Will a mercury relay blow up?

Peter Walsh wrote 12/27/2016 at 02:24 • 1 point

My hackerspace was throwing out a bunch of capacitors, so I snagged them and now I have a capacitor bank to play with. It holds something like 1700 joules at 200 volts.

I'd like to make a wire vaporizer, and I've been trying to come up with a switch that can handle the enormous instantaneous current that this bank represents. I've already burnt out a 500 amp thyristor in trying.

Would a mercury displacement switch work for this? Such as this one:

http://www.newark.com/durakool/1035a24ac/relay-mercury-spst-no-600vac-35a/dp/24C1554?ost=1035a&DM_PersistentCookieCreated=true&categoryId=800000005435&searchView=table&iscrfnonsku=false

The vaporizer will only exceed the load of the relay for an instant - the *average* power will be much less than the rated relay current, but the *instantaneous* power is much, much bigger.

This would vaporize and ionize some of the mercury in the relay, but I'm thinking that the mercury displacement design, which is encased a steel tube with non-oxidizing atmosphere, wouldn't be damaged by doing this, and the steel tube could withstand the pressure.

Anyone have experience with these switches and could comment?

Will this work?

Discussions

Jim Thompson wrote 02/01/2017 at 12:26

I can't comment on mercury relays, but about the thyristors: in a pulsed current application, the limiting factor is the bond wires between the die and the package, which is where your original device probably failed. So even using the same part in a beefier package will give you higher pulsed current tolerance. My suggestion would be multiple thyristers in parallel.

Are you sure? yes | no

Peter Walsh wrote 02/01/2017 at 23:12

The mercury relay has arrived, so I think I'll try that first. Someone else burned out 2 hockey pucks, so I think trying the mercury relay is the right next step.

I'm planning on rebuilding the capacitor box to have space to contain the relay (as well as the capacitors) in case it blows up. It *probably* won't do that, because the core is a steel tube, but you never know.

In theory the average power dissipation of the switch will be minimal.

I'll report back once I've tested it.

Are you sure? yes | no

andres ortiz wrote 01/11/2017 at 20:51

use a hockey puck style SCR, they go for cheap on ebay

Are you sure? yes | no

Peter Walsh wrote 01/11/2017 at 21:18

I appreciate the reply, but are you certain that these would work?

I ask because a mercury relay goes for about the same price on eBay, and another hacker online burned out two hockey-puck SCRs trying to do exactly what I'm trying to do.

(The hockey pucks on eBay are only about twice as much current as the really beefy (500A) thyristor that I've already burned out.)

If you have experience with these than I'll go with your solution, but if not I have to assume that hockey pucks will burn out as well.

Are you sure? yes | no

andres ortiz wrote 01/11/2017 at 21:22

http://www.powerlabs.org/gaussgun.htm this guy's work has some good info on pulsed power applications. a quick read through the site should guide you in the right direction.

Are you sure? yes | no

andres ortiz wrote 01/11/2017 at 21:26

also this http://www.electricstuff.co.uk/destructotron.html

Are you sure? yes | no

Peter Walsh wrote 01/11/2017 at 21:31

Ha! Half the battle in these projects is figuring out the right search terms to use.

Now I can look for "surge generator" and see what other people have done.

Thanks!

Are you sure? yes | no

andres ortiz wrote 01/11/2017 at 21:38

glad to help

Are you sure? yes | no

Mars wrote 01/05/2017 at 06:28

With all the RoHS craziness, I'm surprised they're even allowed to sell those anymore. Whatever you do, do it outside. The minimum order if 5 for $100 each, so have fun. I can think of a lot of funner projects for $500 though.

Are you sure? yes | no

.Stack

Will a mercury relay blow up?

Discussions

Become a Hackaday.io Member