Full Text
99-Channel Amplitude Pulse Analyzer
Recently several reports have appeared on ingenious designs of differential pulse analyzers. The article under review¹ contains a description of a stable analyzer with a large number of channels, distinguished, nevertheless, by moderate dimensions and low power consumption in comparison with ordinary 20–30-channel systems.
The principle of operation of the instrument consists in converting the input pulse into a series of standard pulses, the number of which is proportional to the height of the input pulse. Registration is carried out by two counting decade rings; their established position (after the passage of the standard pulses) is recorded by the corresponding electromechanical counter.
Fig. 1.
The conversion of the input pulse into a certain number of standard pulses is carried out in two steps. First, the input pulse is “tipped onto its side,” i.e., converted into a rectangular pulse whose duration is proportional to the magnitude of the input. This pulse “illuminates” a locked generator with period $\tau$, which then produces a number of standard pulses proportional to the height of the input.
Since it is very easy to make a generator whose stability is much better than 1% over a very long time, the stability of the operation of the entire device is determined by the correctness of the first part of the conversion.
Let us consider in greater detail the operation of the electronic circuit (Fig. 1) that performs this conversion. The input is a cathode follower. Capacitor $C_1$ is charged to the peak value of the positive input pulse through series-connected diodes and the output impedance of the cathode follower, which in total amounts to about 1 m$\Omega$. Thus, in order for capacitor $C_1$ to have time to charge to the peak value of the input pulse, the rise time of the latter must be large in comparison with $RC$, i.e. $1000 C_1$. When the input pulse begins to fall, the first diode is cut off,
FROM CURRENT LITERATURE
and the discharge of the capacitor \(C_1\) through \(R_1\) begins, the rate of change of the pentode anode potential being equal to the supply voltage divided by \(R_1C_1\), in volts per second. While this fall of the anode potential is taking place, the voltage on the pentode’s screen grid also falls to a low value and, consequently, a more or less sharp positive voltage pulse appears on the screen. This pulse is shaped by the two double triodes following the pentode and is then fed to an inverter tube, which makes it possible to take rectangular pulses of either polarity. The values of \(R_1\) and \(C_1\) are determined by the following considerations. In order that an insignificant curvature of the diode characteristic should be less important, the pulses must be as large as possible. The maximum value of the input pulse was chosen as 150 V. With a supply voltage of 300 V this means that the maximum duration of the pulse is determined from the condition that
\[ \frac{150}{T}=\frac{300}{R_1C_1}, \]
or
\[ T=\frac{1}{2}R_1C_1. \]
The maximum duration of this pulse (i.e., the maximum height of the input pulse) must correspond to
Fig. 2.
100 standard pulses (according to the number of channels), i.e. \(T=100\tau\), where \(\tau\) is the period of the standard-pulse generator. In order that the rise time of the input pulse should not affect the calibration, it must be less than one period of the generator. \(50\ \mu\text{sec}\) is a reasonable upper limit for the pulse rise time required for accurate analysis from proportional counters and fast ionization chambers. Thus, \(\tau=50\ \mu\text{sec}\), and
\[ 100\tau=\frac{1}{2}R_1C_1, \]
or \(R_1C_1=10\ \text{msec}\). But \(1000\,C_1\) must be shorter than the smallest
for the pulse rise time, i.e., it is advantageous to take \(C_1\) as small as possible. However, an excessive increase of \(R_1\) leads to instability in operation.
A compromise is provided by the values \(R_1 = 50\) megohms and \(C_1 = 150\) micromicrofarads, giving \(R_1 C_1 = 7.5\) msec, which permits an amplitude change for a rise time of 1 µsec. The operating accuracy of the entire instrument is determined by the stability of these components, as well as by the stability of the standard-pulse generator and the stability of the power supply (0.5%), which presents no difficulty. But if the long-term stability of the power supply for some reason gives rise to suspicion, one may use the stabilizer in the proportional counter circuit^2.
The second part of the “digitizing” process is carried out by applying a time pulse to the circuit blocking the passage of standard pulses to the counting circuit.
Having thus considered the conversion process, we may proceed to the operation of the instrument as a whole (see the block diagram, Fig. 2).
A pulse from the ionization chamber or the proportional counter, through an external linear amplifier and the blocking circuit of the instrument, is fed to the input of the already described “inverting” device. The converted pulse of negative polarity causes the blocking circuit to pass a certain number of standard pulses to two consecutively connected ring counting decades.^3 Each of the ten elements of each ring is connected, through a current-amplifier tube, with a simple relay, which closes if the ring remains in such a position that this element has operated, i.e., the operation of the relay is analogous to the operation of signal neon lamps in ordinary counting rings. The elements of the rings are numbered 0–9; thus, if 27 signals are applied to the input, the relay of element 7 of the units ring and relay 2 of the tens ring will close. The internal contact of each relay of the units ring is connected through an ordinary telephone-conversation number register to the internal contact of each of the 10 relays of the tens ring (Fig. 3), not merely to connection 0—0. The external contacts of all relays of the units ring are grounded, while the external contacts of all relays of the tens ring are connected together and brought, through a high-speed relay (its purpose will be explained below), to a lamp providing a current pulse sufficient to operate the register.
The counting system and the two operated relays remain for the time being in this position.
As already stated, the maximum rise time for the input signal is chosen as 50 µsec. In order that pulses following the accepted one should not distort its digitization into a time pulse, 50 µsec after its acceptance the blocking circuit operates, cutting off all subsequent pulses until the entire operating cycle of the instrument is completed, which is determined by the operating time of the telephone registers (it is about 0.1 sec). The time for which the blocking circuit is engaged is constant and does not depend on the magnitude of the input pulse.
The same pulse that controls the operation of the blocking circuit is fed, through a delay sufficient for operation of the relays of the counting system, to the tube supplying the current pulse for switching on the register by the counting system. After an additional delay (during this time the register has time to operate), a current pulse is produced which passes through a relay disconnecting from ground the leakage grids of the tubes of the counting system, thereby resetting the decades.
If the input pulse exceeds 150 volts and, consequently, is converted into a number of standard pulses greater than 99, then the tens ring will make a full revolution and give a signal to the trigger lamp, which controls the operation of the mentioned selector relay. Thus, if the input pulse produces, say, 113 standard pulses, then the counting rings will be set to position 13, but this recorder will not operate, since the current pulse required for this will go to the residual-pulse recorder. Resetting of this trigger is carried out by the same relay that resets the decade.
A careful examination of Fig. 3 shows that, for the current pulse which causes the recorder to operate, besides the direct path through the recorder 27 switched on by the decades, in the case shown there also exists
Fig. 3.
many other paths through three consecutively connected recorders. However, the fact that in one of them the current is necessarily directed in the direction opposite to the normal one makes it possible to eliminate this harmful side effect by connecting, in series with each recorder, a diode allowing current to flow only in the proper direction.
After all the processes are completed, the blocking returns to its initial state and permits reception of the next pulse.
Summing up the operation of the instrument, the following may be said. The abundance of channels, the very great stability of operation (according to the tests carried out, the error was less than 1% over a period of several months of continuous operation), good linearity (noticeable deviation only in the initial channels), small dimensions, and the insignificance of the required energy very favorably distinguish this instrument from the complex and extremely bulky designs of ordinary differential
pulse analyzers using a series of simple discriminators with successively increasing bias, connected to adjacent anticoincidence cells. However, the field of application of such a device is limited by the impossibility of correct operation when the rise time of the leading edge of the pulses under investigation is less than 1 μsec. It is to be hoped that ways will be found to improve the operation of the input part of the interrogating circuit.
N. Lazarev
References
- D. H. Wilkinson, Proc. Phil. Soc. 46, No. 3 (1950).
- D. H. Wilkinson, Journ. Scient. Instr. 27, 36 (1950).
- D. H. Wilkinson, Electronics, March, 1948, p. 122.