Transistors are finding new and ever-increasing uses. Some of the circuits in which transistors are being applied are well-known through prior employment in conjunction with vacuum tubes. The applications shown in the chapters which follow illustrate some of the more unique ways in which to take advantage of the desirable properties of transistors. The circuits that are described are practical, have been built, tested, and are now in use.
Oscillators and Triggers
THOUGH transistors generally demand their own circuitry, there are some good vacuum-tube circuits that function nicely when transistors are substituted. As the phase relations of a grounded-emitter transistor resemble those of a grounded-cathode vacuum tube, it is possible to use junction transistors for vacuum tubes in most oscillator circuits.
The multivibrator and the two-terminal sine-wave generator perform well with junction transistors, with the transistorized two-terminal L-C having a great advantage over other transistor oscillators. As with vacuum tubes, the circuit oscillates on higher frequencies than single-transistor oscillators with the same components. While transistors must be selected for other oscillators, most transistors, even those not oscillating in single oscillators, work well on rf in this circuit.
The multivibrator is in principle a two-stage R-C coupled amplifier (Fig. 501). The output voltage of the second tube is fed back to the grid of the first tube. Because of this feedback the circuit starts to oscillate at a frequency determined by the time constants of R and C. The circuit will also oscillate if you substitute an element with similar amplifying and phase-shift relations in place of the vacuum tubes.
Before getting into further circuitry, let's discuss the principle of analogy. Fig. 502 shows a junction transistor and a vacuum-tube triode in common-emitter and common-cathode connections. The symbols e, c and b represent the emitter, collector and base, respectively.
Fig. 501. Basic multivibrator circuit. Tile operating frequency is determined by the values of R and C.
We can immediately draw an analogy between the terminals of the two devices: The emitter corresponds to the cathode, the collector to the plate, the base to the grid. The analogy may be extended by noting that an ac signal undergoes a phase reversal between input and output electrodes—base and collector for the transistor, grid and plate for the tube.
Fig. 502. These diagrams show the basic analogous vacuum-tube and transistor circuits.
If you make the grid of the vacuum tube more positive, plate current increases and as the plate current increases the plate becomes less positive due to the voltage drop across the load resistor. Thus, all "mountains" of a sine wave applied to the grid are trans¬formed into "valleys" of the anode voltage. Now look at the grounded-emitter circuit (Fig. 503) of a p-n-p junction transistor. If the base is made more negative with respect to the emitter, col¬lector current increases and the voltage drop across the load resistor causes the collector to become more positive. So, a sine wave is "turned over" the same as with a vacuum tube, a grounded-emitter transistor producing the same 180° phase shift as a grounded cathode vacuum tube. Fig 503 shows how the grounded-emitter circuit of a junction transistor corresponds to the grounded-cathode circuit of the vacuum tube.
There are, of course, certain differences between the two. Whereas the grid is returned to a negative source and Ig is zero for normal operation, the base return is positive (in transistors with a positive collector) and I„ is not zero. This means that the input resistance of the transistor is not infinite, as is the tube's, but has a finite, actually low, value.
Fig. 503. Tube-transistor comparison. Note phase inversion of the waveform in both instances.
The magnitudes of the output resistances are also dissimilar, the transistors' being much higher than that of the tube. However, as alpha, the ratio of a change in Ic to a change in Ie, increases to unity, the differences decrease and the analogy becomes closer. The input resistance of the transistor rises and its output resistance drops correspondingly.
Fig. 504. Transistorized multivibrator. The output has a rich harmonic content.
An important difference that does not change with alpha is the input voltage at which conduction begins. For the tube, the grid cathode voltage must be more positive than the cutoff level. The transistor, however, conducts only for a base-emitter voltage equal to, or greater than, zero; it is cut off for all negative values.
Fig. 505. Multivibrator output pattern showing complex waveform.
Thus, to transistorize a vacuum-tube circuit by analogy, we must make use of the similarities of the two devices and use transistors with alphas close to unity to minimize the differences. The operating principles of the resulting circuits are the same as for their tube counterparts.
Fig. 504 shows the circuit of the transistorized multivibrator. It produces an output waveform rich in harmonics, though they are not as rectangular as those produced by tubes. Nevertheless, there are harmonics up to 30 mc when the circuit oscillates at about 100 cycles. An oscilloscope pattern of the output waveform is shown in Fig. 505. The waveform is more complex than with tubes because the transistor requires input power.
Fig. 506. Diagram of sine-wave generator. This is a two-terminal circuit with the frequency of oscillation determined by the values of L and C.
There is a sine-wave oscillator circuit which resembles the multivibrator: the two-terminal circuit of Fig. 506. In principle it is a multivibrator with a resonant L-C circuit between the grids. In the old days of radio the two-terminal oscillator (then called a balance generator) was used because of two advantages: there is no need for a tapped coil or feedback winding, any resonant circuit connected between the grids will oscillate; the balance generator worked up to very high frequencies even with the poor tubes of those days.
Fig. 507. Transistor sine-wave genera¬tor. Note the similarity to Fig. 506.
Some say this was the first circuit which ever produced CW in the 2-meter band. Today we have similar difficulties—only a relatively * few individual transistors oscillate in the radio-frequency band. It would seem that the balance generator might help to reach higher frequencies even with components not suitable for oscillating in ordinary circuits. It has been found that it is so. The transistorized balance generator (Fig. 507) oscillates over the entire medium-wave band with individual transistors which are otherwise suitable only for audio purposes. If you omit the resonant circuit, you have a multivibrator again. Therefore, it is possible to combine both circuits with the oscillator being used either as a multivibrator or sine-wave oscillator.
Fig. 508. Schematic of the combined multivibrator and sine-wave generator. The multivibrator circuit becomes a two-terminal sine-wave generator when the parallel L-C circuit is placed across points A-B.
The combined circuit
Fig. 508 is the circuit of the combined oscillator. Two CBS-Hy-tron 2N36 transistors work with grounded emitters. The bias is obtained by 330,000-ohm resistors between the bases and negative supply. In each collector lead there is a 3,900-ohm load resistor. The two .05-iif coupling capacitors connect the output of each transistor with the base input of the other.
Without any L-C circuit connected to terminals A and B, the circuit works as .a multivibrator. The waveform between A and B with the coil-capacitor combination in place is shown in Fig. 509.
Fig. 509. Waveform between terminals A-li of Fig. 508.
The circuit oscillates at about 100 cycles. If you connect terminals A and B to the input of a radio receiver, you will hear the 100-cycle signal over all bands up to 10 meters. Thus the circuit makes an excellent multivibrator for test and alignment purposes. With only a penlight cell used as a voltage source, the unit may be built ultra-compact.
When you connect a resonant circuit to terminals A and B, the oscillator produces sine waves of a frequency determined by the L-C circuit. For low-frequency purposes any capacitor and choke arrangement may be used. The sine wave produced in that case may be seen on a scope. If you want to hear the audio frequency, use your headphones as the inductor. If the capacitor is omitted, higher frequencies will be generated. If the L-C ratio is too high, the sine waves are distorted as shown in Fig. 510.
To use this circuit as an if alignment generator, connect a transformer to A and B. For covering the medium-wave band, any wave-trap or crystal set may be used. The battery voltage may be as little as 1.5, though higher voltages of about 4.5 are desirable to obtain higher output on high frequencies. With a battery voltage of about 4.5, a current of 1 ma will be drawn by the circuit when oscillating, which increases to 2 ma when oscillation ceases. The sine-wave output between A and B is about 2 volts rms at 100 cycles, decreasing to about 0.5 volt at 1.5 megacycles.
Fig. 510. Distorted sine-wave pattern produced when the ratio of L to C is too high.
The unit is laid out so that it can be easily soldered into any hookup. The layout may be seen in the photo, Fig. 511. A 2 x 3-inch polystyrene sheet serves as a "chassis." Only four soldering connections have to be made if the oscillator is connected into any larger hookup. When assembling the unit be careful not to apply excessive heat to the transistors when soldering.
As it is much easier to get transistors oscillating with this circuit, it should be of interest to everybody interested in experimenting with transistor oscillators at high frequencies.
Trigger circuits are an indispensable part of modern electronics. We find them not only in such complex equipment as radar sets and digital computers, but also in the "ordinary" devices that the radio and television technician is and will be called upon to service —black-and-white and color television, the multiplex radio sets of tomorrow and many others which are as yet still in the dreams of design engineers.
The transistor is having a profound influence on such developments and thus transistor trigger circuits are of prime interest. Investigations have almost exclusively aimed at the point-contact type in the past, because of its inherent negative resistance. Progress was less encouraging than anticipated and research is now being directed almost entirely toward the several types of junction transistor.
Fig. 511. Polystyrene-mounted multivibrator-sine-wave generator. The plastic chassis is 2"x3".
Junction types hold more promise for several reasons: They are more rugged and reliable than point contacts and, due to present manufacturing emphasis, are readily available. Earlier it was stated that junction circuits can be designed by analogy with conventional tube circuits; the tried and proved techniques of vacuum-tube practice can be almost painlessly carried over to transistors.
In this chapter we have three trigger circuits developed by the analogy method. Each has different applications: the bistable multivibrator is a gating device, the one-shot multivibrator produces a gate of variable duration, the blocking oscillator generates short pulses. Type 2N98 n-p-n junction transistors are used because the high collector supply voltage and high pulse output with good rise-and decay-time characteristics are particularly suited for these and similar switching circuits. Thus the figures all show n-p-n rather than the more common p-n-p circuits.
The transistor version of the popular bistable multivibrator is shown in Fig. 512. Feedback is provided by the dc coupling fromcollector to base, through the voltage divider (the 22,000- and 8,200-ohm resistors) and through the 680-ohm common-emitter resistor. This differs slightly from the usual tube circuit where the common-cathode resistor is omitted and the shunt arms of the divider are returned to a negative voltage. Actually, the 680-ohm resistor, in addition to biasing the off transistor beyond cutoff, stabilizes the on transistor and aids the start of regeneration whenever a trigger pulse is received
Fig. 512. The bistable multivibrator. The values of the capacitors across the 22,000-ohm resistors may have to be determined experimentally.
The 200 tiiif speedup capacitors allow rapid changes, such as trigger pulses, at the collectors to be coupled immediately to the opposite base and, therefore, also help start the switching action. Their values are determined experimentally.
Fig. 513. Output waveform of bistable multivibrator at 100 kc, 10 volts p-p.
Although the triggers are shown applied through the isolating diodes, to the collectors, other points could have been chosen.
The output waveform of the device is shown in Fig. 513. The peak-to-peak amplitude is approximately 10 volts and the repetition rate 100 kc. This does not represent the maximum frequency at which the circuit will operate. The photograph was taken while the circuit was still on a breadboard, where no particular attention was given to circuit dress.
Fig. 514. This emitter-coupled circuit is a one-shot multivibrator.
The emitter-coupled circuit shown in Fig. 514 is derived from the cathode-coupled tube version. There are two advantages of this over other types of one-shot multivibrators—it is highly stabilized by the common-emitter resistor; the width of its output pulse can be easily controlled by varying the setting of the 1,000-ohm potentiometer. Stabilization is obtained by the negative feedback pro¬duced by the 680-ohm resistor when the circuit is not in transition (switching states). Control of the pulse duration is not quite so obvious.
Fig. 515. Output waveform of one-shot multivibrator, 10 volts p-p.
In the stable state V2 is on because its base return, through the 68,000-ohm resistor, is positive. When a negative trigger is applied, through the isolating diode and the 50-LiLif coupling capacitor, the circuit switches to its quasi-stable state—VI on and V2 off.
Fig. 516. Transistorized version of the basic blocking oscillator.
At the end of the transition, the timing action begins as the capacitor charges. This causes the base of V2, which has been driven negative, to rise toward the 13-volt supply. This rise continues as long as V2 is off, i.e., until the base attains the emitter voltage. But the emitter voltage is developed across the common-emitter resistor and is equal to the difference between E and the base-emitter drop
Fig. 517. Output waveform of blocking oscillator at 60 kc, 2.5-microsecond duration.
of VI, the on transistor. For all practical purposes, the base-emitter drop of a conducting transistor is zero. Thus, the voltage of the emitter of V2 is E. The base, therefore, rises until it reaches E volts, at which time the circuit flips back to its original state. Thus, the duration of the output pulse is determined by the magnitude of E and can be varied by changing the potentiometer setting.
This is demonstrated in Fig. 515 which shows the output waveform at two values of E. For each, the amplitude is approximately 10 volts peak-to-peak and the triggering rate is in the order of 100 kc.
The transistorized version of this familiar circuit, as shown in Fig. 516, is free-running. The time between pulses is controlled by changing the setting of the variable resistor or the value of the 2.5-volt source. This can be seen by considering the timing process.
During the pulse, the capacitor charges. The charge, after the pulse has terminated, biases the transistor beyond cutoff. As the capacitor now discharges through the variable resistor, the base rises toward 2.5 volts. When it reaches zero, the transistor conducts and another pulse is generated. The off period is thus determined by the slope of the discharge curve, which in turn is a function of the capacitor, the variable resistor and the 2.5-volt power supply source.
Fig. 517 shows the 6-volt output pulse at a repetition rate of 60 kc, having a duration of 2.5 Lt sec. Ringing is prevented by the damper diode across the transformer tertiary, which acts to short out the positive excursion of the oscillation.
The waveform is remarkably free from jitter as compared with a tube blocking oscillator. Jitter is caused by fluctuations in the conducting potential. The transistor conducts a zero base-emitter voltage and the total variation of its conducting potential is very small. However, the cutoff level of a vacuum tube may readily change by large amounts.
The principle of analogy provides us with a powerful method of design. We can directly transistorize tube circuits. It is especially applicable to trigger circuits and brings us that much closer to the time when the technician will be servicing transistor radios and television sets daily.