the feedback factor equals one.In a cathode follower tha gain is sonewhat less than one. Therefore, by-keeping the feedback also just sli.^htly less than one, the feedback canbe kept below a critical value that i^rould caase oscillating to ensue. Thefeedback provides, in eff-^ct, a negative resistance that cancels a part ort^e equivalent parallel resistance of the tuned circuit thereby increasingits 0. It was fo'ond that Q's of 2000 were po3sible at 2 Kc sigial inputwith a resultin.^ bandwidth of 1 cyclao vVith the circuit shown in figure _,IB a bandvddth of 2B.7 cycles was measured for an input frequency of 2182cycle '.vith resistor Rf all in As Rf was reduced, the Q of the input cir-cuit increased to 200 while r-^taining good stability

It was found that if the C was increased to $70 a slight overshootdue to instability occurred when a sine wave was suddenly impressed. Thisis shown in the three photographs of figure 19 . The first photo shows thesif^nal tl^t resulted when the feedback resistor was disconnected. The riseof si.J^nal to final amplitude was very rapid. (Signal ;vas supplied byHewlett Packard audio si-^n.al generator). The Q of the tank circuit at thegrid of the tube when considered alone -A-as found to be 57. The secondpj.cture shovv's th-^ rise due to impressing a -^ve on the circuit with Kiadjusted so that the Q was increased to 20C. There is a measurable risetime evident as wouU be expected. In the third photo the Q was $70 andsome overshoot is evident, Q's used during the tests never exceeded 273.

It is oVviously impossible to read the bandwidth of a circuit witha of over 500 from th-i frequency dial of jn audio signal generator \henthe ceriter frequency is 2182 cyclr^s. The band^vidth measurements were


90/202


91/202

Figure 19. Photoc^raphs showing result of impressing a sine wave an ahigh Q circuit. (Continued on next page)


92/202


93/202

Firare I'^o Photographs shovdng result of impressing a sine wave on ahigh Q circuito

made by H.?, 5?4\ oountsr pl'os the binary scaling circuit which was con-structed to oe used as an integral part of the overall system. Thiscounting? system will be described in ^.Jri'^ next section.

Ffavinp; constructed a circuit to provide a narrow bandwidth it wasalso n-^ce3sary to have twD tintuned stages of amplification in order tohave an output signal of approximately 10 volts rraximum amplitude.

Narrow band amplifier.

B-K


94/202


95/202

6o CSatin?; and 'ôuntin.^ circuitsThere is available out of the amplifier a sigial of frequency near

2182 cycles. Its initiil amplitude is approximately ten Tolts and it isexponantiaily decaying with a time constant of approximately 1,6 seconds.It has an initial S/N ratio of 20 or 30 to 1. During the 1st 100 milli-seconds a transient condition exists that must be avoided. The frequencyof this signal must be detennined to v/ithin a small fraction of one cycle.

The H.ô 524A recîred ,85 volts of input signal to trigger the firstsquarinf^ and amplifying stages. An input signal sine wave greater than,85 volts would (with the fr^squency-perlod switch set to period) producepositive rectangular pulses (shaped signal pulses) of about 30 volts amnli-tude at oin one of the decade divider socket. This decade divider thensupplies output pulses to pin 2 of tha same socket that are l/lO the fre-quency of the incoming si^îîj. . These pulses act as gating pulses for theinternally generated 100 Kc


96/202


97/202

cm^

-_i rI

Figure 21 One stage of binary circuit


98/202


99/202

CD

Fip;ur 22o Binary gating counter.

-.\.A


100/202


101/202

It was important that the counting not commence until 100 milli-seconds after the polarizing coil was cut cf f in order to avoid a tran-sient condition. (See page 23 and Appendix I) However, no further delaywas desirable if the maximum counting time .vas to be utilized. The delay-was introduced by placing a second microswitch so that it was on theopposite side of the slowly rotatirg wheel from the polarizing coil micro-switch o

cot I


102/202


103/202

The physical position of the second microswitch was arranged sothat it dropped into the notch 135 milliseconds after microswitch #1,This introduced sufficient but not excessive delay. The second micro-switch caused a relay to open and close that was connected to the reset ofthe binary gating counter. By dropping into the nctch^ switch s^^ in draw-ing 22 was closed bypassing the 50000 ohm resistor. When microswitch #2was out of the notch, the 50000 ohm resistor was placed in series withthe grid resistor in one stage of each of the pairs of bistable circuitsin the binary gating counter. This unbalanced each pair and caused each

of the stages which had 50000 ohm introduced to become the on tube ofits pair. The first squaring amplifier of the H.P, 52/A counter limitedits output to 20 volts regardless of input signal amplitude. This out-put when fed to the binary gating counter out of pin 1 of the DecadeDivider Socket was not sufficient to trigger ths binary gating counterprovided s^ was open with the 50000 ohm in the grid lead of every otherstage. Therefore large transients which cccurred prior to the tiire thatmicroswitch #2 dropped into the notch did not start an erroneous count

When microswitch #2 dropped into the notch closing 35 and bypassingthe reset resistor, the binary gating counter was ready to receive inputsignal pulses. It was necessary for ths first input pulse to produce anoutput pulse which would open the gate on the 100 Kc and start a counton the HoP. 52/fA counting stages. Since this was true, the binary gatingcounter was not arranged in the conventional manner. If it were, thenthe first output pulse would occur only after 4096 input sigial pulses andadditional pulses would follow spaced 4096 cycles of the input sigial fre-quency apart. In order to cause the first input signal pulse to produce


104/202


105/202

an outout puls , i.., go all the way through, the reset resistor waintroduced into the opposite stage of each pair than it would have beennormally. This is the condition that would usually exist after 4095cycles of the input signal frequency. Thus the first pulse started acount and /+O96 cycles later the count was stopped by the next pulseo

In order to determine tbs overall accuracy of the counting equip-ment, a Hewlett Packard low frequency standard of 1000 cycles was countedfor 8 seconds. The results showed the device to be accurate within thelimits of the crystals which is 1 part per ndllion plus 1 cycle of 100

KCo For data see appendix III,7. Analoging circuit

For the data collected the precession frequency was counted and theanswer indicated on the lighted dials of the H.Po 524A counter. For scaneapplications, it would be preferable to have the result of the successivemeasurements continuously recorded on a tape recording voltmeter such asthe Esterline-Angus or Browno It is possible to accomplish this by usinganother binary counting circuit that will count to 64. To this circuitcan be fed the same 100 Kc freouency that is counted by the HoPo 52AAwhen the gate is opened by the 2 Kc signalo The total number of 100 Kccycles occurring during a 2 second count will be about 200000 and theactual number of cycles will differ from this sli^tly depending on theexact frequency of precession. When these cycles are fed into the binarycounter capable of counting to 64, it will count thru its full range alarge number of times and when the 100 Kc is gated off will stop counting.The count indicated at this time on the binary will detennine some refer-ence numbero If the next count then leaves a different remainder on the


106/202


107/202

binary indicator lights, this will correspond to a change of field. Achange of one cycle of tha 100 Kc will cause a change of one count inthe binary indicator vrfiich corresponds to a change of very nearly k^inthe earth's field. If, in the plate circuit of one of each pair ofstages in the binary circuit is placed a relay switch that switches ina voltage of a ceiiiain amplitude then the 100 Kc frequency count can beconverted to a voltage amplitude. The weight given to the six stagesof this binary counter are 1, 2, U, S, 16, and 32 so that the voltagesswitched in by the vairLous pairs of stages should be proportional to

these numbers. The total voltage or scsne fraction thereof can then beapplied to a recording voltmeter.

Such a circuit was built and tested with a synthetic signal but timedid not permit its incorporation into the overall systemo


108/202


109/202

CHAPTER VaXPSRIMSNTAL RESULTS

Data were collect-^d far a plot :^ the diurnal change of the earth'smagnetic field. In general during the day data were collected for ten min-utes out of each hour, (A total of 40 pieces of information was collectedin a ten minute period) c Data were collected continuously for certain per-iods of the day in order to establish short term trends. A plot of thedata 15 given in figure 24 o The tabulated data is appendix IV Severalthings of interest are: the polarizing and receiving coils were placed

approximately 50 feet from the main laboratory building in order to reduceinterference o However, some interference still existed but was less atnight when the laboratory was quiet o Also, the effect cf the sun is lessat ni^ht The data collected starting at 2331 shows a period of over 7minutes when the indicated change in the earth's field was only t -j | Data taken starting at 0029 shows a drift in the earth's field of 3 Ifdur-ing a ten minute period with a clear trend apparent in the data. Thissame type of trend is evident at other points in the datao Data collectedbetween 0750 and 0820 is believed to have superimposed on the normal changean effect d'je to th? arrival on the Varian parking lot of ab^ut 200 auto-mobiles during this period. Data were collected continuously and a totalchange of 7 Y ^^^ noted during this time. It is also noted in the data thatat the end of the 24 hour cycle the readings indicated a return to the start-ing point of the previous dayo

An additional piece of information on equipment sensitivity was obtain-ed by moving a 3 foot length iron pipe 5 inches in diameter from a point 50feet distant fron the polarizing and receiving coils to within 15 feet. Animmediate shift of over 12 f was noted in the data, (See appendix V)


110/202


111/202


112/202


113/202

TiAPT?:R VICONCLUSIONS

1, Possible Errors:It is believed that changes in the earth's field of t i can be

detected, vdth the present equipment c onfi.juration . The limitation isimposed by the fact that the count is for only 2 seconds and the basecounting frequency is 100 Kc, If this base frequency were increased to5fXI) Kc then changes of l/20 Y should be detectable. Accuracy of courseis also deoendent on the stability of the oscillator creating the basefrequency but with crystal controlled oscillators errors from this sourceshould be less than one part per million o Noise is another possiblesource of error but with initial S/N ratios of 20 or 30/l, this is not aproblan unless long counts are attempted.

Accuracy with ^ T for measurement of the absolute value of the earth'sfield was achieved Greater accuracy than this is not claimed since thegyronagnetic ratio as measured by the Bureau of Standards was only accurateto within one part in 40000. Also the exact way in which the squared upsignal wave crosses the axis affects the accuracy of the count slightly.2. General Comments and Applications

The accuracy of the s^'stem which has been described depends solely onthe accuracy of measurement of the frequency of precession. The orien-tation of the coils need not be exact and the precession frequency is notaffected by temperature or humidity changes. Further, the equipment doesnot operate as a rate of change instrument and does not require motion.

The equipment exclusive of the frequency counting devices could bereduced to a 25 pound packagesomething that could be carried by one mano

46


114/202


115/202

The 2 Kc sip^nal could be telemetered to a remote spot where its exactfrequency could be determined.

This system would obviously have applications as a station magneto-meter for a magnetic observatory. It could also be used for magneticprospecting either as an airborne or ground equipment. Possible militaryapplications in harbor defense and anti-sutanarine warfare are also evident.


116/202


117/202

APPKJiDIX J

rI

Q

V

v


118/202


119/202

r --- - X o) ( (i / ^)


120/202


121/202

^4^ ^ ^ A j^

A .- - i f^T ,- ^-^>


122/202


123/202

/ (^ ) -- K (V U^(^ f / - .> )(^ ^ X i-jis) 0-i u^ yy.^N 'j y)

^;;>a- f ^J) g


124/202


125/202

r) < ^i ^^iif) f^/i ^^


126/202


127/202

^ei (^i ^-.frf^)- S


128/202


129/202

M o^ .- 1 (A- ^00 ci^ci f^^^J t l/^< ^ T- ^ o oo ^

- Vo o

K./- c >.?

^1'^

/^

f'/^r( K ^0

tct 2. ^

- j/.f

I. ^ >I o

fv ^x- ^0


130/202


131/202

-^ 1/ 3^^o ^ i:i.sL

-i*e-^ / i f)-

^ >.. k'4x/oJ.to.g1o


132/202


133/202

I > . 5r4

r^ - zi \h)r ^ OJ

^^ - |, 3>i^ e3I J jLfO

^0CaJ , 4 ^ ^^ J J t/ H^ - ^ l^* *^ \ oUS V


134/202


135/202

APPENDIX II

- a/ - o

l.o

,^-^4,7


136/202


137/202

APPENDIX III1 8192082 8192083 819208k 8192085 8192086 8192097 8192098 8192099 819209

81920881920881920881920981920981Q209819208819208819209819209819209


138/202


139/202

172B-381820-281933-432030-382129-392229-392331-a0029-390123-330215-250315-250415-250515-250615-250715-250952-021052-021?08-181243-5314^1-411532-421632-42

APPENDIX IVSUMMARY CF DATA TAK^N 2, 3 I^iRCH187710.6187747 08187754 0818776lo7187713c187711cl187712 187708 08187705 o187712,4187715.4187711c187701 06187689 o187707 ol187802 ol187830 oO187809 oO187802.4187761 o1877 51 o187704 ol

59


140/202


141/202

172B1 17,70$2 093 07U 095 086 117 108 109 14

10 1111 1112 1413 13U 0915 1016 1117 1318 1219 1420 11

17^21 1877102 123 134 125 116 127 148 129 14

10 1211 1412 1613 1514 1415 1616 1417 1618 1419 1620 16

60


142/202


143/202

17^81 1877U2 173 194 U5 166 207 198 16

9 1610 1711 2012 1813 17U 1815 1716 2017 1918 1819 1720 18

174>1 1877172 183 20k 175 206 21

7 218 21

9 a10 2311 2212 2213 2414 2215 2216 2517 2618 2719 2320 29


144/202


145/202

174^.31 1B77282 283 32U 285 276 307 308 28

9 2910 2811 3012 3013 29IZ^ 2715 3016 3017 3218 3319 3320 31

17?3o?1 1877312 313 344 325 356 377 338 31

9 3410 3411 3212 3213 3514 3615 3516 3217 3418 3419 3420 35

62


146/202


147/202

17591 1877342 353 38U 365 386 447 428 44

9 4110 44ll 4412 4713 44U 4915 4916 4317 4418 4819 4720 47

I8O41 1877462 503 514 495 496 497 518 50

9 5010 5311 5012 5513 5214 5415 5516 5517 5318 5719 5620 56

63


148/202


149/202

1^081 1B77542 563 554 575 576 607 608 62

9 5810 ~ 5911 6012 5713 6314 5815 5916 5917 6218 6119 5920 59

18141 1877642 613 594 595 596 587 59d 59

9 6110 5911 5812 5913 6014 5715 5616 5817 5718 5519 5620 54


150/202


151/202

18201 18775A2 533 524 505 536 507 498 509 4710 4611 4712 4613 51li^ 4515 4516 4517 4618 4219 4420 42

^m1 1877532 533 544 545 576 557 568 559 55

10 5611 5612 5613 55U 5415 5516 5517 5418 5519 54

20 55


152/202


153/202

1 1877092 093 054 045 046 037 028 019 01

10 1876 9711 9612 9613 9114 9215 9016 8717 9218 8719 9020 85

21 1876 8622 8523 8524 8525 8726 8727 8828 8729 8630 8531 8632 8733 8734 8735 8436 8537 8538 8839 87

40 86


154/202


155/202

21291 1877 132 13

3 124 125 126 127 148 129 11

10 1111 1112 1113 11U 1315 1116 1217 15IS 1319 12

20 13

2L 1322 14

23 1424 1425 1426 1427 1428 1529 1530 1531 1632 1633 1734 1735 1736 1837 1738 1739 1740 18


156/202


157/202

22291 1877122 133 134 125 106 127 118 11

9 1210 1211 1212 1213 UU 1215 1116 1217 1118 1119 U20 11

21 18771022 1023 1124 1125 1026 1027 1026 1029 1130 1131 1132 1233 U34 1135 1136 1037 1138 1039 10

40 10


158/202


159/202

23311 1877132 U3 124 125 126 137 128 12

9 1210 1211 1312 1213 1214 1215 1216 1217 1318 1319 1220 11

21 18771122 323 324 325 226 227 328 2

29 230 231 232 233 334 235 236 337 238 239 3

40 2


160/202


161/202


162/202


163/202

012;?

1 1877072 83 74 75 76 77 78 79 7

10 711 612 713 7U 615 716 617 718 719 720 7

21 622 423 624 625 526 527 328 2

29 430 231 332 333 234 335 236 237 338 339 4

40 4


164/202


165/202

02151 1877132 33- 2U 35 26 27 28 2

9 510 211 212 213 3U 215 216 117 2Itt 1

19 120 2

21 18771122 223 224 225 126 227 228 229 230 131 332 333 334 335 336 437 438 439 4

40 5


166/202


167/202

^un1 1P7714? 213 '154 235 266 197 17fl 099 16

10 0911 1712 1813 1914 1315 1216 1717 24IB 1919 20

PO 10

n 18771122 1423' 1924 1825 1726 1927 2028 1729 1930 1431 0932 1333 1334 1135 1236 1137 1138 1139 11

40 11


168/202


169/202

COp1 1^710? 113 10U 095 096 107 108 10

1010 0911 09IP 101-3 0914 1015 1016 1017 . 101^ 101^ LI20 09

21 ifr771022 11^3 122k 1225 1226 1227 122^ 1129 1130 1231 1332 1333 1334 1535 1536 1437 U3B 1439 14

40 16


170/202


171/202

0^1?1 1877132 103 094 085 076 077 068 069 0510 0711 0612 0413 0214 0215 0316 0117 0118 0019 00

20 697

21 18768722 8723 9724 9825 9826 9927 9928 70029 70030 69931 70032 70033 70134 70035 70036 70137 70238 70139 701

40 702


172/202


173/202

06151 1876912 913 924 935 946 947 938 929 91

10 9111 9012 8913 89U 8515 8616 8617 8418 8619 87

20 87

21 18768722 8623 8424 8625 8826 8927 8828 8829 8930 8931 8832 9033 9034 9135 9136 9137 9238 9139 92

40 91


174/202


175/202

07131 1877082 63 7U 85 66 57 78 79 910 8

07481 187699 21 6952 97 22 6953 99 23 696U 700 24 6975 699 25 6956 699 26 6967 697 27 6968 698 28 6979 699 29 694

10 698 30 69711 699 31 69712 700 32 69813 69914 69815 69716 69717 69718 69619 696

20 696


176/202


177/202

1 1876992 6983 698U 6995 7026 7027 6998 7009 700

10 69811 69912 70013 69914 700

15 70116 70017 70118 70119 70220 . 70221 70422 70523 70324 705

(0800)

(0802)

(0805)

25 703 54 71826 706 55 71827

28 705 (0807.5)29 70730 70531 70332 70533 70434 70835 708 (0809.5)36 70737 71138 709

39 70840 70641 70742 71043 710 -44 71245 71446 . 71547 71748 717 (0814)49 719 '

50 71851 71852 718


178/202


179/202

1 1877972 98

3 98k 985 976- 987 968 979 98

10 80211 80112 80313 80114 80215 80016 80217 80118 80019 80020 803

21 18780322 803

23 80324 80325 80526 80427 80328 80329 810


180/202


181/202

10^21 1878312 2

3 3U 35 2

6 47 48 59 3

10 411 512 413 314 313 416 117 318 219 320 2

21 18783122 3223 3724 3025 2726 2627 2828 2929 2830 2731 2832 2933 2934 2835 3336 3237 3238 3239 33

40


182/202


183/202

12081 187B082 083 094 085 086 087 078 079 06

10 0511 0612 0713 0714 0715 0816 0917 1118 1019 1020 U

21 18781122 U23 1124 U25 1026 1027 1028 0929 0930 0931 0932 1033 0934 0935 1036 U37 io38 1639 11

40 12


184/202


185/202


186/202


187/202

U311 187768 21 1877622 67 22 633 67 23 65U 66 24 615 64 25 616 69 26 627 62 27 598 187762 28 599 62 29 5810 60 30 5911 61 31 5812 61 32 5913 62 33 5814 62 34 5915 62 35 5816 62 36 5817 61 37 58IS 61 38 5819 62 39 57

20 62 40 57


188/202


189/202

1 1877522 523 44 15 26 17 18

9 , 21011 412 413 314 315 216 317 118 519 320 4

21 18775422 123 224 325 226 22728

29 330 131 5132 4733 5134 4835 4736 4737 5038 49J) 4840 48


190/202


191/202


192/202


193/202

kpp'^mix V

Pipe at 50

1 1875762 1875753 187578h 1875795 1875766 1875777 1875748 1875789 18757910 187581

Pipe moved to 15 '

11 18752112 18752413 187521

14 18752015 18752216 18752217 18752318 18752219 18752520 187525


194/202


195/202

1. Bloc'r, Fo

2. Bloc' , F Hansen, V/.'V.and PiCkari, Mo

3. ''llectror.ics Hay 1951Uo Lf^vinthal and Rodgsrs

5 . Mc '^omb ,

BIBIJOGR/iPHYPhysical Revisw 70 kSO (1946)Physical Review 20 474 (1946)

Simplified Q MulUoliftr P 130Measurements of low magnetic fieldsusing paramagnetic resonanceReport prepared fcr University ofCalifornia Rad lab P.O. #H-246$6 April6, 1951Magnetic Observdtor>' Manual UoSoDepartrrert of Commerce SpecialPublicaUon 283 1952

6a NOL Laboratory Report 9377. Packard, Dr. Martin

Purcell, F.N'c Torey, J.C,Pound, R.VoWyckoff, R.D,

1945Method of nuclear induction 'thesis3tanfordPhysical Review 69, P 1%6The gulf airborne magnetometerGeophysics volume 13 pp4'^-56 1936


196/202


197/202


198/202


199/202


200/202


201/202

1 N0V7^15 U

the fre^ ' Qprotons. I^,* ?

et

^42^263

Bowen.\n Earth's field iragiietom^ter

that utilizes the free precessicaiof protons.


202/202

thesB74

^If, : }^'^ f'eld magnetometer that utili

3 2768 002 07364 5DUDLEY KNOX LIBRARY

Documents

An Earth's field magnometer that Utilizes the free precession of Protons