HV dc/dc

25

What is your input voltage? I have similar need with 12V to 24V input.

Anything I want. I'll probably use a MeanWell brick for the main supply, 24 volts maybe. That would give me a reasonable flyback ratio.

This is a weird custom pulse generator box for a mysterious ultimate customer. Parts cost isn't a big deal, but it would annoy me to spend a kilobuck per box on little dc/dc things.

I was just looking at ISOW7821, as the gate driver. It has a freebie

5-volt dc/dc built in.

--

John Larkin         Highland Technology, Inc 

Science teaches us to doubt. 

  Claude Bernard

e=25

12V to 110V AC inverters are cheap and available, but they have a very narr ow input range (around 11V). 110V AC to 170V DC regulators are also cheap. So, for my app (4 li bat in serial), i am buck regulating 12V to 16V DC t o 12V, 12V DC to 110V AC, then 110V AC to 170V DC. Lots of hardware, but c heap and available.

Didn't you show some design earlier for a high voltage supply?

Swap your standard caps for the charge pump with y caps. Then it's safe wrt approvals and you reuse an earlier design

Cheers

Klaus

Even 12 to 24 volt 5 watt isolated converters are relatively expensive off the shelf, $20-25. Bleh

How about a C-W multiplier driven from an off-the-shelf boost regulator? In other words, don't use use one diode and capacitor, use a string of them.

-- john, KE5FX

At the 1-watt or 2-watt level, there are lots of nice little surface-mount dc/dc things for a few dollars. I use them a lot, even when I don't need isolation.

--

John Larkin         Highland Technology, Inc 

Science teaches us to doubt. 

  Claude Bernard

e=25

This one is pretty good for $7. It can display input or output voltage as well. But it does not have output protection, and i blow up the mosfet. I am goi ng to add output protection for cascaded booster.

Something like this should work.

Feedback into the LTC chip is interesting.

They want a working demo next week. Gotta warm up the Dremel.

--

John Larkin         Highland Technology, Inc 

Science teaches us to doubt. 

  Claude Bernard

I was going to say I used a LTC3803 design I copied from you... or someone else here.

George H.

If each channel needs some degree of control, the ultimate low-cost solution looks like this thing:

Strobe G, load whatever you want to SER IN. One of my faves; works like a charm.

If no control is needed, use 8 MOSFETs fed with the same 50% waveform, then use each of them as the switch driving a forward/flyback transformer, whatever you like, and burn excessive power in a zener.

Best regards, Piotr

I think a flyback with a minimal amount of secondary capacitive boost is the right choice here, even if isolation weren't required.

For that voltage step-up ratio with a boost and CW you can either put most of the step up into the boost, but then the small duty cycle increases stress and losses on the switch, or into the multiplier, which compromises your ability to drive pulsed currents without large caps, which again craps on your efficiency due to esr.

I made a EL panel driver good for a couple watts using a similar topology it works very well, 12V to about ~150 V DC into an HV809.

Didn't need isolation and the current draw is approximately constant so the dumbshit resistor + zener-thing off the secondary works well enough for feedback.

It could probably go lower than that I expect the flyback controller IC could be boostrap off the HV809's regulated output pin with some thought.

It's a wonderful little part. It just works, and the slope compensation is magical.

I can probably peak rectify the fet drain voltage, to make the feedback into the 3803. That will sorta regulate all the +200 things.

--

John Larkin         Highland Technology, Inc 

Science teaches us to doubt. 

  Claude Bernard

We use a ton of TPIC6595, a simlar part, as our universal relay and LED driver. But I don't need independent control, and there will be no uP or FPGA or anything. It's refreshing to design an all-analog box once in a while.

I think one big fet can drive all 8 flyback transformers.

--

John Larkin         Highland Technology, Inc 

Science teaches us to doubt. 

  Claude Bernard

With independent flyback transformers, you'll get lousy cross-regulation, and poor output voltage tolerance, due to Lp differences (that determine energy transfer).

You'd be better off going for preregulated voltage inverters, which are (at least) voltage regulated, to begin with.

Or fit all the flyback secondaries on the same core. ($)

RL

All 8 transformers will have the same turns ratio. I was thinking their primaries would be in parallel. Same input, same output.

I was planning to regulate in each isolated channel

so the primary-side flyback regulation doesn't need to be very good. Just good enough to make about 200 volts in each channel.

It doesn't need to be flyback. Some h-bridge forward converter would work, if I can get the right transformer ratios. But the flyback looks easy and versatile.

--

John Larkin         Highland Technology, Inc 

Science teaches us to doubt. 

  Claude Bernard

e

A single flyback with multiple secondaries seems like the cleanest solution . Since you would need 18 pins on a bobbin, it might be more practical to use two independent flyback converters, each with 4 outputs. There might a ctually be an off-the-shelf transformer that works for this.

I've never considered driving multiple flyback primaries from one switching FET, but it seems like the outputs could vary by a lot. ... ok you pique d my curiosity, I scribbled out the equations, using V = L dI/dt , and E = 1/2 L I^2, if you vary L by say 20% it looks like the transferred energ y will also vary by 20% (if I did that right). So for your idea it might j ust work. It might be a good idea to add a DC load at the outputs to soak up any energy from the transformers' leakage inductance.