little high voltage supply

If you could wind your own transformers (or find you local transformer winder and get them to wind some for you) you could build the much simpler Baxandall inverter

Yarrow C.J., Proc. I.E.E 106 B, 1320 (1959) "Transistor Converters for the generation of High-Voltage Low Current D.C. Supplies"

was presented at the same conference, back in 1959, if two days later. I've got a copy which I could e-mail to anybody interested, but only for private study.

Much later Jim Williams wrote a series of application notes for Linear Technology, on high frequency inverters for driving cold cathode back-lights used in laptop computers (application notes AN45, AN49, AN51, AN55, AN61, AN65), using the same circuit, though he doesn't seem to have known where it came from.

It works better with MOSFET switches, and doesn't "squegg" as it can with bipolar switches, if you make the feed inductor too inductive.

We make high voltage APD supplies more or less that way, starting with a

Cheers

Phil Hobbs

"Bill Sloman" snipped-for-privacy@ieee.org wrote in message news:105fff1$2k200$ snipped-for-privacy@dont-email.me...

...

This should be able to make a KV or two depending on what you follow it with.

The simulation starts only if R2 and R3 are slightly different values.

Version 4.1 SHEET 1 880 696 WIRE 288 -176 288 -224 WIRE 288 -176 208 -176 WIRE 368 -176 368 -224 WIRE 512 -176 368 -176 WIRE 592 -176 512 -176 WIRE 208 -144 208 -176 WIRE 288 -128 288 -176 WIRE 368 -128 368 -176 WIRE 224 -16 160 -16 WIRE 336 -16 304 -16 WIRE 368 -16 336 -16 WIRE 512 -16 448 -16 WIRE -16 48 -128 48 WIRE 336 48 336 -16 WIRE 336 48 64 48 WIRE -128 96 -128 48 WIRE 160 112 160 -16 WIRE 304 112 160 112 WIRE 512 112 512 -16 WIRE 512 112 368 112 WIRE 160 192 160 112 WIRE 192 192 160 192 WIRE 272 192 256 192 WIRE 416 192 400 192 WIRE 512 192 512 112 WIRE 512 192 480 192 WIRE -128 224 -128 176 WIRE 160 240 160 192 WIRE 512 240 512 192 WIRE 272 320 400 192 WIRE 272 320 208 320 WIRE 400 320 272 192 WIRE 464 320 400 320 WIRE 272 368 272 320 WIRE 400 368 400 320 WIRE 160 384 160 336 WIRE 512 384 512 336 WIRE 272 496 272 448 WIRE 272 496 192 496 WIRE 400 496 400 448 WIRE 400 496 272 496 WIRE 192 528 192 496 WIRE 192 656 192 608 FLAG 160 384 0 FLAG 512 384 0 FLAG 208 -144 0 FLAG 512 -176 do-not-touch FLAG 192 656 0 FLAG -128 224 0 SYMBOL nmos 208 240 M0 WINDOW 3 61 71 Left 2 SYMATTR Value IPA126N10N3 SYMATTR InstName M1 SYMBOL nmos 464 240 R0 WINDOW 3 64 73 Left 2 SYMATTR Value IPA126N10N3 SYMATTR InstName M2 SYMBOL ind 208 0 R270 WINDOW 0 32 56 VTop 2 WINDOW 3 5 56 VBottom 2 SYMATTR InstName L1 SYMATTR Value 7000nH SYMATTR Type ind SYMBOL ind 352 0 R270 WINDOW 0 32 56 VTop 2 WINDOW 3 5 56 VBottom 2 SYMATTR InstName L2 SYMATTR Value 7000nH SYMATTR Type ind SYMBOL ind 384 -144 R90 WINDOW 0 4 56 VBottom 2 WINDOW 3 32 56 VTop 2 SYMATTR InstName L3 SYMATTR Value 40mH SYMATTR Type ind SYMATTR SpiceLine Rser=1000 SYMBOL res 384 -240 R90 WINDOW 0 0 56 VBottom 2 WINDOW 3 32 56 VTop 2 SYMATTR InstName R1 SYMATTR Value 1k SYMBOL cap 368 96 R90 WINDOW 0 0 32 VBottom 2 WINDOW 3 32 32 VTop 2 SYMATTR InstName C1 SYMATTR Value 100n SYMBOL diode 256 176 R90 WINDOW 0 61 32 VBottom 2 WINDOW 3 -34 26 VTop 2 SYMATTR InstName D1 SYMATTR Value UPSC600 SYMBOL diode 416 176 M90 WINDOW 0 60 34 VBottom 2 WINDOW 3 -32 30 VTop 2 SYMATTR InstName D2 SYMATTR Value UPSC600 SYMBOL res 256 352 R0 SYMATTR InstName R2 SYMATTR Value 68 SYMBOL res 384 352 R0 SYMATTR InstName R3 SYMATTR Value 67 SYMBOL voltage 192 512 R0 WINDOW 123 0 0 Left 0 WINDOW 39 0 0 Left 0 SYMATTR InstName V1 SYMATTR Value 12 SYMBOL voltage -128 80 R0 SYMATTR InstName V2 SYMATTR Value 24 SYMBOL ind -32 64 R270 WINDOW 0 32 56 VTop 2 WINDOW 3 5 56 VBottom 2 SYMATTR InstName L4 SYMATTR Value 56000nH TEXT -104 464 Left 2 !.tran 0.01 startup TEXT 400 -104 Left 2 !K1 L1 L2 L3 0.97 TEXT -304 -168 Left 2 ;Quick HV generator by ER 19 July 2025

And you can cook your porridge on L3.

..>>

LOL I was hoping to see what Bill might say.

Are there any capacitors which could handle being C1?

A low loss polypropylene would be a candidate.

Changing L3 series resistance to 5 ohms gives a more reasonable power dissipation in L3 but whether that's achievable with a real transformer I do not know.

Fun, but not very practical.

Assuming you rectify the output sine wave, you only get about 600 volts. How do you adjust the output voltage?

Where do you get the transformer and C1?

..

With this circuit you likely have to adjust the input voltage.

Very good question. A custom designed transformer is likely required. And C1 might be a few components in parallel. Looking at the current waveform through C1 it's true that I wouldn't like to be C1 in this circuit.

"john larkin" snipped-for-privacy@glen--canyon.com wrote in message news: snipped-for-privacy@4ax.com...

I can do a nice clean simulated 1.3 KV DC if you only need 75W.

If you want it stabilized I'd suggest adding a shunt regulator.

Version 4.1 SHEET 1 1016 696 WIRE 432 -512 288 -512 WIRE 544 -512 496 -512 WIRE 432 -400 368 -400 WIRE 544 -400 544 -512 WIRE 544 -400 496 -400 WIRE 592 -400 544 -400 WIRE 672 -400 592 -400 WIRE 832 -400 752 -400 WIRE 896 -400 832 -400 WIRE 944 -400 896 -400 WIRE 592 -384 592 -400 WIRE 944 -384 944 -400 WIRE 832 -368 832 -400 WIRE 288 -288 288 -512 WIRE 432 -288 288 -288 WIRE 544 -288 496 -288 WIRE 592 -288 592 -320 WIRE 592 -288 544 -288 WIRE 832 -288 832 -304 WIRE 832 -288 592 -288 WIRE 944 -288 944 -304 WIRE 944 -288 832 -288 WIRE 288 -224 288 -288 WIRE 304 -224 288 -224 WIRE 368 -224 368 -400 WIRE 944 -224 944 -288 WIRE 368 -176 368 -224 WIRE 432 -176 368 -176 WIRE 544 -176 544 -288 WIRE 544 -176 496 -176 WIRE 288 -112 288 -224 WIRE 368 -112 368 -176 WIRE 224 -16 160 -16 WIRE 336 -16 304 -16 WIRE 368 -16 336 -16 WIRE 512 -16 448 -16 WIRE -16 48 -128 48 WIRE 336 48 336 -16 WIRE 336 48 64 48 WIRE -128 96 -128 48 WIRE 160 112 160 -16 WIRE 304 112 160 112 WIRE 512 112 512 -16 WIRE 512 112 368 112 WIRE 160 192 160 112 WIRE 192 192 160 192 WIRE 272 192 256 192 WIRE 416 192 400 192 WIRE 512 192 512 112 WIRE 512 192 480 192 WIRE -128 224 -128 176 WIRE 160 240 160 192 WIRE 512 240 512 192 WIRE 272 320 400 192 WIRE 272 320 208 320 WIRE 400 320 272 192 WIRE 464 320 400 320 WIRE 272 368 272 320 WIRE 400 368 400 320 WIRE 160 384 160 336 WIRE 512 384 512 336 WIRE 272 496 272 448 WIRE 272 496 192 496 WIRE 400 496 400 448 WIRE 400 496 272 496 WIRE 192 528 192 496 WIRE 192 656 192 608 FLAG 160 384 0 FLAG 512 384 0 FLAG 192 656 0 FLAG -128 224 0 FLAG 944 -224 0 FLAG 896 -400 hvout SYMBOL nmos 208 240 M0 WINDOW 3 61 71 Left 2 SYMATTR Value IPA126N10N3 SYMATTR InstName M1 SYMBOL nmos 464 240 R0 WINDOW 3 64 73 Left 2 SYMATTR Value IPA126N10N3 SYMATTR InstName M2 SYMBOL ind 208 0 R270 WINDOW 0 32 56 VTop 2 WINDOW 3 5 56 VBottom 2 SYMATTR InstName L1 SYMATTR Value 6800nH SYMATTR Type ind SYMBOL ind 352 0 R270 WINDOW 0 32 56 VTop 2 WINDOW 3 5 56 VBottom 2 SYMATTR InstName L2 SYMATTR Value 6800nH SYMATTR Type ind SYMBOL ind 384 -128 R90 WINDOW 0 4 56 VBottom 2 WINDOW 3 32 56 VTop 2 SYMATTR InstName L3 SYMATTR Value 47mH SYMATTR Type ind SYMATTR SpiceLine Rser=5 SYMBOL cap 368 96 R90 WINDOW 0 0 32 VBottom 2 WINDOW 3 32 32 VTop 2 SYMATTR InstName C1 SYMATTR Value 100nF SYMBOL diode 256 176 R90 WINDOW 0 61 32 VBottom 2 WINDOW 3 -34 26 VTop 2 SYMATTR InstName D1 SYMATTR Value UPSC600 SYMBOL diode 416 176 M90 WINDOW 0 60 34 VBottom 2 WINDOW 3 -32 30 VTop 2 SYMATTR InstName D2 SYMATTR Value UPSC600 SYMBOL res 256 352 R0 SYMATTR InstName R2 SYMATTR Value 68 SYMBOL res 384 352 R0 SYMATTR InstName R3 SYMATTR Value 67 SYMBOL voltage 192 512 R0 WINDOW 123 0 0 Left 0 WINDOW 39 0 0 Left 0 SYMATTR InstName V1 SYMATTR Value 12 SYMBOL voltage -128 80 R0 SYMATTR InstName V2 SYMATTR Value 24 SYMBOL ind -32 64 R270 WINDOW 0 32 56 VTop 2 WINDOW 3 5 56 VBottom 2 SYMATTR InstName L4 SYMATTR Value 56000nH SYMBOL diode 432 -496 R270 WINDOW 0 32 32 VTop 2 WINDOW 3 0 32 VBottom 2 SYMATTR InstName D3 SYMATTR Value VS-E7MH0112 SYMBOL diode 432 -384 R270 WINDOW 0 32 32 VTop 2 WINDOW 3 0 32 VBottom 2 SYMATTR InstName D4 SYMATTR Value VS-E7MH0112 SYMBOL diode 496 -272 M270 WINDOW 0 32 32 VTop 2 WINDOW 3 0 32 VBottom 2 SYMATTR InstName D5 SYMATTR Value VS-E7MH0112 SYMBOL diode 496 -160 M270 WINDOW 0 32 32 VTop 2 WINDOW 3 0 32 VBottom 2 SYMATTR InstName D6 SYMATTR Value VS-E7MH0112 SYMBOL res 928 -400 R0 SYMATTR InstName R1 SYMATTR Value 22k SYMBOL cap 576 -384 R0 SYMATTR InstName C3 SYMATTR Value 47n SYMBOL cap 368 -240 R90 WINDOW 0 0 32 VBottom 2 WINDOW 3 32 32 VTop 2 SYMATTR InstName C2 SYMATTR Value 100p SYMBOL ind 656 -384 R270 WINDOW 0 32 56 VTop 2 WINDOW 3 5 56 VBottom 2 SYMATTR InstName L5 SYMATTR Value 5mH SYMATTR SpiceLine Rser=1 SYMBOL cap 816 -368 R0 SYMATTR InstName C4 SYMATTR Value 330n TEXT -104 464 Left 2 !.tran 0.02 startup TEXT 384 -88 Left 2 !K1 L1 L2 L3 0.97 TEXT -304 -168 Left 2 ;Quick HV generator by ER 19 July 2025

Regulating the output of a Baxandall inverter can be done by mark-to-space switching the input of the feed inductor between OV and your chosen supply voltage.

I've simulated it, centring the high voltage period around the peak voltage at the centre tap, and the 0V periods around the point where the regular Baxandall switches swap.

It would be pretty efficient - much better than any kind of shunt regulator.

Setting up the switching waveforms would takes a effort. It's the sort of thing you could do easily with a fast single chip micro, but it needs feedback to match the switching frequency to the actual resonant frequency of the transformer, which you could get by sampling the leading and trailing shoulders of the half-sine wave you get at the centre tap.

This wouldn't work well for very low output voltages.

There a Cuk-type low ripple variant were you add a pair over-windings to the inductor to let you create a very nearly square wave output.

That dual-winding DRQ127 is a wonderful surface-mount part. We pay 89 cents.

MMBD5004 is a dual 400 volt diode in a SOT-23 package. It costs us 5 cents.

Applying engineering should make things simpler, cheaper, better, and done.

I have a nice old Bertan 3KV bench supply, but I'll have two or three guys playing with GaN circuits who need supplies, and they can't have mine, and HV bench supplies are crazy expensive. We can piggyback a couple of this circuit on a 4-layer proto board.

I should include an easy way to measure current.

Just bog it down until you get the voltage you want?

I was thinking of making a bipolar HV supply starting with 2KV maybe, and using a full bridge of high voltage phototransistors as the regulators.

One can run a dual C-W string, positive and negative, from one AC source. A lot of C-W stages makes the magnetics much easier.

Cameras don't have xenon flashes much any more, but people still sell "capacitor-charging" transformers.

Lawrence Livermore considers 9 joules as the threshold of lethality. A little C-W multiplier is probably millijoules.

Sure, if you can get enough bandwidth. Phototransistors are pretty high-Z beasts.

For protos I usually use 2N6520/6521 class-B outputs, possibly Darlington connected, run from a CE stage, with capacitive local feedback and an op amp wrapped around it all. That’s great for unipolar stuff like photodetectors and most piezos.

Yup.

That’s a threshold I prefer to stay far away from. My power limit for HV products is that I personally have to feel comfortable shorting anywhere in the circuit with my hand. So I use a lot of LND150 depletion FETs and let the C-W sag some at heavier loads.

With the FETs inside the regulator loop, you get a nice rectangular current limit.

Cheers

Phil Hobbs

Phototransistors (and photo-FETs) pnly make sense in the context of a fully floating high voltage output. For most non-floating applications you end up with a resistive divider to ground which you can tap at a few volts off ground to get you feedback signal.

Particularly when you cripple yourself by only using off-the-shelf transformers.

The stored energy does rise as CV^2. Lots of volts can make a small capacitance lethal.