Chemical Instrumentation feature
S. Z. LEWIN, N e w York University, Washington Square, N e w York 3, N. Y.
and pulseshaping circuitry. In the ease of amplifier8 for scintillation and proportianal counters, these features are usually indispensable.
T h i s series of articles presents a survey of the basic principles, characteristics, and limitations of those instruments which Jind important applications i n chemical work. The emphasis is on commercially available equipment, and approximate prices are punted to show the order of magnitude of cost of the various types of design and construction.
13. Nuclear Radiation Pulse Amplifiers The amplifiers that have been described in the foregoing are intended for use with detectors that measure the total ioniaation produced by the nuclear radiation source, and their principal requirement is for very high amplification combined with stability, i.e., freedom from excessive drift and noise. When i t is desired to count individual radioactive events, it is necessary to employ an amplifier with properties that are appropriate to the nature and frequency of the detector pulses, as well as to the type of information i t is desired to obtain about the pulses. The first requirement of a pulse smplifier is that it must be capable of accepting and passing signal8 with rapid rise times and short durations. A pulse is fed onto the grid of a. vacuum tube by being spplied to one plate of a capacitor, the other plate of which is connected both to the grid and, thraugh a large resistor, to a grid bias voltage or to ground, as shown in Figure 11. If the pulse is thought of
magnitude. The rate of decay is governed by the product RzC, which is called the time constant of the circuit. If the time constant is extremely 8hort (RC very smdl), d l thc electrons repelled from the righthand plate of the condenser may leak away through R so quickly that no signal ever builds up on the grid of the vacuum tuhe. On the other hand, if RC is too large, the charge released by one pulse may still be on the grid when a second pulse comes in, and the amplifier will be unable to resolve individual pulses. I n pulse amplifiers, the time constant of the input stage, as well as of subsequent, stages, is usually in the range of 0.1 to 10 microseconds, this being a long enough timo constant to permit a ufiablc signal to be transmitted to the grid of the vaouum tube, yet short enough to permit the signal t o decay substantially completely between pulses for most counting rates. (If C = 1 micromicrofarad and R = 1 megohm, RC = 106ohms X 10-I% farads = 10-eseeonds). A pulse amplifier for use with a GeigerMueller detector can be a reletivcly simple instrument, for the detector pulses are large and fairly uniform in size, and their nulse heieht has no soerinl sienificance.
Figure 12. A negative feedback stobitired stage of arnplificdion. The feedback of o portion of the output in opposition to the input tokes ploce through RI, Cl.
The use of negatwe jeedhack to stabilize the gain and improve the linearity of an amplifier is illu~trated by the schematic diagram in Figure 12. If a negativegoing signal is applird to the grid of the pentode shown in the Figure, the tube current will decrease. This causes the potential of the plate to become more positive (since I S R4 = E&is reduced, the plate pdential rises toward the supply voltage, Ex),and a positive-going signal is transmitted through C3 and R3 to the control grid, partially cancelling the primary negative-going signal. Thus, a. portion of the out,put is fed back in opposition to the input, and the net output signal is smaller than i t would be without the feedback. However, the new output is stzbiliaed by this feedback arrangement,
Figure 11. Pulrer are fed into on amplifier through an RC circuit, the time conrtont of which determines the rhope ond height of the pvlre that reacher the grid of h e vocvum tube.
as a burst of electrons approaching the lefehand plate of the capacitor in the figure, i t will be clear that a similar burst of electrons must be repelled from thc righehand plate. These electrons are repelled onto the grid of the vacuum tube, making it more negative, and thus the electron pulse has been transmitted through the capacitor as a negative-going signal on the grid. This charge, however, immediately begins to leak away through R, and the signal on the grid decays in
Figure 13. An example of o feedback loop, in which a frattion of the output of the second stage ig fed back to the grid of the flrrt stage.
Volume 38, Number 5, M a y 1967
/
A305
Chemical instrumentation and can be made substantially independent of drifts and random fluctuatioan occurring in the characteristics of the vacuum tube. When several stages of amplification are used, it is often etrective to employ a feedback loop, such as is shown in simpli&d form in Figure 13. Here, a fraction of the amplified signsl in the second stage is fed back in opposition to the primary signal in the first stage. A negative-going signal impressed on the grid of pentode T-l causes the tube current to decrease; this causes the potential of the plate of T-l to rise, and as a consequence a positive-going pulse is transmitted through the inter-stage coupling condenser to the grid of the second stage pentode, T-8. The current through this tube then increases, and the increased IR-drop in the plate resistor, Rz,causes the plate potential to decrease (become more negative relative t o the supply voltage, E). This negative-going effect st the plate of T-8 is transmitted through the feedback coupling circuit, RrCp, to the cathode of T-1, causing it to feel a negative pulse. But driving the cathode negative is equivalent to making the grid positive relative to its cathode. Hence, the partion of the second stage output signal fed back to the first stage diminishes the original negative-going grid signal that is bang amplified.
Pulse-Shaping Pulse-shaping circuits are employed to modify the characteristics of the input signsl so that they are better suited to the requirements of the amplifier than is the original pulse from the radiation detector. One such circuit is the univibralor, shown in schematic diagram in Figure 14. In this arrangement, one of the vacuum tubes is normally conducting, and the other is cut off. A suitable pulse applied to the outdff tube causes the circuit to "flip aver," i.e., the tube that was off comes on, and the one that was on goes off. This situation lasts only briefly, and then the circuit reverts to its original condition. Thus, a pulse applied to one tube of the pair can be translated into a differentlyshaped pulse taken from a suitable place in the circuit of the other tube. The principle of operation of the univibrator circuit can be understood relatively simply on the basis of the following qualitative considerations. The d u e s of the various resistors and voltages in Figure 14 are chosen so that tube T-1 is conducting, and tube T - 8 is cut off. The grid of tube T-1,although connected through RZ to the +300 v source, is actually a t about ground potential, for it picks up grid current, I,, until I,zR, is approximately equal to 330 v. (If I, = 0.3 milliamperes, I, z R. = 3 X 10-4 amp X lo8ohm = 300 v). The current flawing from cathode to plate in tube T-l under these conditions is about 10 milliamperes, and since this makes the IR-drop in plate resistor RI equal to 200 v (10 ma X 20 Kohms), the plate potential of T-l is about +I00 v. Resistom (Continued on page AS08)
A306
/
Journol of Chemical Education
Chemical Instrumentation Ra and R+ serve, therefore, as a voltngo divider circuit h e t w ~ r nthe volti~gpsof +I00 a t one end and -15U v a t the other end. Thus, the potentist1 of the grid of T-B is more positive than - I50 v by the amount:
That is, the grid of T-B is a t -50 v relative to ground, and this is sufficiently negative to keep the tithe from passing any current. If, now, a positivc-going signal is applied to the grid of T-B in sulfcei~ntmagnitude to muse this tube to begin to conduct, the
following regenerative effect sets in. The current starting up through plate resistor R, generates a n ZR-drop across this r e siutor, causing the potential of the plate of T-B to decrease, relative to the +300 v power supply. This falling potential is transmitted through the coupling eondenser, C2, 8.8 a negative-going signal on the grid of T-I. This causes the current flowing through T-I t,o decrease, which reduces the Zli-drop across RI and causes tho plate potential of T-1 to r i ~ e(become morc positive). This, however, rcquircs the grid of '1'2 to hceome more positive also, ~ i n c ethe latter is connected to the former throngh t,he voltage divider resist,or
R,. Thus, we hnvc seen that thc starting up of current in 7'-B causes t,he current in T-1 to diminish, which causes the grid of 1'-2 to breomp still more positive, which d l m u s c the currcnt in T-B to incrcsse still furthcr. That is, any incrrsse in currant in T-B esusrs this current to increase some more, and the new increment in current lcsds to yet another incrrnont, d e . , in the manner of n selfhouri ~ h i n g explosive rcaetion. This rcgenerativr proccm eontinups unt,il tuhe 1'-2 is conducting hravily, xnd tuhe 7'-1 is cut, off. With T-I cut off, the regeneration ceasos, and the circuit is in a new, tcmpornrily stablo state. However, tube T-1 is cut off hccsusr of the n ~ g a t i v echarge which has been built up on its grid by transmis~ionfrom the plate of T-8 through the capacitor Ca. This charge will leak sway through R2 a t
'7:j 150 K
Figure 14. The univibrotor circuit, in which o positive puke applied to the grid of T-2 turns this tube on and T-1 off, after which the circuit returns to its starting condition.
A308
/
Journal of Chemical Educafion
(Confint~edon page AS10)
Chemical Instrumentation a rate determined hy the time constant Re z Cp (cf. Fig. 11). Soon the grid of T-1 has lust enough of this negative charge that a current can begin to flow in this tube. The starting up of a current in T-l causes the potential a t its plate to decrease, transmitting a negntive-going signal to the grid of T-d, etc., snd setting in motion o. new regenerative effect leading to the shutting off of T-2,and the full conduction of T-1.
signal can he reshaped by passage through the univibrator circuit. Another commonly used pulse-shaping circuit is the delay line, simplified schematic of one type of which is shown in Figure 16. The combination of induetsnec and capacitance in this network has the effect of retarding the rise of the steepest portions of the pulse relatively more than of the less steep leading and trailing edges, so that s, result similar to that shown in Figure 15 is obtained. That is, the top of the pulse may he "clipped" and spread out as a result of passage through the delay line.
30pSEC
Figure IS. A. Pulses coming from the radiation detector may vary in height and shape. B. After reshaping in a univibrator circuit, the pulses hove identical shapes, and o deflnite duration.
Thus, any pulse of sufficient magnitude fed onto the grid of T-d can initiate the flipping aver of the univihratar, but the time during which tube T-1 is cut off, and the delay before the univibrator returns to its starting state are independent of the characteristics of the original pulse. If the output pulse is taken from the plate of T-1, Figure 15 shows how the input
A310 / Journal of Chemkol Education
Figure 16. A simple deloy line used for the reshaping of pulses.
Non-Overloading Amplifiers When a. pulse amplifier is used for linear amplification of the signals from proportional oor scintillation detectors. i t is often
Chemical Instrumentation -
-
-
the case that the radiation source generates both smell and large pulses, and the presence of tho large pulses may prevent the amplifier from faithfully magnifying the small pulses. That is, a large input signal may overbad an amplifier, hy which is meant the fact that the tube currents may be large enough to push the operation of the tuhe into the non-linear portion of the tuhe characteristic curve, and to alterthe effective plate and grid voltages from their desired range of values. Because of the finite time constants of the various parts of the circuitry, an amplifier does not recover instantaneously from such overloading. However, by careful design of the circuitry, and by careful selection of the amplifier's vacuum tubes, it is possible to minimize these effects. A nmoverbading amplifier is one which has been so designed. The criterion of the performance of such an amplifier is the time it takes for the output base line to recover its initial value after an overload of several fold aver the maximum signal for which the circuit is intended; i.e., several times the signal which gives full-scale deflection of the output stage read-out.
Coincidence Amplifiers
A type of amplifier that is often employed when it is desired to eliminate the interference of high energy radiations from the measurement of a low energy component, or to reduce the background count when measuring a low level source, i. the anti-coincidence amplifier. The operation of this type of circuit may be most easily understood by considering first the eoineidenee eimuit, shown in simplified form in Figure 17. In this ease, consider that the circuit components and applied voltages are chosen 60 that both tubes are normally conducting. If a negativegoing signal is placed an the grid of tube 1, the current through that tube d l decrease or be cut off, depending
Figure 17. A coincidence circuit, in which the IRdrop ocrotr the common cathode resistor doer not show o large change unless pulses el and el are simultmeous.
upon the magnitude of the signal. This would tend to decrease the current Bowing through the common cathode resistor, R. However, if the IRdrop across the r e &tar decreases, the cathodes of both tubes become more negative (closer to ground, in the diagram as shown), which (Continued on page ASlB) Volume
38, Number 5, May 1961
/
A31 1
Chemical Instrumentation is equivalent to the grids becoming relatively more positive with respoct to their cathodes. If the input signal is large enough t o keep tube 1 cut off, this positive-going effect on the grid of tube 1 has no effect, hut the positive effect on
.
IR-drop. Thus, if the output is taken as ,he ,,a,, dm, .c,s, t h e resistor, large signs1 is obt,ained only if hoth grids are pulsed in coincidence. Two radiation detectors are used with this type of amplifier; s high energy nucleon traveling along a line defined by the locations of the two detectors will trigger essentially simultaneous discharges in hoth, and a cainci-
. Figure 18. A discriminator circuit based on o rectifier diode. Pvlrer cause current to Row through the rectifier only if they exceed in voltoge the reverse voltage bias applied ocrorr the rectifier by the battery and rlidewire.
the grid of tube 8 causes the current through tube 8 to increase. Thus, the decrease in current in tube I causes sn increase in current in tube 8, and the net current through the cathode resi~tarR does not change much. However, if the grids of both tubes are fed negative-going signals simullaneously, the cathode resistor R must show a large decrease in
A312
/
Journal of Chemical Education
dence will be rerorded hy the amplifier. I n this way, directional information or selectivity can be schicved. The anti-coincidence circuit is very similar to tho coincidence cirouit described ahove, with the exception that one of the tubes is biased so that it is oanduoting, while the other is biased off. The de(Continned on page A S 1 4
Chemical Instrumentation teetor that feeds pulses to the tube that is normally on is arranged to feed its grid negative pulses, while the other detector feeds positive pulses to the grid of the tube that is normally off. If either one or the other detector sends a pulee to the circuit, the two tubes will become either both an, or both off, and the cathode resistor eurrent shows a marked change in magnitude. However, if both detectors send in pulses simultaneously, the current in the common cathode resistor is not affected, since under these conditions one tube is on and the other is off (viz.,the circuit is in the mirror image of its resting condition).
Discrimination
It is often important to amplify only one type of pulse in the presence of another, or to pick out a pulse of given size range from accompanying random noise pulses, and the circuitry employed for this purpose is called a dismiminatm, or pulse height seleetw. One type of circuit used for this purpose is shown in Figure 18. The heart of this circuit is the IN34 diode, which psases current anly when the vol& age difference across it is in a certain direction (e.g., positive to the left and negative to the right of the rectifier symbol in the figure). By means of a battery, E, and a slidewire, a voltage is applied across the diode in the sense opposite to that which produces conduction. If s
A314
/
Journal o f Chemical Educofion
-
SIGNAL
OUT
INPUT
Figure 19. The Schmiw trigger circuit. When the voltage applied to the grid of T - l reashes o critical value, the circuit triggers rhorply, causing the tube that was conducting to go off, and visa vewo.
pulse is fed into this circuit, it will result in B flow of current through the diode anly if its voltage exceeds the biassing voltage. Thus, the circuit can beadjusted to pass only those pulses which are bigger than any predetermined size. Another type of discriminator circuit that is widely used in nuclear detection amplifiers is the Schmitt trigger, shown in Figure 19. In this circuit, one of the tubes is normally conducting and the other one is cut off. When the voltage applied to the grid of the cut-off tube is made more positive, a critical value is reached at which a regenerative effect
sets in (see discussion above of univibrator circuit) that turns this tube on, and turns the other tube off. This triggering effect can be set to occur for any desired voltage level, so this circuit can be employed to announce when a signal has reached a given value, and, in fact, can tell whether the signal has approached that value from above or from below. The action of the Schmitt trigger can be understood qualitatively as follows. Consider that tube T-I in Figure 19 is not conducting because its grid is negative with respect to its cathode. If this grid (Continued on page A316)
Chemical instrumentation
to permit some tube current to flow. When this current starts in T-1, regeneration sets in, for the current causes the potential of the plate of T-1 to decrease, which makes the grid potential of T-8 more negative, and causes the current in T-8 to decrease. The net effect of the increase in ourrent in T-l and the decrease in T-8 is a small decrease in the total current through the common cathode resistor, RE. This makes the cathodes more negative than they were, and is equivalent to the grid of T-l becoming more positive. This further increases the current in T-1, further reduces this tube's plate potential, causes the grid of T-8 to become still more negative, etc., and the regeneration rapidly leads to the cub ting off of T-2, and the full conduction of T-1. After this triggering has occurred, making the grid of T - l more positive by means of the input signal merely causes this tube to conduct more heavily, but does not change the state of the circuit. If the input signal is reduced in magnitude, the regeneration sets in in reverse when the critical level of potential is reached. A single channel differential analyzer contains two discriminator circuits, so arranged that one of the discriminators acts ss a lower gate or window to pass d l pulses larger than a preset value, and the other disrriminator sets as an upper gate to pass only those pulses having smaller amplitudes than its preset value. Thus, the two disoriminators provide a selected pulse height interval, and only pulses whose heights lie within this interval are accepted for amplification. By varying the locations of the upper and lower windows, the entire distribution of pulse heights produced by a radiation source can be determined. This is the basis of the relatively new and rapidly developing field of pulse height spect~mn'etry. A variety of pulse height analyzers employing single channel and multi-channel selectors is now commercidly available.
Cathode Follower I n the coupling of a n amplifier to a radiation detector on one side, and to a resd-out device on the other, it is often essential to have a n impedance matching stage between these units. If the input impedance of a device is low, the device will draw appreciable current from any source t o which it is connected. If the output impedance of a source is high, i t may he incapable of delivering rtppreciable currents. A circuit that can he used as a coupling stage to permit the output of a high impedance source to control a low impedance circuit is the cathode follower, shown in schematic diagram in Figure 20. The signal from the high impedance source is applied to the grid of the vacuum tube through the high impedance eepaeitor-grid resistor circuit, and negligible current is drawn from the source. The signal modulates the tube current, and the changes in the
A316 / Journal o f Chemical Educaiion
Chemical Instrumentation latter can be ohserved ae variations in the IR-drop across t h e cathode resistor, Rx. Since the tube current is relatively large, and RK is a small to moderate resistance, the ZR-drop across R K can be used to drive low input impedance devices. The circuit derives its name from the fact
F0 Figure 20. The cathode follower circuit for impedance matching. The input to this circuit is fed in through a high impedance; the output is taken across the low impedance, RK.
that the variations in the I R d r o p across the cathode resistor precisely follow the variations in the input signal. Scalers The output from a radiation detectorpulse amplifier combination oonsists of pulses that are randomly spaoed in time, and occur with a frequency determined by the strength of the radioactive source. I n order to count the individual pulses, and register their number, i t is necessary to dense a high-speed counting instrument capable of responding to impulses that are only microseconds apart. The function of a sealer is to provide an electronic means of scaling down the frequency of pulsing of the output of the amplifier without losing any information, so that the scaled-down rate of pulsing can be presented in a form suitable for reed-out. One of the most widely used of scaling circuits is the scale-of-two, binmy, or flip-flop circuit, shown in Figure 21. This circuit has two stable states, vie., the two mirror images in which one of the pair of tubes is conducting, and the other is cut off. Since the plate of each tube is connected through a n impedance to the grid of the other tube, the current that Hows in the tube that is conducting produces a n IR-drop in its plate circuit that drives the grid of the other tube negative enough to cut i t off. The trick in the scale-oftwo circuit is to introduce the pulses coming from the amplifier in such a way as to cause the circuit to Hip over from one of its states to the other one. The detailed operation of this circuit can be understood by reference t o Figure 21. Assume that tube T-S is conducting, and T-4 is cut off. The twin diode, T-l and T-3,has its two plates connected to the plates of the two scaling: tubes, (Contimed on page A.719)
Volume 38, Number 5, May 1961
/
A317
Chemical h m m e n t a t i o n to that of the piates, aid no current
flows through T-1 and T-2. The pulses to be scaled down are applied to the twin diode cathodes, driving them negative (Continued on page AS22)
Figure 21. The scale-of-two circuit, for yielding on output having exactly one-half the frequency of the input. T-1 and T-2 are rectifier tubes through which the input pulses are fed to the two trigger tuber, 7-3 and T-4.
Volume 38, Number 5, May 1961
/
A319
going pulse on this grid causes the tube current to begin to decrease. This raises the potential at the plate of T-S, whieh in turn raises the potential of the grid of T-4, which ca:lses current to start up in T-4, lowering the potential of the plate of T-4, whieh causes the grid of T-3 to become more negative, decreasing still further the current in T-3, etc., etc. The regenerative process rapidly leads to the
Chemical Instrumentation
I
tubes. These currents flow through the plate resistors RI and R8 of both scaling tubes, causing the plates of T-S and T-4 to become more negative in potential than they were. Since these plates are coupled t,o the grids of the other member of the pair, the ~ R e c of t the original pulse
Bt
I 0
Q2
INPUT
?
4 -
dd
OUTPUT
T
n
-
i =
Figure 22. A transistorized scale-of-two circuit. D - l and D-2 correspond t o t h e rectifier tubes in Figure 21; 0-1 and 0-2 correspond t o the trigger tubes.
is to produoe negative-going potentials on the grids of both scaling tubes. Since the grid of T-4 is already cutting this tube off, an additional negative-going pulse on this grid has no effect. However, T-3 was conducting, and a negative
A322
/
Journal of Chemical Education
cutting off of T-3 and the full conduction of T-4. Thus, the introduction of a pulse has flipped the circuit over from (T-S on, T-4, o f ) to ( T 3 off, T-4 on). The neon bulb, N, is in parallel with (Continued on page AS841
Chemical Instrumentation the plate resistor, R,, of T-3. Whm T-R is conducting, the IR-drop through 732 corresponds to a, sufficient volt,age to light, the hulb. When T-3 goes off, the light is extinguished. Hence, the light goes on for every other pulse fed to the circuit; e.g., if four pulses have been fed in, the hulb has been lit twice, and extinguished twice. That is, the frequency of lighting of the hulb is exactly one-half of the f r o quency of the pulses; if the hulb (or, equivalently, the IR-drop Bcrcss one of the plate resistors) is taken as the output signal, the frequency of the input ha8 been scaled down by a faotor of two. The output from this soalwf-two eircnit can now serve as the input to a
FLIP-FLOP 1
FLIP-FLOP 2
FLIP-FLOP 3
FLIP-FLOP 4
Figure 23. Schematic representation of the principle that permits four rcale-of-two stages to be corcoded and modified with feedback loops to yield o scole-of-ten counting orrangernent.
~econd seslaof-two circuit; i.e., the circuits can be cascaded. If two binary circuits are csaeaded, the output a t t h e last stage has been scaled down by a factor of 4; if three, hy 8; if n, by 2".
Binary scalers are currently available with final-stage scaling factors of 1034, and higher. Figure 22 shows a scale-&two circuit hased upon the principles outlined above, but in which NPN-type transistors are employed in place of vacuum tubes. I n this circuit, the collector corresponds to the plate of tho triode, the emitter to the cathode, and the base to the grid. The binary system of recording counts is simple electronically, hut creates some arithmetic difficulty in making computations according to our derimal (i.e., scale-of-ten) system. Therefore, considerable ingenuity has gone into the design of circuits that permit the fundamental scale-of-two to be cascaded in such a way as to yirld a scale-oi-ten counter. The principle of this method of achieving a decimal scaler is shown schemsticelly in Figure 23. This eonsists of 4 scsleof-two circuits, such 8 s would be found in a scale-of-sixteen scaler. However, when tho tenth pulse is fed in, V 2comos on and a signal is sent over path A to cut otY V7 and turn V Son. 4lthough Vz was also turned on a t the second, fourth, sixth, and pighth pulses, VTwas off a t thosr times, and the signal in path A had no effect. When V2 comes on, it. normslly turns Va on, hut on this tenth pulse, the turning an of V8 sends a pulse back over path B that prevents Vs from being affected by V 2 . I n this WILY, there result ten different combinations of the four bistable flip-flop states produced - successively by ten pulses. Each combination can be assigned a. digit, and s soale-af-ten decade tinit is the result. An example of such a decade scaler eircuit is given in Figure 24. A scaling decade can be achieved in a. much simoler fanhion by the use of a beam switching, or glow transfer tuhe. This is rt tuhe containing a filling of neon or other noble gasas,a permanent magnet surrounding the tube, and three sets of ten each of eleotrade elements plus a single cathode. The electrodes consist of a grid, a plate, and a spade, respectively, and ten groups of these three electrodes are spared around the central cathode as shown in the Figure. Consider that a discharge exists in the tube so that an electron beam flows from the central cathode to one of the plates (No. 0 in the figure). When a pulse is applied to this tube, it is arranged to make the ten grid electrodes ( d l of which are connected to a oomrnon point) become more positive. The glow discharge will then move over toward that grid and then (Continued on page ASB7)
A324
/ Journd o f Chemicol Educofion
Chemical lnsfrumenfafion climb along the beem-forming spade electrode as the pulse dies away. The discharge "locks in" on the plate (No. 1, in the figure) toward which the shape of t h e spade has directed it. The glow will then remain here until the adjacent grid is made positive, by virtue of a new pulse coming in, as a consequence of which the glow will move aver t o plate Na. 2. Each pulse, applied simultaneously t o all grids, advances t h e glow by only one position a t a time, and always in the same direction (clockwise, in the figure). Thus, the location of the glow discharge
Figure 24.
a t any instant shows how many pulses have been fed iuto the scaling unit, in units of ten. These beem switching tubes can also be cascaded, so that the second stage tube gets its pulses from the No. 10 (i.e., 0) position in the first stage tube. Rate Meters
The actual counting of total mimbers of discrete pulses ovrr imcisely defined time intervals can be avoided through the m e of x circuit which totalizes the pulses in a differential way, so that the eondition of t h e circuit is scnsitivo to the rate of pulsing. This is called counting-rate (Conlinz~edon page A8f8)
Circuit diagram of o decimml scaler based on the principle illustrated in Figure 23.
Volume
38, Number 5, May 1961
/
A327
Chemical Instrumentation
5
Figure 25. Schemotic diogrom of the electrode configuration in 0 beem witching tube. The grid electrode%are the small circler located between each plote and spade. A glow dixhorge occurs between the cen!rd cathode, and one of the plate elec!roder.
I-0
POWER
SUPPLY
Figure 26. A portion of the circuit diogrom of the counting-rate meter employed in the Anton ElecIronic Labs' Model 6 Rodiologicol Survey Meter.
circuitry, and iuvolves two principal fatures. Provision must he made, if it is not nlrondy included in the amplifier t h a t prcccdes thi8 circuit,ry, t o shape the pulses so that every pulse is equal in hcight and dnmtion. This is commonly sceompli~hedby the use of n ~mivihmtorcircuit (see above for detailed f.xplsnation of this circuit,). The counting-rate circuit is then arranged so t h a t ciirh pulse puts a definite amount oi chnrgc on a enpneitar. Simultaneously, some oi tlrr charge on the espscitor 1r:aks off through n high resistance. The ste:~dy state rhxrge on (02 voltage xeros~ithe eitpnritor is thus due t o the dymmic hnlancc hr,twen the rate a t wtiich pulses are frd in, and the rate of leakage of thc stored charge. A vokmeter oonnectcd across t h r cnpacitor can he calibrated to read rlirrctly in counts per minute. An example of a iwr~n-ratemetrr usod in a eommcrcial survey i n ~ t r u m e n t(Anton Electronic Labs hlodrl 6 Radiological Bmwy Mcter, $139.30) is given in Figure 25. The two trsnsistars, T'-I and V-Z, constitute a. univihrator; pulses from the G-M tube cnum V-b to go off and V-1 t,o come on. V-1 t-tays on for a time determined hy the product of R-10 times - t h e capncitancc chofien by the selector swit,ch. This rcsults in a voltage pulse a t the collector of V-1; this pulse passes current through thp silicon rectifier, CR-1, and adds charge to the cspscitor, C-5. The motcr, M , is in parallel with the storage condcnscr, sa that some of the charge leaks off through it. The deHection of the metor is thercfore proportional t o the rate of p ~ ~ l s i nofg tho G-M tube. A counting rate mcter in very ust~iulif rhanging lrvcls of radioactivit,~are to he determined, as in the surveying of a n area, or the monitoring of a flowing stream. I,h e counting ratc circuit adapts very simply t o tho driving of a continoous recorder, so that a continuous and permanent record of t h c radiation lcvel is obtained. However, in relating the output of a count,ing rate meter t o radioactivity level, i t is important t o take into nrcaunt the errors t h a t may result from the relatively largc time constant necessary in the tank rircuit (storage capacitor plus its shunting rt.sistanee).
.
(Co,dinr~edon paye .l.7.30)
A328
/
Journal of Chemicol Education
Chemical ln~trIIment~ti0n Since tho storage condenser is simultaneously charging and discharging, the time required for its equilibrium stateto be achieved depends upon the rates of both these processes. The former depends upon the radioactivity level ( N counts per minute) being memured, whereas the latter is constant for a given setting of the selector switches of the instrument (viz.. , , given values of R and C). Hence, the equilibrium lime of a counting rate meter, defined as the time necessary for the deflection to change from zero to tho average value of the oaunting rate, is a function of that counting r t e , in a manner that is given by the following equation. , t = RC(1.15 log,, 2NRC 0.394) This time can be of the order of several minutes for counting rates of less than 1000 counts per minute. When switching from the measurement of one radiation source to another, i t is not neeasmy t o wait the full time specified by this equation, since the storage condenser is not starting from a state of eero charge. Indeed, the smaller the change is in intensity level, thesbortermay be the waitingperiod before reading the meter. The probable error of an instantaneous reading of a counting rate mcter is a fonotion of the counting rate and of the time constant of the t,nnk circuit, as shown by the equst,ion: ~~~
+
P.E.
A330_'/ Journal o f Chemical Educafion
=
0.6745
----.
V'2mc
This error can be reduced by averaging the ,,,ding over a period of time.
~ i b l i ~ ~ ~ ~ ~ h ~ BOSQUET,A. G.,"Counting Rate Meters Versus Sealers," Nucleonics, 4, 67 (1949). ELMORE,W. C., snd Sands, M., "Electronics; Experimental Techniques," NaLimrill Nuelertr Energy Series, Division V, Vol. 1, McCraw-Hill Book Co., N. Y., 1949. FLUEGGE,S., A N D CREUTZ,E., editom, Encyclopedia of Physics, vol. XLV, 11." "Nuclear Instrumentation Springer-Verlag, Berlin, 1958. HINE, C. J AND BROWNELL,G. L., editors, "Radiation Dosimetry," Academic Press, N. Y., 1956, Chaps. 4-10. KORFF, S. A., "Electron and Nuclear Counters," D. Van Nostrand Co., N. Y . ,2nd Ed., 1955. LION,K. S., "Instrumentation in Scientific Rosesreh. Eleotrical Inout Transducers," ~c~raw-m ill ~ o o Co., k N. Y., 1959, Chap. 5 . RENNE,H. S., "Atomic Radiation Detection and Measurement," Howard W. Sama & Co., Indisnspolis 5, Indiana, R o s s ~ ,B., AND STAUB, H., "Ionization Chambers and c ~ , , ~ McCraa~~~~ Hill Book Co,, N, Y., 1g49, WILK,NSON, H,, ~ . ~Cha,mbers ~ ~ and Counters,n cambridge University Press, 1950.
,,
Nezl: Conclusion of the sumey of nuclear radiation electronic gem.
," i
~
