Back to Homepage & Index
This Page - The Unijunction Transistor and Applications

As with the tunnel diode on page 19 of this website, another almost forgotten semiconductor device is the Unijunction Transistor.
It is difficult to believe that anyone set out to specifically develop the unijunction transistor and it is possible that it was a spin off from jfet. development.
An N type UJT. apparently comprises an N type silicon bar with an off-centre P diffusion into one side part way along its length, as is shown conceptually in Figure 1.

The N type bar is lightly doped and thus resistive and probably 4 to 10 kilohms from end to end.
The bar has end connections labelled B1 and B2 and the heavily doped P region is closer to the B2 connection than it is to the B1 connection. The connection to the P region is called the Emitter. There is therefore only a single P-N junction, which is why the device is called a unijunction transistor.
Like the tunnel diode on Page19, the unijunction transistor exhibits a negative resistance region in its forward conduction characteristic; making it suitable for use in very simple oscillator circuits. However it is not capable of operation at very high frequencies as is the tunnel diode and historically its most common application was the production of sawtooth waveforms for timebases etc. A typical characteristic of the type TIS43 UJT is shown in Figure 2.
Without going into detailed discussion of its operation, a typical 'linear' saw-tooth oscillator is shown in Figure 3.
An oscilloscope trace of the waveform at the Emitter terminal is shown in figure 4.

It can be seen from Figure 4 that the output at the emitter terminal is a linear ramp which rises approximately 12 Volts over 600us with a fall time (fly-back for a timebase) which is fast in comparison.
The reader might ask what occurs at terminals B2 and B1. The answer is 'not a lot' or more correctly a lot for a very short time.
As might be expected there is a negative going voltage transient at terminal B2 and a positive going one at terminal B1; which are the volt drops across their respective series resistors for the current flowing during the precipitous discharge of the capacitor connected to the emitter terminal.
The B2 and B1 waveforms are shown in Figure 5 and Figure 6 respectively.


 

From these traces it appears that, during discharge, a current pulse of approximately 47mA flows into terminal B2 and a current of approximately 385mA flows from terminal B1 and that consequently the capacitor is discharged by a current of approximately 340mA into the Emitter terminal. Quite impressive considering that the power supply current limit was set to 20mA!
It might be expected that, should a negative going pulse be applied to terminal B2 prior to the discharge, the discharge can be started early i.e a timebase fly-back can be triggered; which is in fact the case.
Similarly the positive going excursion at terminal B1 might be applied to the cathode of a cathode ray tube
for fly-back blanking.
Among the author's (my) 'junk' is a 1" diameter cathode ray tube type 1CP1 ( Figure 7) dating from the 1950s and a possible ongoing application might be to develop a timebase for this device at least as a breadboard circuit. Such a timebase was conveniently featured in Wireless World in December 1970 and it will consitute the basis for further development -there is no point in re-inventing the wheel.
However it will be necessary to establish a power supply system for the for the 1CP1 crt. before timebase development can be undertaken. Also very conveniently a schematic of the tube requirements is published, free to use, by James Millen & Co and this is depicted in Figure 8; redrawn in accordance with my practice and components to hand. Thank you James Millen.
Rather than to develop just a timebase circuit, the possibility for taking things a little further and progressively building a very simple oscilloscope is attractive; particularly because the author is currently locked down by the global coronavirus pandemic. Of necessity this will not be a design by any stretch of the imagination as it will be cobbled together with whatever components are to hand. What is required in addition to the test circuit of Figure 8 is a converter circuit to produce the 650V supply for the tube and a lesser voltage to supply X and Y amplifiers. A separate mains transformer will supply the 6.3V 0.5A for the crt. heater. If the project comes to fruition it will be called 'Waveform Monitor type Covid 19'.
A schematic of the the converter circuit for derived supplies is shown in Figure 9. Ideally such circuit would be based upon an ad. hoc. transformer, with a soft ferrite core and windings for all supply voltages including the crt. heater, running at a few tens of kHz. The converter depicted in Figure 9 uses an iron cored mains transformer with a self running frequency of around 160Hz. The running frequency is determined by the transformer inductance and resistance and the 47nF capacitor serves only to suppress ringing on the transients. The prime mover dc. supply voltage is 20V to match that used for the basic timebase cicuit of Figure 3.
The breadboard layout for the converter circuit is shown in Figure 10. No effort has been put into making the layout compact. The circuit is laid out on a piece of stripboard salvaged from another project and space is available for further development as necessary. The waveform at the collector of one of the BD139 transistors is shown in Figure 11 with a running frequency which is determined by the characteristics of the transformer alone.
If input signals of around 1volt amplitude are to be displayed, simple 'X' and 'Y' amplifiers are required to drive the tube deflection plates via the 50nF capacitors shown in Figure 8 and these amplifiers will need to have gains of around 40dB. The amplifier schematic is shown in Figure 12.
As can be seen, it is a simple differential amplifiier configured to permit output voltage swings of +/- 50V or so. Ac. coupling is employed with a low frequency cut off to reduce the response to mains frequency pick-up.
An LTspice simulation of the frequency response of the Delection Amplifier is shown in Figure 13. The gain is to the collector of a single transistor so the differential gain will be 6dB greater than that shown in the response and close to the nominal 40dB target.
The stripboard layout of the Deflection Amplifiers is shown in Figure 14. The circuits for Y and X amplifiers are laid out graphically i.e. as they appear in the circuit schematic. This makes inefficient use of board area but it is ideal for trouble shooting and testing.
So far as it has progressed to date, W.M. Type Covid 19 is shown in Figure 15, where a Lissajous figure combination of a 400Hz. sinewave and a 400Hz square wave appears on the screen.
A timebase circuit having switched scan periods of 1s, 100ms and 10ms and based on the unijunction circuit of figure 3 is incorporated. This circuit includes a potentiometer which provides fine control (+/- 10% variation) of these periods. It is installed in the chassis but currently it cannot be used due to the existing capacitive coupling to the X amplifier and delection plates. What is required, and intended, is to convert to a negative supply for the cathode ray tube, thereby allowing the deflection plates to operate close to ground potential and thus to be directly connected. However the purpose of this page is to show a basic application of the unijunction transistor and further development of the monitor itself will not be recorded here.
Thank you for reading
All information is free to use.