Developing HV DC Pulses

images.google.com has a load of examples of "HV wiring corona ring" and some of 'em are just twists of wire.

At irregularities (like splices or joints) the field adjacent to the pointy is much lessened if you put the same HV on an encircling tube or spring-like item

At 20 kV, outer diameter of uninsulated wire bits should be large (8mm?) like auto spark plug wires' insulation. It's OK if all the current flows in a tiny wire inside. As long as the outer shielding loop (or cylinder, ring, ball) is connected to the same HV as the wire, there's zero field in the gap (and thus, at the pointy places).

Kapton tape has much better dielectric strength and it comes with adhesive as well! John

My son caught my mistake with a 4 output voltage divider using 4kV and 8kV. I was thinking,

1kV, 2kV, 3kv, and 4kv, and second test at 8kV output 5kV, 6kV, 7kV and 8kV., But the second test is 2kV, 4kV, 6kV and 8kV, not what we want. So he wants 8 outputs, 1kV thru 8kV in 1kV steps. He wants to use banana plugs for quick change. Can we get away with this? Would it help to cover the Resistor divider with a few layers of Kaptan tape? Thanks, Mikek

Amazon has these HV banana jacks and plugs, seller says rated for 10kV.

HV cable should fit inside of them, seems like a good idea!

Looking at Kaptan tape, I find 7500V per mil, if I cover the voltage divider with a couple of layers of Kaptan Tape will that prevent corona? Thanks, Mikek

Put a row of banana jacks on a piece of plastic. Solder the resistors between the jacks the obvious way. Ordinary banana leads can run out to the cells. Ignore corona. It's only an experiment and it's only

8KV.

Amazon has great banana jack assortments.

These are cool on PC boards too.

Gross overkill at 8KV.

As I keep telling my engineers, keep it simple and get it done.

John, I'll take that under advisement, however, there could come a point in experiments, that I add a second transformer and go up to 18kV or even 30kV. When we go to a larger diameter vessel, the electric field per cm goes down, to compensate our input voltage needs to go up. I'm now hopefully going to start spending someone else's money on parts, so I have less concern. :-) I'd like to build for possible future developments. What about kaptan tape over the Resistive divider, does that help prevent corona? Mikek

P.S. adding in my ignorance about HV projects, I'd rather err on the side of gross overkill!

If you do simple stuff fast, you'll learn more.

søndag den 14. maj 2023 kl. 16.04.00 UTC+2 skrev John Larkin:

too expensive, too late, or .. good enough

Our business is a struggle against the seductive force of complexity.

Arrgh ! :-) Mikek

But will you learn enough? Simple stuff is a way of getting started, but to stay competitive your products have do more than the competition, or become cheaper, and you need the best product, which isn't necessarily the simplest.

In this thread John has been touting the high-voltage valve/tube which isn't all that simple to use, though it may look that way to him.

Bill, I want your input, because I'll soon start putting the high-voltage valve/tube pulser together, I'd rather get it right the first time.

Also, on the oil/water emulsion, Do you have a thought about about the relationship between the field strength used and the rate of separation? Is it linear with drive voltage? Is there a voltage where we would get diminishing returns? Also, I got the idea from the literature that it is the water that is polar, not oils, so we are only moving water bubbles, and bigger bubbles move farther under the electric field, colliding with smaller bubbles that move less. Does that sound right? Thanks, Mikek

P.S So far we have found some of the feedstock is harder to separate then others using an electric field.

Needn't be a single 10kV 1pF cap - you could have ten 1kV rated 10pF in series or five 2kV 5pF etc etc

piglet

You won't. With luck you may learn enough to get something together that will work, but that's as much as I'd hope for.

If the droplets are really are small enough to form an emulsion the charge on each droplet should pretty uniform,and the rate of separation should be more or less linear with field. You probably won't get any kind of space charge in the emulsion as the droplets move towards the electrode - the charge is split between the droplet and the surrounding fluid and that averages out.

Probably.

When some other sort of conduction mechanism sets in, but that would probably end up as some kind of arc, so you'd likely notice.

Bigger droplets may distort into a more extended spindle/elliptical shape under the electric field, so you would expect them to move faster, but also to break up into smaller, more spherical bubbles

Watch out for surfactants - soaps - in the feedstock. That changes the charged double layer around the droplet that creates the charge in the first place, and messes up the surface tension that makes the droplets sort of spherical, which will changes the way the droplets will distort while being dragged along by the electric field.

That's disheartening! Maybe you could post a circuit you think would get the job done.

There is data on small bubble chains forming (a detriment), but I will need to read again to find out why this happens.

The stubborn oils have got better results with de-emulsifier added. I'll need to look up how de-emulsiers work.

Mikek

Not really; Kapton is a strong material, the tape is circa 0.001" thick and that's not adding much to the diameter. Large-diameter wires inhibits corona (if you have some coax cable, use the shield because that has the large diameter). For a small apparatus, it's probably not worth buying rated HV cable, just keep fingers and such away from the HV parts. Any insulation OVER the coax shield is unlikely to hold off HV for handling purposes; some experimenters use transparent PVC tube (Tygon) when they need insulation; the vacuum-rated tube is thick-walled.

Putting bits of brass tubing over splice points is better than nothing.

If you smell ozone, something needs correcting.

It's merely an opinion.

Not one that you would find easy to put together, because I've got no clear idea of what you could put together, and what you have to hand to put it together from.

I've worked in university laboratories and in commercial electronic development. It's fairly obvious that your set-up doesn't look much like either.

Surface tension at the interface between body of droplet and surrounding fluid means that any non-spherical shape will tend to become more spherical.For can extended shape that can be breaking up into a string of smaller, more-nearly spherical droplets.

Good idea.