12-24v to 400v DC to DC boost converter help

[PCB.Wiz]

I am wanting a DC to DC boost circuit that can take a variable 12-24v and turn that into around 300-400v @ 20-30mA.
.... I'm working on this as a power supply to be a part of a high voltage mosfet audio preamp. The audio preamp circuit is figured out already just need a way to power it.

Do you want +/-200V or just +400V

How stable is the load current ?  What is your preamp PSRR, over frequency - that determines how low a ripple you really need.

The pre-built module linked above, looks like a good place to start. That uses a good looking transformer, which helps when trying for a wide voltage change.

A search finds a fuller spec of the same thing is here
https://www.aliexpress.com/item/1005004091462517.html

- says they have split rail or single rail options, 40W continual 70W peak,  US $2.92
Run one of those up, and measure the ripple.

Do you have an oscilloscope ?

Looking for an output of a steady 400v.

Getting a decent enough low ripple is fine for now, as I can tweak it later as need be, I just need a schematic that has the right components and voltages for my specific use case. Just need it to do the voltage boost for now, filtering can be chosen later but may be important to pre choose a circuit where that kind of modification isn't too difficult.

I do have an oscilliscope.

The very simplest design could use a single inductor, but the high voltage ratio here needs > 95% conduction times, and your power is not mW.

Start with the AliExpress module and measure it, and look up the ICs. Their data will give a circuit, which they likely just cloned.

A decent transformer is likely needed, and the Aliexpress module gives you a nice transformer and everything else essentially 'for free'.

[PCB.Wiz]

The requirements of this circuit are very much akin to what a tube preamp would be, however it uses a high voltage mosfet in the preamp stage of the circuit to sort of "emulate" tube behavior. So we can essentially treat it as if we are powering that kind of a circuit.

So that's going to be largely Class-A, which should make ripple and regulation easier.

I did see this new-tech part, GAN039-650NTB,  a cascode GaN with very low Crss, might be fun to play with ?  Not cheap, but easy to cool.
http://www.nexperia.com/ccpak

[badchardev]

This looks super interesting! I am gonna check this out!

[Martinn]

Made a few pictures of this Chinese module. I intended to build a simple HV lab supply with this, but finally just got a real (Chinese) one.
Scope shot is from switch node. Some really strange switching here, does not look as if the UC3843A does what it is supposed to do.
Also as mentioned earlier not the lowest noise option one might imagine. Bipolar output with center tapped secondary is a nice bonus. Even somewhat regulated due to flyback connection.

Here https://impulsite.ru/viewtopic.php?t=1294#p23960 you can find the schematic. Googling for yh11068a will give you many more hits.

If you read the whole article from page 7 Jim Williams has a whole section on low noise flyback converters, difficult I agree, but obviously not impossible.

I have not memorized all of JW's writings. These examples are a bit more low power though.

I assume the TO will not need isolation. For an isolated supply, getting rid of CM noise is a whole different story (and much more difficult than filtering output ground referenced noise).

I'd probably start with the yh11068a module - if it's good enough, then that is hard to beat.
Next would be a current-fed resonant Royer (Baxandall), these are very easy to DIY and winding a transformer (N87 RM core, maybe RM10?) pretty easy.
Or a push-pull forward, maybe even fixed duty cycle? Depends on what regulation quality is required.

[Martinn]

For fun, I wired it up with my Tube load.
I do have a proper electronic load, but not for 400 V.

[Vovk_Z]

Some really strange switching here, does not look as if the UC3843A does what it is supposed to do.

Looks like it totally lacks proper snubbing and filtering. You may consider it as Lego brick where you have to add your own required parts (filters, snubbers, etc).

[KerimF]

Perhaps this could be solved by a conventional low-power square wave inverter and a diode full-bridge at the transformer output.   

[Martinn]

Perhaps this could be solved by a conventional low-power square wave inverter and a diode full-bridge at the transformer output.

Like driving a 230V mains transformer backwards? If it had two say 12 V secondaries, one could make a simple push-pull with just two transistors and fixed 50% 50 Hz drive.
Not sure what a mains transformer would do with square wave input. For sure it would be less noisy than a flyback.