a period measuring proton magnetometer with a direct readout

7
A Period Measuring Proton Magnetometer with a Direct Readout Giinter Schulz and Uwe Carstens UDC 550.380.8 Summary A period measuring proton magnetometer (PLUM) with a programmable digital computer connected to it is described. The computer converts the time-propor- tional measured value into a frequency-proportional, and thus field-proportional, readout. Further operations like the calculation of mean values are also applied to the measured result on line via the computer. In addition, a printer protocol is supplied. A highly stable quartz is the basis for the period measurement. Its data and the dimensioning of the counter depend of the requirement of a field-independent resolution of about 0,1 nT. Results of a long-time comparison speak in favour of the concept of period measurements. This comparison was carried out at the Deutsches Hydrogra- phisches Institut, Erdmagnetisches Observatorium, Wingst, using this instrument and a frequency measuring PRM. Ein Perioden messendes Protonenmagnetometer mit Direktanzeige (Zusammen- fassung) Es wird ein Perioden messendes Protonenmagnetometer (PRM) mit einem nachgeschalteten programmierbaren Digitalrechner beschrieben. Dcr l~echner wandelt den zeitproportionalen 1V[el~wert in eine frequenz- und damit feldpropor- tionale Anzeige urn. Weiterfiihrende Operationen, wie Mittelwertbildungen, werden ebenfalls fiber den l%echner on-line auf das MeBergebnis angewendet. Zus~tzlich wird ein Druckerprotokoll erstellt. Ein hoehstabiler Quarz bildet die Grundlage ffir die Periodenmessung. Seine Daten und die Dimensionierung des Z~hlers sind durch die Forderung nach einer feldunabhi~ngigen AuflSsung yon etwa 0,1 nT bestimmt. Fiir das Konzept der Periodenmessung sprechen Ergebnissc eines Langzeitver- gleichs, der am Deutsehen Hydrographischen Institut, Erdmagnetischen Observa- torium, Wingst, mit diesem Ger/~t und einem Frequenzen messenden PI%M durch- geffihrt wurde. Magn6tom~tre h protons h mesure de p6riode avec lecteur direct (R6sum6) On d6crit un magn6tom~tre s protons s mesure de p6riode (P1RIV[) avec calcu- lateur num6rique programmable connect6. Le calculateur eonvertit la valeur mesu- r6e, proportionnelle au temps, en une lecture proportionnelle ~ la fr6quence, donc au champ. Des op6rations suppl6mentaires, telles que le calcul de valeurs moyennes, sont aussi effectu6es en temps r6el par le calculateur sur les donn6es mesur6es. En outre, un 6tat imprim6 est fourni. Un quartz de haute stabilit6 est la r6f6rence pour la mesure de p6riode. Ses donn6es et le dimensionnement du compteur sont fonction d'une r6solution sp6cifi6e d'environ 0,1 nT ind6pendante du champ. Les r6sultats d'une comparaison de longue dur6e sont en faveur du concept de la mesure de p6riode. Cette comparaison fur men@ h bien au Deutsches ttydro- graphisches Institut, Erdmagnetisches Observatorium, Wingst, en utilisant cet instrument et un PI%M ~ mesure de fr6quence.

Upload: guenter-schulz

Post on 16-Aug-2016

218 views

Category:

Documents


4 download

TRANSCRIPT

Page 1: A period measuring proton magnetometer with a direct readout

A P e r i o d M e a s u r i n g P r o t o n M a g n e t o m e t e r w i t h a D i r e c t R e a d o u t

Giinter S c h u l z and Uwe C a r s t e n s UDC 550.380.8

Summary A period measuring proton magnetometer (PLUM) with a programmable digital

computer connected to it is described. The computer converts the time-propor- tional measured value into a frequency-proportional, and thus field-proportional, readout. Further operations like the calculation of mean values are also applied to the measured result on line via the computer. In addition, a printer protocol is supplied.

A highly stable quartz is the basis for the period measurement. I t s data and the dimensioning of the counter depend of the requirement of a field-independent resolution of about 0,1 nT.

Results of a long-time comparison speak in favour of the concept of period measurements. This comparison was carried out at the Deutsches Hydrogra- phisches Inst i tut , Erdmagnetisches Observatorium, Wingst, using this instrument and a frequency measuring PRM.

Ein Perioden messendes Protonenmagnetometer mit Direktanzeige (Zusammen- fassung)

Es wird ein Perioden messendes Protonenmagnetometer (PRM) mit einem nachgeschalteten programmierbaren Digitalrechner beschrieben. Dcr l~echner wandelt den zeitproportionalen 1V[el~wert in eine frequenz- und damit feldpropor- tionale Anzeige urn. Weiterfiihrende Operationen, wie Mittelwertbildungen, werden ebenfalls fiber den l%echner on-line auf das MeBergebnis angewendet. Zus~tzlich wird ein Druckerprotokoll erstellt.

Ein hoehstabiler Quarz bildet die Grundlage ffir die Periodenmessung. Seine Daten und die Dimensionierung des Z~hlers sind durch die Forderung nach einer feldunabhi~ngigen AuflSsung yon etwa 0,1 nT bestimmt.

Fiir das Konzept der Periodenmessung sprechen Ergebnissc eines Langzeitver- gleichs, der am Deutsehen Hydrographischen Insti tut , Erdmagnetischen Observa- torium, Wingst, mit diesem Ger/~t und einem Frequenzen messenden PI%M durch- geffihrt wurde.

Magn6tom~tre h protons h mesure de p6riode avec lecteur direct (R6sum6) On d6crit un magn6tom~tre s protons s mesure de p6riode (P1RIV[) avec calcu-

lateur num6rique programmable connect6. Le calculateur eonvertit la valeur mesu- r6e, proportionnelle au temps, en une lecture proportionnelle ~ la fr6quence, donc au champ. Des op6rations suppl6mentaires, telles que le calcul de valeurs moyennes, sont aussi effectu6es en temps r6el par le calculateur sur les donn6es mesur6es. En outre, un 6tat imprim6 est fourni.

Un quartz de haute stabilit6 est la r6f6rence pour la mesure de p6riode. Ses donn6es et le dimensionnement du compteur sont fonction d'une r6solution sp6cifi6e d 'environ 0,1 nT ind6pendante du champ.

Les r6sultats d 'une comparaison de longue dur6e sont en faveur du concept de la mesure de p6riode. Cette comparaison fur men@ h bien au Deutsches t tydro-

graphisches Insti tut , Erdmagnetisches Observatorium, Wingst, en utilisant cet i n s t r u m e n t et un PI%M ~ mesure de fr6quence.

Page 2: A period measuring proton magnetometer with a direct readout

120 Dr. hydrogr. Z. 32, 1979. It. 3. Schulz e~ al.: Period Measuring Proton Magzaetometer

Introduction Starting from the relation

B = 2=y -1 ]p

where : ]p = the frequency of the free precession of the proton;

= the gyromagnetic ratio of the proton;

the measurement of the amount B of induction using a PRM is referred to a frequency de- termination.

The transfer factor is

2=7 -1 = 23,48743 (1 +_ 5 �9 10 -6) nT s ( W e y a n d [1978]).

Its tolerances limit the absolute precision of the measurement. In addition, the measured value is systematically falsified by transverse relaxation processes (Wiese, S e h m i d t et al. [1960]) and by possible magnetic impurities in the sensor.

With earth field measurements the amount of the transfer factor may be expected to cause precision frequencies of up to some 103 s -1. The relaxation processes cause an exponen- tial decay of the signal and thus of the signal-to-noise ratio. For an individual measurement a counting interval of several seconds is therefore available in the most favourable case. A direct measurement of frequencies by counting of periods throughout this time allows only a quantization of the order of magnitude of 10 nT.

Two methods are suited for achieving a smaller qnantization step :

1. Measurement of the duration of a given number of periods of the precession signal in periods of a crystal time base whose natural frequency of some 10 -5 s -1 is sufficiently high (instrument of the first generation).

2. Multiplication of the frequency of the precession signal by a suitable factor (abt. 102) prior to its counting within a fixed time interval (instrument of the second generation).

Whilst, in the first case, one obtains a value which is inversely proportional to the in- duction to be measured, the second method supplies a direct display which may be adapted - with appropriate adjustment between time interval and factor - to the valid SI unit (nT).

The multiplication of a frequency, however, is most sophisticated if - like in this case - the signal is disturbed by noise and if only the order of magnitude of the frequency is known. This problem is solved technically by means of a voltage-controlled oscillator (VCO) whose frequency is higher than the precession frequency by the factor of the desired mul- tiplication. The VCO is controlled in a phase-locked loop (PLL). After a division of the VCO frequency by the constant factor this frequency is traced in the PLL of the precession fre- quency. The induction B to be expected determines the coarse tuning of the base frequency of the VCO and thus, via the dynamic range of the VCO (VCO range), the acquisitionrange of the PLL. The signal controlling the VCO passes a variable low-pass which attenuates the disturbing influences of all irregularities of the precession signal on the VCO frequency.

(~omparison At the Erdmagnetisehes Observatorium, Wingst, measurements with both methods

have been undertaken since 1973 (with some interruptions). The principle of the first gene- ration has been represented by an instrument of the type V4931 of the firm Varian, and the principle of the second generation by an instrument of the type V75 of the same firm. The long-term comparison shows significantly systematic differences of the measured values between both instruments of some 0,1 nT. As the same probe-cable configuration is used for both instruments, the cause lies in the different counting electronics.

V o p p e l [1977] suggests that the error in the instrument must be sought in the second generation. I t is important for this statement that the distribution of the differences is ob- viously correlated with the VCO ranges of this instrument.

In the first approximation the correlation is linear. Fig. 1 shows the regression lines of two VCO ranges after an adjustment of the instrument V75 by the firm Varian, with

Page 3: A period measuring proton magnetometer with a direct readout

Dr. hydrogr. Z. 32, 1979. I-I. 3. Schulz et al.: Period Measuring Proton Magnetometer 121

T

AB

1,0

n T

O_

-o,s

- -1,0

T - %%'".~0 ~ / S ~ ~r ~

�9 . . /~r / i m t " ~ "-'~ ~

/ �9 / / 3 d-confidence belt J f / �9 /[ ~ /�9 �9 of the regression line

~: " Z F t I

io io io 6b B "-

Fig. i. Differences of induction values measured with the Proton Magnetometers V75 and V4931 in the sense 'V75 minus V4931' depending on the induction itself. The values are linearly smoothed

by the principle of least squares for two VCO ranges of the V75

the generating values. A jump of about 0,7 nT in the overlapping area of both VCO ranges becomes obvious, a feature which is well secured by additional relative measurements and which is also true of the non-represented subsequent VCO ranges.

Fig. 2 shows the time-dependent processes of the deviations for three discrete values of induction _F -- 4,85 �9 104 nT, H -- 1,8 �9 104 nT and Z = 4,5 �9 104 nT by running quarterly

T / (I- r onge %0 T T +

+ + / j _ v z \ V ~ \T + + +

p ~ A B T ~_ T

T

nT ~\ "L~ T T T

- V l . . . . . A Z • T i i

+-+ I I 1

S\TT :-:! ! ! - ! - ~ T

o d "A N'L, - I ~\~/~ ~\~

0 ~ %--'-

i

-o,,-- I

- I ~L--i I\T ~\~x -T T/I •

] T.!' RT 7/ • 1 H,Z:VCO r o n g e l I"/ i I • ~\!/ F." VCO ronge 2 x/~ i 1 -%0- I

I ~ 1 I I I [ I I I I I I I I I I I I I I I I I

1973 1974 1975 t --,,,.-

Fig 2. Differences of induction values measured with the Proton Magnetometers V75 and V4931 in the sense 'V75 minus V493t' during a time interval of two years. The values are smoothed by

overlapping quarterly means

Page 4: A period measuring proton magnetometer with a direct readout

122 Dr. hydrogr. Z. 32, 1979. I-I. 3. S c h u l z et al . : Per iod Measuring P ro ton Magne tome te r

means throughout the period August 73 to May 75. Their amounts correspond to those of the components of induction in Wingst, measured by means of the method of compensation ( V o p p e l [1972]). H and Z were measured in the VCO range 1, and F in the VCO range 2.

has not been represented for 1973, since it is uncertain which VCO range was used for this period.

The enlarged symbols for H and Z in August 1973 mark the values to be expected accor- ding to Fig. 1. Conversely, the respective symbols in the field-dependent representation show the situation of the temporally coinciding three-monthly means. The results are nearly identical within the la- l imits .

What is important is the knowledge tha t the long-term changes of the deviations are of the same order of magnitude as the deviations themselves. So, e.g. AH changes within two years by about 0,7 nT. This behaviour in particular is not in accordance with the re- quirement of high stability for base-line instruments.

Therefore, the IMS standard will also in future be derived from the older version of the first generation. To meet the requirement for a direct reading with a field-independent quantization of about 0,1 nT, the original counter of the instrument was replaced and a calculator was connected to it.

Technical description

P r e s e l e c t i n g c o u n t e r The preselecting counter was built in the experimental workshop of the Deutsches

Hydrographisches Insti tut . Input of the counter is the highly amplified and limited pre- cession signal (rectangular) as well as a trigger signal of the analogous unit. Output is the BCD-coded counting n of the period trigger pulse for the subsequent computer.

] I supply

Components of the Vor iometer V 4931

ampl i f ier , voriable narrow- band f i l ler , l im i te r

trigger- signal

r I

N [

preselecling counter ]

memory : p rogrum

2 TF'f N , - - ~ - ~

processor

B" 2 ~ N f

1 K B = ~ E B i

i=I

l i s l ing : p r o g r a m

ni , B

I computer RD - 3Pj

I Tows Elektronik

I Fig. 3. Block diagram of the Direct Reading Proton Magnetometer

The t ime interval for counting the crystal frequency is given by the preselected number of passing periods of the precession signal. I t has proved useful to wire frequently occurring oscillating numbers - besides the sensitive incremental preset N - so as to enable their being dialled quickly. Field-dependent and configuration-dependent opt imum values are obtained for this preset N: the error of quantization represents the lower limit, the signal- to-noise ratio of the precession signal at the end of counting represents the upper limit.

Page 5: A period measuring proton magnetometer with a direct readout

I)t . hydrogr. Z. 32, 1979. H. 3. S c h u l z et al . : Period Measuring Proton Magnetometer 123

The sca t te r ing of t he wave f ront of the signal a t the level of the t r igger threshold , a t the t ime of t r iggering, is a measure for the por t ion of noise. The measu red va lues using a hexan probe are shown in Table 1. The r e l a t i ve ly low usable count ing in te rva ls for the in- duct ion components Z and H follow f rom the remain ing inhomogenei t ies of the superim- posed bias fields dur ing the measu remen t of components .

Table i

Component

F

Z H

Induct ion

104 n T

4,85

4,51 1,80

Count preset

(7500) 5000

4000 400

Count interval

S

(3,63) 2,42 2,08 0,52

in i t i a l signal -to -

noise ratio

dB

30

3O 2O

Error of noise,

relative

3" 10 -~ < 10 -~ 8" 10 -7 4" 10 -~

Error of quantization,

relative

7- 10 -~ 10-6

1 ,2 .10 -~ 5- 10 .8

Tak ing as a bas is a count ing f requency of / = 4 �9 10 a s -1, the quan t i za t ion steps, too, become sufficiently small .

I n order not to fal l shor t of t he usua l sampl ing r a t e of 1/(6 s) the wired prese lec t ion for F was es tab l i shed to be 37 = 5000.

Dr i f t s of count ing frequencies are p r e d o m i n a n t l y due to t empera tu re . They fa ls i fy the measurements sys t ema t i ca l ly and m u s t be contro l lable as far as t h e y are no t compensa ted b y t empera tu re - s t ab i l i s ing devices. The control of the count ing frequencies is done b y com- par i son wi th a h igh ly s tab le reference f requency of 1000 �9 (1 +__ 2 �9 10 -7) s -1 before and af ter each measurement of the vector.

The funct ion of the presetect ing counter m a y be seen f rom the block d i a g r a m in Fig. 4. The t r igger signal of the analogous un i t causes, a t first, a reset of the d i sp lay COmlter and a

Out: GATE o CON TROL Out: SIGNAL CON TROL

Cfystdl I Oscitld tot Di v i der

In. fp e ~ I SIgnal- PRECESS/OiV-sIGNAL l, Amplifier

#RIGGER~LEVEL

+ I

In: I Reset/Stort TRIGGER ~ I Deley

I

SIGNAL 1 ISTART DELAY

Signal - Gate

__ stort 5top

Sote-FF

- - ~ T e s t

4 Digit L~Nd Preset Coun ter

l l l l l l l Comparc~tor I

6Digi t

Displdy Counter n

Computer Interface ]

Out: PRESET n

o Out: COMPUTING C OMMA ND

Fig. 4. Direct l%cading Proton Magnetometer. Basic block diagram of the prcselecting counter

loading of the prese lec t ing counter wi th t he p resen t N. Af te r a d e l a y of 30 ms to 300 ms the measu remen t s tar ts . Via t he signal ga t e and the t ime base ga te t he precession signal fp and t h e t ime base signal [ get in to the preselect ing counter or into the d i sp lay counter , res- pec t ive ly . The preselect ing counter now counts backwards f rom 37 a n d finishes the measu-

Page 6: A period measuring proton magnetometer with a direct readout

124 Dr. hydrogr. Z. 32, 1979. H. 3. Schulz et al.: Period Measuring Proton Magnetometer

rement at zero-coincidence by a pulse from the comparator to the gate FF. The display coun- ter then .indicates the number n of the time base units. One unit corresponds to a time of 2,5 �9 10 -6 s. The beginning and end of the gate time are triggered in the gate I~F by the pre- cession signal. Both this signal and the gate signal are controllable via BNC plugs. The switch "test" enables a self checking of the instrument concerning the counting functions. For this purpose it is possible to start the measurement manually,

D i g i t a l c o m p u t e r The computer of the type I~D-3P was delivered by the firm Tews Elektronik. Input

of the computer is the counting n which - after triggering by the trigger pulse - is received by the preselecting counter. In the real-time mode the computer has to fulfil three tasks : 1. Computation of the induction B from the counting n and the preset N by means of

B = 2~ ]?-1 N n- :

where ] is the frequency of the crystal time base. The magnitude

2~ ]7 -1 - 0,9394972 �9 10 T nT

is stored in the computer as constant value. The display is in the unit nT. 2. Averaging of a sequence of measured values.

Averaging serves for smoothing statistical variations of individual measurements and for damping of effects of aliasing by superimposed pulsations. This holds in particular true of control measurements of the base lines of variometers with low-time resolution, as e.g. photographically recording instruments of the classical type (20 mm h-l).

3. Calculation of the differences (and of their means) used for the base-line measurements of the declination after the method of addition. Multiplication of the means of difference by a stored scale value yields the corrections for the actual adjustment of the coil theodo- ]ire in the appropriate unit. An azimuthal tracing of the coil system for the bias field can thus be done immediately.

l~or the solution of these tasks the computer is capable of storing 72 programme steps. The input-, provisional- and final data may be put out via a printer.

The programme can either finish itself or initiate a renewed start by the computer's remaining sensitive for the arrival of the next trigger pulse.

In the first case a new start may only be initiated via the keyboard. This kind of ope- ration is suited for sampling- and time-dependent random test measurements. Time marks may be indicated by the printer's changing the colour.

I f no measurements are carried out the computer is available for other tasks, especially for evaluating the base-line measurements.

Conclusion The concept of the period measurement is clear and easily realisable by means of a

preset counter. A digital transformation of a measured value into a field-proportional indi- cation can be achieved with any accuracy.

Thus, the possibilities of errors are reduced on principle. Moreover, the possibility of preselecting the number of periods allows a control of the influence of noise on the measu- ring result.

I t must be emphasised tha t the described restrictions of the instrument of the second generation are considerable only in the case of their being operated within the region of an observatory. There is a long-term fluctuation of the deviations which can therefore not be added as constant corrections to the measured values.

For field operation or at sea these restrictions are negligible. Here, too, the simple method of period measurements must be given preference since, in the case of failure, the instru- ments must be repaired on the spot. Due to the advanced development of microprocessors the fixed and compact installation of a digital unit with the described performances is now- adays neither a technical nor a financial problem.

Page 7: A period measuring proton magnetometer with a direct readout

Dr. hydrogr . Z. 32, 1979.14. 3. S c h u l z e t al . : P e r i o d Measur ing :Proton M a g n e t o m e t e r 125

Technical data 1. S e n s o r , boise cancel l ing , V4931 (Var ian)

F lu id : h e x a n I n d u c t a n c e : 16 H

2. C a b l e L e n g t h : 80 m

3. T i m e r , V4931 (Var ian) S a m p l i n g r a t e s : 1/(6 s) a n d 1/(60 s)

4. A m p l i f i e r , V4931 (Var ian) Ampl i f i ca t ion : 7 �9 10 ~ Narrow band filter, passive, variable, V4931 (Yarian)

F Z H

Limiter, V4931 (Varian) Output: 0,4 V maximum

Point of l~ange of Component Frequency 3 dB [ 20 dB overlapping

S--I

2O 15 5

10 8 s -~ s-1 s-1

2,06 _ 18 _ 90 1,92 -[-15 __85 0,~7 _+ 8 + 45

5. Preselecting counter (workshop of the Deutsehes Hydrographisches Institut) input : a. trigger pulse from the timer : 0V to -{- 28 V, time delay continuously adjustable : 30 ms

to 300 msl b. precession signal from the analogous unit: 0,8 Vpp c. preset N: 300, 400, 500, 3000, 4000, 5000 prewired and I, 2 ..... 9999 incremental d. trigger level for precession signal adjustable : 0V to -[- 0,4 V

Output: a. trigger pulse to the computer: TTL, negative b. counted periods of the crystal time base n, 6 digits

code: BCD (8, 4, 2, i), bit-parallel level: TTL, positive

c. LED display, 6 digits d. signal control for the precession signal e. gate signal

Crystal, temperature-stabilized (Philips TCXO) frequency: I0. 106 s -I adjustable: ___ i00 s -I error: -{-2 �9 10 -6 within variations of temperature of-20 ~ to -{-70 0C aging: I" 10 -6.a -I

6. Desk computer, RD-3P (Tews) Input: a. trigger pulse from the prese]ecting counter

b. counted periods of the crystal time base n, 6 digits, code and level see output of the preselecting counter

c. keyboard for operating, data input, and programme input Output : listing, 16 places for numerical and 4 places for alpha-characters, printer double co]ou-

red Arithmetic : 16 decimal places Memory: 3 working registers

i constant register 1 register for counting' 72 programme steps

References V o p p e l , D. , 1972: The p r o t o n v e c t o r magne - W e y a n d , 1978: (Persona l c o m m u n i c a t i o n )

t o m e t e r a t ~Vingst O b s e r v a t o r y , D e u t s c h e s P h y s i k a l i s c h - T e c h n i s c h e B u n d e s a n s t a l t , I-Iydrographisches Institut. Erdmagn. Jahrb. Braunschweig. 17, 133-149. Wiese, 11., 14. Schmidt et al., 1960: Geomag-

Voppel, D., 1977: On the absolute accuracy netische Instrumente und MeI~methoden. of Direct l%eading Proton Magnetometers. Berlin. 382 S. (Geomagnetismus und Aero- Paper presented at the Third General Scien- nomie. Bd. 2. Hrsg. : G. Fanselau.) tific Assembly of the IAGA, Seattle.

Eingegangen im November 1978 Anschrift der Verfasser: Dipl.-Geophys. Giinter Schulz, Deutsehes I-iydrographisches Institut , Erdmagnetisches Observa- torium, 2177 Wingst, Am Olymp 13. I_Vwe Carstens, Deutsches I-Iydrographisches Institut, Bernhard-Nocht-StraBe 78, 2000 Hamburg 4.