Developing HV DC Pulses

Trying to solve for self capacitance.

Ahh, great John, I have been looking at grounded grid schematics, but not with your twist. Now, I'm waiting for parts.

I need to start thinking about a filter cap for my DC. I have seven, 1uf 2000V caps from microwave ovens. That's 140nF.

Any reason I can't series them, as a first iteration filter?

Upgrade when if we go to higher voltages. Pretty sure they have built in 1MΩ parallel resistors.

Thanks, Mikek

Bill, great idea, but for later, it could be very useful as so far, it seems different oils need different treatments. Developing that would be time killer for me, I could probably get it done with enough time but for these preliminary tests, I think a 555 will be fine, but it's in the note book. Thanks, Mikek

I thought it was cute to have a single 6-volt power supply, but that nails the positive grid (I keep wanting to say gate) drive at 6 too.

Whit's idea of push-pull drive is cute but needs another power supply, and in its most basic form forces the + and - grid voltages to be equal, although that is probably OK and easily fixed. One 556, the dual 555, could drive both the grid and the cathode.

You should build in the scope pickoff, so you can see what's going on. We usually build in a signal pickoff/monitor circuit on pulsers.

You don't need much filter capacitance, driving a 10 Meg load, and ripple won't matter much here. Equalizing/discharge resistors are a good idea. Safety warning: a neon bulb relaxation oscillator. [1] [2]

I used to get military surplus HV oil caps, big metal cans.

Last-gen CRT-type TVs used to have a potted flyback-rectifier thing, which would be another way to get HV. I'd expect that ebay has some. They could make 25KV or some such, conveniently positive, and are small and light and relatively safe. I think they used the CRT bulb as the filter cap, but the flyback freq was 15K so they didn't need much cap. There are old TV doorknob caps too.

I'm already thinking about two neon sign transformers, and another complete tube switcher, with opposite polarity, so it will ± 7kV for a 14KV differential.

Would we build a negative switcher or to positives and invert the input to one?

Agree on that, Need to source a 1nF HV cap.

I have searched for a Red blinking light, that would power on when the power is turned on. I have couple setting in my Amazon cart, but as yet unsure what to do. Now that filter caps are involved, I need something that will display until the filter caps are discharged. I'm not sure a neon bulb would be enough of an attention getter. Maybe a neon bulb and a phototransistor to trigger the red flasher until the HVDC is diminished :-) But yes safety is of great importance.

That should have said I need to source a HV 1pf capacitor. Mikek

The oil emulsion thing is inherently AC-coupled, no? One supply and one tube does that.

And one neon sign transformer could make +-HV if you really wanted that.

I think Whit's idea is to drive the grid and cathode of the one tube,

My suggested circuit works well making low duty cycle negative HV pulses. Positive-going pulses would be harder.

Make it!

High voltage coax cable tend to work out around 1pF per cm (or 2.5pf per inch) like every other kind of coax cable.

It's electrophoresis. It's actually DC coupled. The droplets of oil suspended in water (or droplets of water suspended in oil) have a small electric charge with respect tot eh surrounding fluid in which they are suspoended - which is how the field moves them, and the suspension gets broken when the droplets get pulled onto the electrodes and the charge flows on into the voltage source generating the bias. <snip>

I was thinking the differential to get higher voltage, in the end I would like to have 30kV, whether that is single sided or ± I don't know what is best.

I don't understand that.

OK, I think that is a logic thing. Since the circuit is acting to bring the output to 0V, you are calling that negative HV pulses?

Ya, I already looked at dielectric breakdown voltages, it seems it takes a thick dielectric.

A 1" long copper pipe with a 0.25" copper pipe in the center is 2pf. If I fill it with silicone sealer (~3) I'm up to 6pf and can shorten it considerably.

Is silicone sealer better than air for breakdown voltage?

Make a 0 to 12 volt positive pulse and apply it to the grid. Invert it, still 0 to +12 but out of phase, and connect that to the cathode. So the grid is usually -12 relative to the cathode and +12 during the pulse. You can do it all with one NE556.

I hate the concept because I didn't invent it.

If a voltage is high positive and occasionally blips to ground, I call that a negative pulse. Or negative-going maybe. Like the OUT here:

That's complex. I was thinking of a variant on the gimmick capacitor, namely two pieces of insulated wire, parallel, with shrink tubing overall. Or two layers on a PC board.

You could hack a small piece of copperclad FR4 too, and add wires. The usual stuff is about 15 pF per square inch. You could make the entire attenuator out of one piece of dremeled FR4!

Maybe I'm being to concerned with arc over. I have made disc caps using Rogers 5880 (0.036"), using a paper punch, they are about 1.3pf. But, I think the 5880 is too thin.

If I use FR4, what is the max voltage you would go? Assumes the FR4 is larger than the copper.

Mikek

Teflon tape is what I'd use. (6000 V/mil according to American Durafilm) Wrap the center conductor with multiple layers so that it fits snugly in the outer pipe. Ed

A quarter-inch square copper patch, 0.062 thick FR4, would be about 1 pF. At, say, 500 volts/mil, that's about 30 KV to the other side.

A clever design could be hacked onto a 1"x2" slab. Maybe go 10,000:1 (discrete 10 nF cap) and include a BNC to go directly to the scope.

You have to worry about corona from the edges, especially if they're square. I'd be looking at ignition cable with a bit of solid wire wrapped round it, a few inches from one end.

Cheers

Phil Hobbs

I have a good bit of HV wire for the neon signs rated at 15kV. I'll make one. Can anyone explain why the wire is so thick in the Neon Sign Transformer wire? The usual max current is 30ma. Mikek

Corona again. The field at the surface goes as 1/r.

Cheers

Phil Hobbs

Press on, next problem. ;)