5kV 40uA from 120VAC

I'm not verygood at ascii art but ... (courier font)

---|(---+------ V out ( | ( - ( ^ ( | --------+------ gnd

Art

You are kidding, right? That produces an offset AC.

Trivial at these low currents to get rid of the ripple:

---|(---+-->|--+- V out ( | | ( - --- ( ^ --- ( | | --------+------+- gnd

...Jim Thompson

DC with a huge ripple. Po-t8-to, po-tah-to. Art

That is not DC, and the OP spec'd DC.

"Why did you put black pane into the windows?"

"Well, you did not specify that you wanted to see through your windows!"

The output voltage never goes negative. Ergo it's not AC. Art

It drops out 60 times per second. Ergo it fails the spec 60 times a second, which is a lot :-)

Excellent point! I sit corrected ..... 5 parts Art

You quoted John Larkin posting both at least one C-W suggestion and at least one other actually achieving such voltage in the transformer - the one he said was "cheap but big".

Do you and the one who followupped you have some sort of hard-on for John Larkin?

One diode and one cap aren't enough for a doubler for DC, which the OP asked for. With only one diode and one cap, the highest DC voltage is across the cap and the diode - these 2 have similar DC voltage value. (Likely less than 5 KV DC, with enough current to electrocute a buffalo.)

The peak voltage across the diode is about twice that, but that's because voltage across the diode varies every cycle from close to zero to close to double the voltage across the cap (assuming the cap has light load or no load).

Not that I think something made to power a 600-watt or 1100-watt output magnetron is 1st choice to supply 40 microamps of high voltage DC. I'd prefer a voltage multiplier applied to a cheap item that can be powered by a wallwart, that has low electrocution risk. (CCFL inverter followed by a quintupler or whatever.)

The DC component is still limited to about 1.4 times the AC voltage from the transformer. Over/under that is ripple.

It's a combo of DC and AC, AKA DC with a great amount of ripple. The DC component is only about 1.4 times the AC voltage from the transformer.

message

If the transformer is a microwave oven one, I think the 2 extra parts work to meet the spec - DC voltage of at least 5 KV. I would not call

============================================

Meanwhile, I still like the idea of a wallwart-powered CCFL inverter feeding a quintupler or whatever other voltage multiplier, whether or not one of Cockroft-Walton configuration.

I give that fair chance (not completely guaranteed) of only "teaching a lesson" if a greenhorn makes a bad move while building or testing this.

As opposed to microwave oven transformers likely capable of killing a hippo...

If you want a bunch or to salvage onesey-twosey, NG ignition transformers are pretty inexpensive and fairly compact, and they're designed to work off 120VAC with a thyristor.

A 4800V to 120V potential transformer could be used in reverse with a HV =

rectifier and a capacitor.

If you can generate audio frequency or PWM (even from a computer = switching=20 PSU) you can use an automotive ignition coil:

If you can generate 400Hz you can overdrive a 60Hz transformer to 8x = normal=20 voltage, but insulation breakdown may be a problem unless the = transformer is=20 made for it.

Some HV transformers:

Here is a company that make high voltage DC-DC converters up to 8000 = VDC:

How about 5000VDC at 5mA from a 12VDC source for under $50?

Here is a battery operated insulation tester with 5kVDC output:

Some "outside the box" ideas:

I have some interest in this because I may soon be designing an = insulation=20 tester up to 5kV. So some of the above seem to be useful for that = project.=20 Others are just FYI and interesting but maybe not very practical.

Good luck,

Paul

Dang, what a whiney twerp you have made yourself into. You rarely even try to answer anyone's questions, unless it's to pontificate on Ohm's Law. You brag about stuff you've designed, and offer to show us pictures, and then chicken out. Well, that's what hens are best at.

Here's a little C-W supply I did for my friend Alex, to demo his cool PLZT-based optical attenuator technology.

ftp://jjlarkin.lmi.net/Z206.pdf

ftp://jjlarkin.lmi.net/Z206_pcb.JPG

It's only 1000 volts or so, but the idea is the same.

If one does the math, it turns out to work better to use a lower voltage ratio on the transformer, and then multiply. Can you guess why?

Old Tek tube scopes, and transistor televisions, often used C-W multipliers, for the same reasons.

Idiot.

John

When I was a kid, there was a neon sign company that would give away old transformers. I had a bunch, but my favorite was an 18KV beast with a big glass insulator on each end.

Modern neon signs generally use a high-frequency inverter supply, and they are fairly cheap, around $30. One could rectify the output.

John

Motorcycle ignition coil?

John