On 6/11/19 10:56 AM, three_jeeps wrote:
A colleague drops by my office toating a Tek R293 Programmable pulse generator and power supply and asks "Can I use a power supply?" followed by, I haven't turned it on in 35 years....
Having never seen one of these before (its circa 1966) and thought it might be an interesting thing to restore, even though I have never and probably will never have use for such a thing.

Looking through the manual scanning for the semiconductor parts list (and looking for the 'special' Tek parts, I see designations on a number of diodes: Snap Diode, Check. No number of any sort.

I never heard of one of these so google is my friend....(aka snap off diode or charge-storage diode or memory varactor) studied the schematics of the pulse generator a bit and realized what a neat device and interesting circuit design.

It probably has a lot of bad parts, given the age, and while repair would be quite a learning experience, my basis questions a
1) how to determine the snap diode characteristics for this unique tek part?
2) and what are the chances of finding a reasonable substitute?

Assuming one of the snap diodes in the unit still functions, I could use a curve tracer and get some general parameters: Breakover voltage, approximate current range, reverse blocking voltage, leakage, etc. But determining the switch time would be critical in this application as that dictates the pulse characteristics.

Another thing that is of interest is this statement from Wikipedia:
"The main phenomenon used in SRDs is the storage of electric charge during forward conduction, which is present in all semiconductor junction diodes and is due to finite lifetime of minority carriers in semiconductors."
Finite lifetime of minority carriers....I assume when switching from forward conduction to reverse the minority carriers are quickly reversed in their flow given the abrupt change of voltage polarity?
J



SRDs have been used for aeons. Besides pulsers, as in the 20-ps-class
Tek SD24 TDR sampling head, SRDs are used as high-order frequency
multipliers. You can still get them from Microsemi and maybe
elsewhere--Arrow has 433 Microsemi GC2510 devices in stock.

One of my first engineering tasks back in 1981 was to rebias an X-band
waveguide frequency multiplier using an SRD--we were changing from a
Microwave Associates device to a Gigahertz Devices one.

The SRD was self-biased with a thick-film conductive ink drop on the
side of the ceramic package. I used a soft pencil to make the resistive
shunt on the new device, and varied it till I found the value that
produced the highest efficiency.

The job was complicated by the tendency of the graphite to ablate.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net