high voltage safety

I've been using EMCO bricks, low-cost 0.5W to more serious 10 to 20 watts, a few hundred volts to 30kV. Their A series, 1 watt, 0.1 to 6kV is nice. The 0.5-watt Q series is tiny and cute.

Flyback is nice and easy for voltage ratios below say 40 to 50, but gets painful above that. The transformer secondary's capacitance is a pain, although specialized bobbin designs and special winding techniques can help.

Most of the commercial designs use a push-pull self-resonant oscillator on the primary side.

On Fri, 15 May 2015 11:29:43 -0700, John Larkin Gave us:

Not after a retarded, insulting bastard like you mouths off. You deserve a plate of s*it, and a plate of s*it is what you get fed, CHUMP. Wade in it, you stupid bastard.

A gentle reminder that the troll is on the loose again.

Post carefully,

Tim

-- Seven Transistor Labs, LLC Electrical Engineering Consultation and Contract Design Website:

A multiplier on the output helps with the ratio issue. A flyback is simple, since you can do everything, regulated, with one control chip and a fet, and maybe not the fet.

I only need a kilovolt at about one watt. Coilcraft has some nice "capacitor charging" flyback transformers.

Isn't that sort of class A, inefficient? Or more complex? Maybe needs custom magnetics.

The output multiplier has a quite high output impedance, which helps limiting the output current and hence the current flowing through a human.

So the needed current is only 1 mA and most likely the short circuit current (through a human) only a few milliamps, which should be quite safe.

I once built an ionizer driven directly from 220 V 50 Hz mains using Y capacitors and 1N4007 diodes (both of which I had plenty of in my junk box) and put several current limiting resistors in series at the output (to handle the resistor voltage limits).

In practice, those current limiting resistors were not needed, since even touching directly the top of the multiplier chain was harmless, due to the high impedance of the multiplier chain.

Anyway, for personal safety, make sure that not a single component failure would allow a large current to flow, use overrated components and check for creapage distances on the PCB.

Push-pull resonant circuit would most likely be class-C ?

On Fri, 15 May 2015 21:56:32 -0700, John Larkin Gave us:

"2 3904s driving a transformer." Is EXACTLY what I said.

Are you so stupid as to think the transformer output would be some other waveform? Much less what feeds it?

Go back to school, child.

IEC 479.

Yes and no, ... current(time)is the criteria. Depends what is the current path, e.g permanent damage on heart can be achieve with only 5mA/1s

Hab.

That doesn't mention "sine."

It could be a square wave forward converter.

Post a schematic and we'll discuss it.

On Sat, 16 May 2015 09:29:24 -0700, John Larkin Gave us:

Only an idiot feeds square wave drive into a transformer.

I did. Look it up.

I went through your posts but didn't see anything.

Most of us use square waves for efficiency. That's because unless we're adding a complicated class-D amplifier, using low-duty-cycle resonant drive, or some other sophistication, we have to suffer the class-A losses inherent in sine-wave drive. Also, square waves do not present any difficulty at all for properly-designed transformers. In most cases the input drive will be a square wave, even if the transformer output may not. I can give chapter and verse on an efficient 550-watt 600kHz 10kV example.

Now, for resonant very-high-voltage transformers, yes, some may be an exception. But to discuss one of these cases, the transformer details need to be spelled out. Also, the driver circuit is usually not all that efficient. Often everything revolves around the transformer. But in the 600kHz resonant example above, the input drive was a square wave.

So... No schematic?

Lots of idiots making lots of money with square wave drive into a transformer but it appears that you're not.

So... No schematic?

On Sat, 16 May 2015 15:27:29 -0700, John Larkin Gave us:

snip

I guess you are unaware of the task a choke performs in a switch mode supply.

What ends up feeding the xformer is not a square wave.

Nice try though, uneducated Johnny.

On Sat, 16 May 2015 20:40:27 -0500, John S Gave us:

did I not already tell you to F*CK OFF AND DIE, f*****ad?

On Sat, 16 May 2015 20:49:35 -0500, John S Gave us:

Sit your retarded ass right back down on that thumb and spin, you retarded POS.

On Sat, 16 May 2015 21:02:02 -0500, John S Gave us:

Nice try, chump.

So... You don't have an oscilloscope? Or, just don't know how to use it? That's a shame.

So... No schematic?

Runs the BOM price up, for one thing.

Can't read app notes and data sheets? You know, the stuff that I furnished and you snipped.

The second stage of kilowatt-level offline switchers are usually square wave forward switchers.

Here's a low power one I did a while back. Built a thousand or so by now.

Works fine. The inductors are ferrite bead EMI filters, but the inverter proper is pure square wave, with just a tiny amount of non-overlap timing.

Why do you insist on being wrong?