Page 1 of 1

Safety shorting bar for HV supply

Posted: Wed Oct 07, 2015 6:11 pm
by Paul W Fontana
In designing my HV supply, I would of course like to design safety in first. One feature I would like it to have is a shorting bar that will automatically short the output capacitors in the event that power is lost or a panic button is hit, or perhaps any time the HV is not enabled. I'm thinking of an electromechanical solenoid relay, and was originally thinking of something like the Ross switches here:
https://www.surplussales.com/Relays/RossRelays.html

or here:
http://www.hvproducts.de/en/hv-active-c ... relay.html.

The solenoid would hold the switch open while there's power to it, allowing power to the HV circuit, and when power to the relay is cut the HV power would be cut and the shorting bar would fall across the output (possibly at a point located after the ballast resistor or a ballast inductor to minimize sudden HV discharge).

But these are very expensive, and appear to be designed for high current switching, which I don't need. Anyone know of a more reasonably priced solution that will be reliable and safe? What is the common practice, if there is one?

Thanks for your input! -- pwf

Re: Safety shorting bar for HV supply

Posted: Wed Oct 07, 2015 8:39 pm
by Jerry Biehler
What you are referring to is typically referred to as a "crowbar" circuit.

I would get a used vacuum relay off ebay. http://www.ebay.com/sch/i.html?_from=R4 ... y&_sacat=0

Re: Safety shorting bar for HV supply

Posted: Thu Oct 08, 2015 1:30 pm
by Paul W Fontana
@Jerry: Thanks for doing some eBay research for me. It's pretty hard to tell from most of those listings, but they don't look like they are rated for many 10's of kV. (Also, most are single throw, so I'd need two, but if they're inexpensive that would be OK.)

I did find a nice Ross relay (not vacuum) rated at 25kV, which looks like it will work fine up to much higher voltage. The price is very reasonable. Unfortunately its coil is 230V, and I'll be using 115V in my supply. I'll have to get a cheap step-up transformer as well, but for the price of this relay it will be worth it. (Any idea whether relay actuators tend to work if you undervolt them?)

(BTW, we may have different usage - to me a crowbar circuit refers to a circuit that prevents overvoltage events, like a surge suppressor. That's not really what I'm after here, except maybe to prevent a surge when the shorting relay is thrown. At these low currents, though (below 20 mA) I think an appropriately sized bleeder resistor in series with the shorting bar will do fine. I figure if the capacitors discharge in a few ms it will be OK.)

-- pwf

Re: Safety shorting bar for HV supply

Posted: Thu Oct 08, 2015 11:46 pm
by Rich Feldman
Paul W Fontana wrote:... Unfortunately its coil is 230V, and I'll be using 115V in my supply. I'll have to get a cheap step-up transformer as well, but for the price of this relay it will be worth it. (Any idea whether relay actuators tend to work if you undervolt them?)
Often the hold voltage is much less than the operate voltage. Characterize your new relay using a variac to power the coil. If your max V isn't enough for the armature to pull in, try operating the armature by hand and see if the coil has enough voltage to hold it. If 115 V is marginal, you might try a boost transformer configuration. For example if you have a little 115:24 transformer you can get 139 volts.
Paul W Fontana wrote:... I think an appropriately sized bleeder resistor in series with the shorting bar will do fine. I figure if the capacitors discharge in a few ms it will be OK.)
A water resistor might be good for this, since the tolerable range of resistance is very wide. I can attest that a few fluid ounces of copper sulfate solution w/ copper sheet electrodes can quietly (!) absorb many shots of 1000+ joules from a HV capacitor. Terminals must be far enough apart to not flash over externally when HV switch is closed.

Re: Safety shorting bar for HV supply

Posted: Fri Oct 09, 2015 12:14 am
by Jerry Biehler
OK, here you go, 60kv at 100A. That should handle any stored energy from caps:

http://www.ebay.com/itm/Jennings-R1G430 ... 419d851466

Re: Safety shorting bar for HV supply

Posted: Fri Oct 09, 2015 5:58 pm
by Paul W Fontana
Jerry Biehler wrote:OK, here you go, 60kv at 100A.
Wow, that's a beauty! And the price isn't bad. I'll have to consider it vis. the Ross switch I found, which is more compact and about half the price, and also DPDT, but only rated to 25kV. Thanks again for shopping for me.
Rich Feldman wrote:A water resistor might be good for this,
Intriguing! What do you make yours out of? I think it might be overkill for this application, but now my gears are turning about making water resistors for the dummy load we've been discussing on my other thread viewtopic.php?f=11&t=10332.

For this shorting bar, I thought maybe a 100k ohm 20W ceramic, or maybe a couple of them, would be sufficient. Or, I would just wire the shorting bar after my 50k ballast resistor, which would be safer anyway. If the shorting bar got slammed down when the supply was at full voltage the power would spike above the rating briefly, but not for very long.

Re: Safety shorting bar for HV supply

Posted: Sat Oct 10, 2015 6:32 pm
by Rich Feldman
Sure you can use a water resistor as a dummy load. You can learn a lot from, uh, searching fusor.net and the Internet. These two links, to generally respectable references, just leaped out from a search result: http://home.earthlink.net/~jimlux/hv/rwater.htm http://www.kronjaeger.com/hv/hv/comp/res/index.html

Electrolytes have a large negative temperature coefficient of resistance, so you'll need to measure both voltage and current independently of the load resistor.

Mine was made for "quickly" dumping kilojoules from a big HV capacitor. Small plastic bottle, dilute CuSO4, sheet copper electrodes. Threaded copper feedthroughs sealed with clear RTV silicone. Today I found the fluid barely covering the plates. Added water to level in this old (2006) photo, before measuring 1.2 amps at 120 volts AC. In a while 1.3, 1.4, 1.5 amps, and a warm bottle.
waterr.JPG
waterr.JPG (10.39 KiB) Viewed 11028 times
For your dummy load voltage and resistance targets, how about a long U-tube or vertical helix of plastic hose or PVC pipe? I think electrolysis from a few mA of DC current won't be hard to manage. Consider a cooling fan.

p.s. Your transformer and rectifier ran an x-ray generator with no HV capacitor, and can do the same for a fusor. Skipping the capacitor might also obviate the HV crowbar relay and load in your emergency-off circuit.

[edit] The transformer's current capacity must be derated if you add enough capacitance to smooth the rectified output. For a given average current, the rectifiers will then conduct only briefly near max voltage. So RMS current and coil heating will be significantly higher than with unsmoothed voltage waveform to a resistive load. Measure DC resistance of your primaries and secondaries to get a handle on the heating power. In your favor, operation at 50% to 70% of design voltage will probably _greatly_ reduce the magnetizing current and core loss.

Re: Safety shorting bar for HV supply

Posted: Tue Oct 13, 2015 12:31 pm
by Paul W Fontana
Well, I ended up getting this one:
http://www.ebay.com/itm/NEW-Ross-Engine ... 7675.l2557

The original listing said there were two available, so if anyone's interested you may see one show up on eBay again. It's only rated to 25kV, and I checked the dimensions - some gaps are as small as 25mm. (The terminals are spaced 1.5" center-to-center on each pole and look to be about 3/8" diam, and the bar itself is another 1/8" thick. There's also only 1" between the tips of the terminals of the two poles.) But everything metal is pretty blunt, so for the price I'm going to take a chance and try it out. With dry air breakdown field at 30kV/cm it should be fine, but then this is Seattle...
Rich Feldman wrote:For your dummy load voltage and resistance targets, how about a long U-tube or vertical helix of plastic hose or PVC pipe?
Yes, that's what I was thinking. I estimate that a 1" diam. x 50cm vinyl tube should dissipate more than the necessary 250W without boiling, and I should be able to find a CuSO4 concentration that would give 2.5M ohm. I wonder if I could use copper hose barbs, epoxy or caulk the through hole, and cap the tube with them so they'll work as terminals? Would cable ties make a tight enough seal, since I wouldn't want to use metal hose clamps?

Another thought was to use a bottle or jar like yours, but use a rod or copper tube through the cap as one terminal and a copper plate or foil in the inside wall as the other. That way I could adjust the resistance by raising and lowering the center rod.

But one of those sites mentioned electrolysis, with makes me a little nervous about safety. I think I might just keep it simple for now and stick with commercial ceramic resistors. (But we've drifted a bit off this thread and should probably take it back over to the other one on dummy loads.)

Thanks to all as always for your input! It's a great help to have folks to talk these things through with.

-- pwf

Re: Safety shorting bar for HV supply

Posted: Wed Oct 14, 2015 12:58 pm
by Richard Hull
Electrolysis is a function of current. This will occur minimally, of course with a few milliamp s of current. Copper sulfate electrolyte and copper electrodes would involve no real plating of signifcance and a simple side arm tube open to air would allow the escape of any tiny amount of gases. Serious electrolysis, gas generation and plating need amps of current.

Richard Hull

Re: Safety shorting bar for HV supply

Posted: Sat Oct 17, 2015 12:47 pm
by Jerry Biehler
Paul W Fontana wrote:Well, I ended up getting this one:
http://www.ebay.com/itm/NEW-Ross-Engine ... 7675.l2557
-- pwf
If all else fails you can always submerge the business end in some dielectric fluid. Something like mineral oil or galden. That will stop any corona discharge too.