UJT curve - mallet to the head needed

[Quantumplate]

Going insane once again, hope you can help. It's the emitter voltage/emitter current curve, the hill and valley one.
I don't get the fact that as one increases the input emitter voltage, it should keep going up and up surely? How can the graph show it plunging into a valley when you are increasing the input voltage?

[rfeecs]

What curve are you referring to?
The usual curve is Voltage vs current, not current vs voltage.

At the peak, the emitter-B1 junction is forward biased.  This causes extra carriers to flow into the region between emitter and B1 contacts.  The extra carriers will lower the resistance, so the voltage will drop, hence the negative slope of the curve.

[Quantumplate]

What curve are you referring to?
The usual curve is Voltage vs current, not current vs voltage.

Yes, the curve is the one you printed.
Thanks for replying. I understand what you said, but I'm still hung up on the fact that Ve-b1 is just the input voltage that is dropped across the pn junction and the resistance of b1. And I thought the graph should be plotting a linear increase in Ve (input) against Ie to see how the current varies. It would make more sense to me (the curve shown) if Ve-b1 labelled on the y axis was in fact the voltage across Rb1 only, so its resistance goes down as you say when it starts to conduct. But clearly one couldn't measure the voltage across Rb1 very easily! So still managing to confuse myself totally.

[free_electron]

well . it is a UJT . unijunction transistor. that is the specialty of such thing. when you go over a certain threshold they shoot into conduction. this collapses the required voltage to maintain conduction.  these things are typically used in oscillators.

[Quantumplate]

when you go over a certain threshold they shoot into conduction. this collapses the required voltage to maintain conduction. 

Ah, so once you get into conduction you can lower the input voltage and still conduct, this time with more current.
But for a given Ve-b1 (above the valley point minimum), how do you know which side of the valley point you are on?

[TimFox]

The curve graphed above, V_EB as a function of I_E, is not monotonic, and therefore its inverse I(V) is multi-valued, and not a "well-behaved function".

[free_electron]

when you go over a certain threshold they shoot into conduction. this collapses the required voltage to maintain conduction. 

Ah, so once you get into conduction you can lower the input voltage and still conduct, this time with more current.
But for a given Ve-b1 (above the valley point minimum), how do you know which side of the valley point you are on?

correct. that's whey they are often use in oscillators. use  r-c at the base. the voltage rises until the trip voltage. at that point the transistor shoots into conduction. it stays in conduction until the base voltage goes below the hold value. this thing is essentially atransistor that has input hysteresis. you need to pierce the upper threshold for it to go into conduction , and pierce the lower threshold to take it out of conduction.

[TimFox]

The UJT can do many things (not all) normally done by a 555 IC.

[rfeecs]

They're not really used anymore.  I think there are only a few models still made.

You might take a look at this GE Transistor Manual from 1964:
https://worldradiohistory.com/BOOKSHELF-ARH/GE-Books/GE-Transistor-Manual-1964.pdf

It has a whole chapter on unijunction transistors and applications, describing all the parameters, how they are measured, etc.

It also has chapters on other near extinct devices like tunnel diodes and silicon controlled switches.

These old manuals are great in that they explain semiconductors in a clear understandable way to engineers and hobbyists that were completely new to solid state at the time.

[Quantumplate]

These old manuals are great

Thanks for that. I like the old books too...well some of them. My UJT info I found in a course book from 1974 - ICS Electronics course distance learning. Quality paper in those days.

[free_electron]

wow. that's an interesting book ! saved. Thanks !
lots and lot of diagrams and pictures to explain it as opposed to a bunch of dry equations where you have no clue where they come from or what they actually do.

[Benta]

For similar behaviour of a semiconductor device, check out Diacs, widely used in dimmers.
Example: https://www.st.com/resource/en/datasheet/db3.pdf

[Quantumplate]

You might take a look at this GE Transistor Manual from 1964

Bought the book! Nice bedtime reading. Still a nagging question as to what that graph would look like if you were just raising the input voltage and carried on raising it beyond Vp.