High Voltage and Physical Contact Switching

[joker12]

Is there a problem with using a contact switch, such as a relay, to apply and remove high voltage? Maybe there is some destructive arcing?

At what voltage is relay switching not a good idea? I haven't installed a house wall switch, but I believe those are directly switching the 120V. I want to switch 500 VDC with a very short current peak less than 100mA. I've been thinking that a physical switch would not be acceptable, but maybe it is okay.

 

[IanP]

Look for high-voltage relays, otherwise you will face arching problem which, by using conventional relays, is not possible to avoid ..

**broken link removed**

Other option for voltages below 1kVdc includes HV-SSR (solid state relay) ..
**broken link removed**

Rgds,
IanP

 

[joker12]

Thanks for the suggestions. Why is the unavoidable arcing bad? Maybe it could weld the contacts together?

 

[IanP]

Here is a good article on this issue:
*Contact Arc Phenomenon*
http://relays.tycoelectronics.com/appnotes/app_pdfs/13c3203.pdf

When an arc ignites between separating contacts, it will be sustained as
long as there is sufficient energy to feed it. As long as the arc exists,
material transfer will continue. In a direct current application, the arc can
be extinguished only by stretching it to such a length that its own
impedance causes it to extinguish, or by opening the circuit at some
other point. In many applications, though, the contact gap is wide enough
that the arc will extinguish before the contacts have fully opened. It is
for this reason that relays of a given contact rating will be rated for, say,
120 volts AC, but will have a considerably lower DC voltage rating—
usually 28 or 30 volts DC.

:!:

 

[joker12]

I see -- the non-extinguishing arc.
I think my arc will quickly extinguish, because there will be a capacitor between the high voltage and ground. As you know, capacitors are open circuits in DC.

 

[IanP]

That may all happen as you predict, but after several actions of this type the contacts will be damaged ..

A more universal form of arc suppression than just a capacitor connected between (+) and (-) poles is a resistor and capacitor in series connected across the contacts to be protected ..
The capacitor provides a temporary alternate path around the contacts until a contact gap is well established while the resistor is necessary to limit the capacitor discharge peak current when the contacts are reclosed and, therefore, should be in the order of 1 and 1/2 ohm per volt of source voltage to limit the current transient at contact closure to max a couple of amps ..

Rgds,
IanP

 

[joker12]

Yes, that is an interesting idea. Although, I'm not sure how or if the cap would eliminated a possible arc. There is still the same voltage between the contacts, right?

I went ahead and tried my idea today. I used a TQ2-4.5 relay and a 100pF. The relay is rated at 2A, 32V. The cap was a regular ceramic cap that probably was not rated for high voltage. Despite the ratings, everything worked perfectly.

I switched the cap from +400V to -400V numerous times. There was a series resistor in the 10s of mega-ohms, so the RC time constant was on the order of milliseconds. That's much slower than what the real circuit will do. I could not hear any arcing over the low noise of the relay.

From **broken link removed**

Voltages below about 500–700 volts cannot produce easily visible sparks or glows in air at atmospheric pressure, so by this rule these voltages are "low".

 

[IanP]

I'm not sure how or if the cap would eliminated a possible arc.

.. The capacitor provides a temporary alternate path around the contacts until a contact gap is well established ..

Voltages below about 500–700 volts cannot produce easily visible sparks or glows in air at atmospheric pressure, so by this rule these voltages are "low".

Not by itself, but opening a contact is a different story ..
Do you know how high (or how low) is the voltage used in a dc arc welding apparatus?

Rgds,
IanP

 

[joker12]

I see, you're only talking about opening a switch. Now I understand that a cap across the switch would work. I don't think it works for closing the switch -- that's was I was thinking about. You were right to think about the opening because that's when the non-extinguishing arc could occur.

However, the non-extinguishing arc does not occur with my setup. I think any spark that does exist is tiny.

DC welding can be done with 12V or probably even lower, but it requires hundreds of amps and they do not put a cap in series. The power supply that I used can only supply 4mA at 500V.

 

[IanP]

However, the non-extinguishing arc does not occur with my setup. I think any spark that does exist is tiny.

OK.