Gunn diode oscillators

Updated February 10, 2008

Click here to go to our main page on microwave diodes

What's so great about a Gunn diode? They are used for cheap sources of microwave power. By the way, when we say "source", it means "oscillator". Just wanted to clear that up. John Battiscombe (J. B.) Gunn is in our Microwave Hall of Fame, check it out!

Another thing you should know: a Gunn device is not actually a diode, even though everyone uses this term! Why isn't it a "real" diode? Because it only uses N-type semiconductor (actually a three-layer stack). Its more accurate name is a transferred electron device.

But wait, here's a second opinion from John:

Your statement "a Gunn device is not actually a diode" isn't true. They ARE diodes, because the word diode simply means it has two (active) electrodes. It doesn't have to have a P-N junction. In fact the term diode was coined for vacuum tube diodes (these actually may have three or four connections, as they require a heater supply, which may or may not form one of the active electrodes) which were made before commercially produced P-N diodes. Semiconductor diodes did exist before this but they were point contact diodes, but weren't called this. They were of course crystals and cat's whiskers, as used in crystal sets.

Good point John, we misspoke in haste! Let's agree that a Gunn diode is like no other semiconductor diode in that it doesn't use a PN junction.

Gunn diodes have been around since John Gunn discovered that bulk N-type GaAs can be made to have a negative resistance effect. Gunn diodes have been a cheap source of microwaves ever since! They are used in many commercial applications for high frequency sources, including police radar, and even K-mart door openers. Ever wonder why your radar detector goes off when you pass a K-mart?

The I-V curves of a Gunn diode will help explain the effect. For low voltages (up to 1 volt perhaps), the Gunn diode behaves nearly as a linear resistor. Then at some point the current stops increasing with increasing voltage. This is known as the threshold voltage. Above this point the diode has negative resistance (curve slopes downward), which mean that it is just itching to oscillate! The operating point is usually about 4X the threshold voltage.

Below is a picture of a Gunn diode oscillator for W-band. Note the WR-10 waveguide, and the cheap heat sink. This bad boy must oscillate somewhere between 75 and 110 GHz, because that is the full extent of W-band. It is something we found in a lab drawer, for all we know it is a blown device. Nice use of a C-clamp to attach a heat sink!

How do you know if this Gunn diode is OK, without a spectrum analyzer that goes to 110 GHz? Put it onto a curve tracer! What's that? negative resistance above 1 volt? Yes, this is a good device!