1kV 2A DC thermal cutout switch

I've never seen a bi-metal switch rated that high (DC ). Maybe a Kilovac switch driven by a Thermik PTC?

Cheers

Yep, an easy task for a MOSFET and a few other parts.

A high-side HV P-Channel with conventional thermal disc pulling the gate low through some R would be easy (though I'm not sure if P-Channels go that high). Or something like that.

But it's a safety issue, I just want to buy something certified and not have to worry.

Cheers

Knowing how they work I can't imagine you'd find one.

NT

Vacuum relay, perhaps?

Thanks for that, you're probably right.

But that's why I asked here - if the answer were in the Farnell catalogue I probably would have found it. It's outwith my experience, but for all I know, there are some fields of expertise where these things are routine. I've already learned some things thanks to other respondents.

Cheers

Winfield Hill wrote in news: snipped-for-privacy@drn.newsguy.com:

One most certainly wants to go with a solid state solution, as HV relays (the little PCB or inline mountable jobs) do not last long at those current levels. Even at low currents we always suggested that switching take place on non-energized circuitry.

We had one customer that wanted an HV supply the used relays to invert polarity (no joke) instead of simply having us make a dual supply for them. We specifically told them (and disclaimed on it) that the unit MUST be switched ONLY when NOT energized.

Clive Arthur wrote in news:pseqpb$cck$1@dont- email.me:

gate

not

They are out there, but good luck attaining a high use/cycle count before a failure mode *happens*... and one will.

whit3rd wrote in news:d0bced4f-0725-4d94-8b82- snipped-for-privacy@googlegroups.com:

They do not last very long switched live.

You would need contact arc suppression for a relay solution, big and expensive, but why can't you switch the low voltage side of the hv generator ?. I wouldn't trust any solid state solution at that voltage and energy level...

If it's a safety issue, you might need a thyristor crowbar or similar and dump resistor to lose the stored energy fast as well...

Chris

It has to be n-channel. Likely with isolated gate drive, analogous to unipolar version of AoE III, Figure 3.107.A

At 1KV/2A, safety will involve physical interlocks that have nothing to do with local electronic limiters. This is thermal protection - you're better off inhibiting the source of this output, rather than trying to interrupt the output itself. Thermal sensors need have no direct electrical contact with the hazardous voltage.

RL