the high voltage pulse generator

Mine came back from the assembly house on Monday. But yours looks nice because it's blue. Ours is green.

I had no choice, had to use RGP02 series with a trr in the 300nsec range because they have to withstand 1.5kV. But nothing gets hot.

Dang, I don't have any blinkenlights in mine.

Mine is a Royer with linear regulator plus linear pre-regulator (for low noise). That stuff gets hot and long term will be replaced with switchers. But I didn't want to risk a noise issue on the engineering run. The Royer itself and subsequent parts remain cool. But I am not going beyond 80kHz.

This one didn't bite me but I also made sure not to touch anything near where it says alto voltaje on the board. It's so easy to become complacent when only 12V goes in.

Schematic? (c;

ok, so what kind of damage control happens when the LED opens?

Jamie

I've posted fragments here, in other threads.

Of course, the latest schematic actually works!

Well, any part could fail. The regulator could short.

But probably nothing would be damaged. The LED might stop making light.

or it could open like LEDS do and no longer have your grn leg connected?

Jamie

Then I'd have 22 volts going into a gate driver instead of 17. It wouldn't do anything terrible.

But decent LEDs, like this Osram, seem to be very reliable, especially at low current.

Right. The MMBDs are typ 50 ns. That sure seems to matter.

It doesn't blink. It just makes orange. I wanted another LED to indicate HV present, but that hogged too much power.

I have a 1400 volt flyback+multiplier supply driving two linear regs, each 0-1200 volts out, controlled by customer DACs. Then the pulsers. They are doing some sort of optical polarization rotation thing.

I worked on this for hours, with a really tacky D25 breakout and a mess of wires, and didn't get zapped. Spooky. My voltmeter is a benchtop Fluke with a big radial-lead 1G resistor at the end of the clip lead, makes a 100:1 (or I guess 101:1) divider.

I decided to discharge the HV supply instead of waiting for it to bleed down. Bad idea. It made a nice PAPP! noise and blew out about 10 parts. Nuisance.

It occurred to me that I've been soldering and desoldering parts since I was about 5 years old.

John, that company logo of yours resembles someone lying in bed. Does that really convey the kind of corporate image you want to project? ;->

Obviously not. He wants them to believe his claims. Or did you mean supine?

Yes, a 10x reduction in trr does matter at 200kHz. Unfortunately it's slim pickens above a kV and no fast diodes available.

I'll put one in on the final production version. Although in my recent case all power to the electronics is cut if someone opens as much as a lid.

Similar here but Royer instead of flyback because it has to be low noise. You can run a Royer almost sinusoidal. The linear post-regulator is controlled by a computer. Well, both are ganged so I don't burn too much power in that linear post-regulator. But no pulsers.

I found out my Fluke 8845A (the one you recommended to buy) goes to almost 1100V on its own.

Thou shalt use some resistance to do that.

I didn't start this early but maybe around 7. Mom told me I can't have a soldering iron unsupervised. So I snuck out one day and bought one at the local hardware place, then hid it. Cost 10 Deutschmarks, made in Italy, a scary thing, _no_ PE connection and the open heating element coil be seen through holes. I guess I was lucky that I never had a wire accidentally poke through one and touch the 230V end of that coil. But it was fire-engine red which was cool.

You have a dirty mind. Or maybe a sleepy one.

It's actually an abstraction of Mount Tamalpais as seen from the Bay Bridge toll plaza, with the moon slightly displaced from its astronomically plausible location.

Mt Tam is 2574 feet high, what passes for "Highland" around here.

I used soldering guns, with discrete axial parts from old TV sets, no PC boards. I preferred a Wen250, which would solder battleship plates.

My uncle Sheldon "helped" me build crystal sets and neon blinkers and things. I don't think I really understood electronics until I was 12, when it all instantly made sense.

Having a knowledgeable mentor from an early age makes a *huge* difference in acquiring early advancement in such a technical field as this. You were lucky.

Nothing dirty about it! It totally reminds me of a late riser reaching out to put his alarm clock on 'doze' for as long as possible. Is that a fair description of how you run your company? ;->

I don't think it included the reg and opto you mentioned.

I begged my dad continuously for a soldering gun. Finally he got me one. But after growing up I became a fan of soldering irons and never went near a gun again. Now I'm a big Metcal fan.

Very lucky. Sheldon never much liked his own kids but loved to babysit me. I sat in his lap at age 4 and soldered parts in TVs that he was fixing. Since he always had a Dixie beer in one hand and a cigarette in the other, I was actually useful. And cheap surplus electronics was everywhere; the US government deliberately seeded the country with surplus radio and radar and all sorts of cool stuff.

I used to go to Radio Parts, an over-the-counter distributor in New Orleans. A 1/2 watt resistor cost 12 cents, more than the price of a Crystal mini hamburger, so the counter guys used to give me parts. The owner saw them doing it, so he started giving me tubes and 10-turn pots and stuff.

It's my turn to show the youngsters what I know. That's fun.

Thought I did. Here's the 1200 volt regulator, close to the final version:

Layout was tight so I removed the unnecessary parts.

Shorting the HV supply blew out Q4 and U9.

The gain is a little low from expected, probably voltage coefficient errors in the feedback resistors. Gotta check linearity to verify that.

There's also a dinky 0 to -400 reg, using an opto-LND250 cascode.