BFR92AW, B-E junction saturation?

I calculated the cap so that, together with the Cbe of the transistor acting as a voltage divider, it can never cause a reverse Vbe even close to the limit. You really don't need much there other than a couple pF as long as then the DC hold makes sure the transistor stays on afterwards.

The only thing I am worried about would be the day when I can't mountain-bike anymore. That could become depressing. I think I busted a suspension pivot bearing on the ride yesterday so it has to go back into the shop, which already causes pre-withdrawal symptoms in me.

Mercury cells have been outlawed long ago. If you use silver cells and the voltage has to match, like it does on cameras, a Ge diode drops just about the right amount. I rejuvenated a Minolta 35mm camera that way. And then 35mm film fell from grace when everything went digital. Now I am wondering what to do with that camera, actually got two of them.

Doesn't that cause awful phase noise? Avalanche is usually not a very time-precise and consistent method of generating a pulse. And avalanching ages transistors.

You can buy "coins" that have a holder for a silver cell on one side and a built-in diode to replace mercury camera batteries. They aren't cheap, though. I have a few for old cameras I bought off eBay some time back.

Many of the Tek samplers used avalanche transistors to make the sampling pulses, including their first sampler, the type N plugin, a single-diode sampler. Later scopes used an avalanche to make the sample pulse or as the SRD driver. Jitter was single-digit ps RMS, and they seemed to last forever.

These avalanches were triggered, not free-running, which seems to make them time-precise.

So what do you do with these cameras?

I know there's an almost religious following of B/W photography but that is a very small group of people who are probably almost artists.

They probably bred their own avalanche transistors, feedings them only barley, caviar on blini and steroids while growing. I've seen lots of avalanche pulser designs because that is what people immediately jump onto when tasked with that. Using commercially available transistors the phase noise was horrid even though all those pulsers were triggered. Mostly these were university projects and then I showed them how to use hotrod RF transistors wham-bam style, followed by a pulse shaper.

Not much, now. No time. I'd like to get back to it, though.

I like B&W photography. At one time I shot about 20x36exp. rolls of Tri-X (and some Ilford and IR) a week. It's a real challenge, as opposed to color which too easily turn into snapshots. For that purpose, my phone is good enough.

barley, caviar on blini and steroids while growing."

Yup, and you apparently managed to pull grid current.

It seems pretty straight-forward--you bias the transistor near breakdown, then hit the base with a fast pulse. Jim Williams used it years later too, probably from reading the Tek stuff.

That's easier and cleaner these days, at least for low voltage pulses and low power. If you want an amp for a nS, avalanche is still worth a look.

Cheers, James Arthur

That's the hardcore stuff. I met one guy whose company actually makes photo lab timers for B/W film development. He gave me a tour of his business and said that there is a solid customer base for this which greatly surprised me.

What what I've seen and measured at universities I was never really impressed with avalanching, not when it comes to noise in the pulser signal.

Amp? Avalanching is normally used in pulsers. But nsec stuff is something I'd never even consider avalanche for, that can be dones with cheap RF transistors. Even sub-nsec like the project I just finished.

The Zetex avalanche transistors are impressive. When they fire, they become near perfect shorts. 350 volts, 80 amps, 28 kilowatt pulses from a SOT23! But they are slow at those currents, ns risetime, wirebond inductance maybe.

They seem to have less jitter than I could measure on my 11802, which runs maybe

Last I heard, they were made in Russia, probably using an old diffused-junction transistor recipe.

For nsec risetimes at high power I'd try LDMOS.

That is quite impressive. I have never seen anything this low with avalanche designs. But they weren't my designs and so I wasn't involved in the layout.

And the day old Anatoly dies they can't produce them anymore because nobody else make the magic potion :-)

There is. There is also a market for this hardware among college students. 35mm photography is still taught in journalism schools. Film photography teaches a lot more about lighting, composition, and texture than does digital. The neat thing is that you can get what once was high end hardware for $.05 on the dollar (or less). Stuff I could never even dreamed to have been able to afford thirty years ago I can get for pocket lint, today.

Betcha you can't pulse 28 kilowatts with a SOT23!

I bet It could be done, once!

Jamie

I just rejuvenated all my Minolta lenses by buying an Olympus PEN plus a micro 4/3rd to Minolta adapter. That also requires full manual skills with the camera. But it's lenses you can't buy anymore. And the 250mm mirror tele lens is now like a 500mm.

I don't want to pulse 28kW with a SOT23.

Some people just don't want to have fun.