FET as UJT?

In my semiconductor physics class, the prof said a past student project involved using a FET as an UJT. Now, the physics behind this is basically injecting carriers into the channel, basically super-enhancing it, in such a way that incremental resistance (with respect to the gate, which is now the emitter) goes negative.

Now, it seems to me that your average FET has a fairly conductive channel, maybe 300 ohms for a 2N3819 (give or take), and that's a weak FET as FETs go. So you super-enhance that, and it's going to drop to what, 30 ohms? 3 ohms? That's pretty low to make use of, at least without going *Phut*. As I recall, UJTs were in the 5kohm range, so they'd be lightly doped, and as a result fairly sensitive (large change in resistance).

Would it work better with source or drain as the common terminal? With lithographic construction, one end is bound to be longer, which means more length to make use of the injected carriers. Actually, that'll only be true if they make vertical FETs... maybe they're all lateral?

Tim

--
Deep Friar: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms
Reply to
Tim Williams
Loading thread data ...

Well, try it.

I did see one circuit in which the UJT was connected wrong, ends swapped. It did still oscillate, but the amplitude was low and the charging curve was nearly linear. We just fixed it and didn't investigate further.

There are jfets with very low Idss and high channel resistances.

I wonder if PHEMTs have a UJT mode. I have observed a sort of super enhancement in some PHEMTs as the gate began to conduct, maybe some bipolar action starting, but I haven't had time to explore that either.

John

Reply to
John Larkin

Depending on the type of circuit you want to build, you might want to try a lambda diode - two JFETs connected together. This forms a device with a negative differential resistance through a certain portion of its transfer curve... it's sort of a poor man's tunnel diode.

See

formatting link
for one interesting application.

--
Dave Platt                                    AE6EO
Friends of Jade Warrior home page:  http://www.radagast.org/jade-warrior
 Click to see the full signature
Reply to
Dave Platt

Perhaps an SCR might be better..

Reply to
Robert Baer

project=20

basically=20

such a=20

now the=20

channel,=20

FETs=20

ohms? 3=20

*Phut*. As=20

and as a=20

With=20

more=20

be true=20

I have seen a variation of them called programmable unijunction transistors (PUT).

Reply to
JosephKK

The foundry has to ship a lot of extra Ts, when Jorge uses them. ;-)

--
Offworld checks no longer accepted!
Reply to
Michael A. Terrell

...if you are adamant about NOT using programmable unijunction transistors, then use a CALL..

Reply to
Robert Baer

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.