A Tunnel Diode Paradox

Check the date on the article or book you've been reading.

Tunnel diodes had a very short life. They appeared really useful, but generally weren't. They were a neat curiosity, and thus made waves in the hobby magazines and hobby books, but there was nothing that other devices couldn't do. They were in those hobby publications because "hey neat, a diode can oscillate and amplify, and sometimes both at the same time".

Commercially, the only really long standing use seems to be triggers in oscilliscopes. That's the only real use anyone has posted about in these newsgroups in over a decade.

Hobbyists ask about them, not realizing they are long out of circulation.

They are expensive because they were made for a short period of time, and that was forty years ago. The real prime was the early sixties. If they are in production today, it is a very small production.

There is no reason (other than for those oscilliscopes) to use a tunnel diode today. Anything you can think of can be done other ways, and with cheaper and more commonly available parts.

If you really want to play with one for the sake of it being a tunnel diode, then you have to pay the price.

Michael

I know one VARIAN Raman-Spektrometer with a TD in it's cicuit. Probably in some sort of discriminator/trigger.

That was about 1975...

Yours, Holger

Oh -- don't forget that they're still fun!

I have a Heathkit tunnel dip meter, which works just fine (and would probably do so with a 2N4416, but wouldn't be such a technological curiosity).

See Michael's comments about availability -- to put it another way, they never took off commercially, so they're not around. But the things they were designed into need them to work, which drives up the value for those few that are left.

Sort of like those few vacuum tubes that are popular with audiophiles -- compare them to those many vacuum tubes that were only used in color TVs, and which you can now get for less than $1.00 apiece.

To answer your circuits question, assuming no oscillation the thing will tend to 'stick' at one of the two stable intersection points (this is probably why they were good for O-scope triggers -- I bet you could build a circuit that would reliably change state lickity split).

Where their perceived value really lay (and where other, more conventional devices quickly caught up) was in biasing them in that negative resistance region with a frequency-selective network, so that you'd get gain in some (way high, for those days) band that could be used for straight amplification or oscillation, depending on how you arranged your circuit.

It won't stay in the negative resistance region if the load line intersects at 3 points; it'll choose one of the two stable states. You can force it to assume the lower voltage state by temporarily making the diode current less than the valley current. You can force it to assume the higher voltage state by temporarily making the diode current exceed the peak current.

I bought a couple on e-Bay. They're rare and expensive.

It avoids the middle one because it's unstable there... it will scoot to one of the outer points. Which of the outer points it settles on depends on the bias history. Power up at that current, and it'll go the the left intersection. Temporarily increase the current past Ip, and then back down, and it'll settle back to the higher voltage right-most point. So it becomes a bistable Schmitt sort of thing.

TDs were super-cool, switching in as little as 25 ps when most everything else was orders of magnitude slower. But their fabrication process was insane.

The only TDs still used in any volume are planar back diodes, used as sensitive RF detectors. They have a sort-of modern fab process. Ip is low by design, so they're not useful as switches.

Too bad nobody makes planar switching TDs.

John

FYI, if all you want is to experiment with negative-resistance circuits, you can get a vaguely tunnel-diode-like I-V characteristic by connecting a pair of transistors:

And I hadn't run across this before today, but apparently you can get a negative resistance out of a common PN2222A, too:

And there is an article somewhere entitled, IIRC, "A crystal that can oscillate". Or something like that. Written in the in the 1920s in reference to 'crystal set detectors' it talks about what for all the world sounds like a negative resistance diode. Note.

A later comment was that this might have been a narrowly missed chance to have discovered the solid state transistor! But tubes were just being perfected and it was not til WWII was over effort was put towards solid state devices!

Recall that when first found the article I figured it was just something about biasing the crystal detector with an adjustable (bias) battery voltage so as to place it in the most efficient region for detecting the modulation content of radio signals! But it was more than that.

Will see if I kept record of it somewhere?

Ah. Maybe this was it (The crystodyne!)

Also this link: