mjj@stda.jhuapl.edu (Marshall Jose) (08/15/90)
Having recently obtained some Gunn sources and disassembled them
(carefully) to understand how they are put together, I became somewhat
confused about Gunn diode polarity. I am referring to the small
(MA/COM style 30) case, as roughly shown below:
_,--|_
|_| |_|
`--|
After consulting the RSGB VHF/UHF handbook and the Gunn data sheets
of both Alpha and MA/COM, my confusion was magnified from "somewhat"
to "utterly".
OK, so you can't tell from the package which end is anode (since no
manufacturer seems willing to mark them). So how DOES one determine
their polarity? I mean, besides the destructive method?
Thanks and Have a Swell Day,
Marshall Jose WA3VPZ
mjj%stda@aplcen.apl.jhu.edu || ...mimsy!aplcen!aplvax!mjjbeams@norvax.UUCP (David Beams) (08/21/90)
> Having recently obtained some Gunn sources and disassembled them > (carefully) to understand how they are put together, I became somewhat > confused about Gunn diode polarity. I am referring to the small > (MA/COM style 30) case, as roughly shown below: > > _,--|_ > |_| |_| > `--| > > After consulting the RSGB VHF/UHF handbook and the Gunn data sheets > of both Alpha and MA/COM, my confusion was magnified from "somewhat" > to "utterly". > > OK, so you can't tell from the package which end is anode (since no > manufacturer seems willing to mark them). So how DOES one determine > their polarity? I mean, besides the destructive method? > > Thanks and Have a Swell Day, > Marshall Jose WA3VPZ > mjj%stda@aplcen.apl.jhu.edu || ...mimsy!aplcen!aplvax!mjj I used to work with M/A-COM K-band Doppler transceivers (which incorporated a Gunn oscillator and Schottky mixer in one package, making a very simple Doppler CW radar for velocity measurement). In these devices (M/A-COM part number MA86859), the Gunn power-supply pin was positive with respect to the cavity (although that really doesn't answer the question, does it?). Gunn diodes are actually not rectifying junctions; the Gunn effect arises because the mobility of electrons in n-type GaAs decreases with increasing applied electric field once a critical E-field intensity is reached. This causes negative differential conductivity ("negative resistance," if you want to use the term); coupled with a cavity, it makes a dandy microwave source. Since a Gunn diode lacks a rectifying junction, it ought to be a bidirectional device. Let me throw in one caveat to this, however. If I recall correctly, the M/A-COM diodes were not completely bidirectional. They were fabricated by ion implantation of acceptors into an n-type GaAs die, leaving only the center of the die uncompensated with acceptors. The reason behind this was, I think, to reduce the capacitance of the diode without decreasing the physical size of the die (they wanted to avoid further increases in the power density in the device itself; the Gunns with which I worked drew approximately 150mA at +5V, and M/A-COM's Commercial Sources engineers told me that the Gunn operated at a steady-state temperature of +260C. . !). The result was that the M/A-COM K-band Gunns showed different V-I characteristics depending upon the polarity of the applied external DC bias. The M/A-COM K-band Gunns would draw considerably more current operated in "reverse" than in "forward," and a relatively simple test would determine correct polarity for these Gunns. Apply the rated voltage to the Gunn in its cavity (should be marked on the cavity itself, assuming you removed the Gunn from a complete cavity as- sembly) and note the current; reverse the polarity and note the current again. The polarity which resulted in the lower current is the correct one. I think that the "reverse-biased" M/A-COM Gunns drew over a half-ampere at 5V, while the "forward-biased" devices drew, as I mentioned, about 150mA. Interestingly enough, the "reverse-biased" Gunns would still oscillate, according to M/A-COM; they drew much more power, though, and were not guaranteed to survive very long. I have had no experience with M/A-COM X-band diodes nor Alpha diodes; I can't offer any advice in those cases. By the way, if your diodes' rated voltage is about 5V, it is probably a K-band device (24GHz); the M/A-COM X-band devices operated at +8V. (The higher the frequency at which the Gunn was to operate, the thinner the active region. A decrease in active-region thickness re- quires a decrease in operating voltage so that the applied electric field did not exceed a second, higher threshold at which the device would revert to positive differential conductivity). There is a manufacturing engineer from M/A-COM's Commercial Sources division whom I knew when I was working with these devices, and I would be happy to pass along his name via e-mail if you'd like. He may not be there now, as I have not dealt with him in over two years; but, if he is, he would probably remember me if you were to mention my name and that of my former employer (which I mention only as a form or curse or imprecation). -- UUCP: ...seismo!uwvax!astroatc!norvax!beams US Mail: David Beams / Norland Corporation / Fort Atkinson, WI 53538-8999 "Let's stop this finger-pointing and get down to some serious name-calling!"