I'm building a small circuit for measuring high(ish) voltages with an oscilloscope. It's based on an isolation amplifier that measures the voltage across a measurement resistor (r3 below), attenuated by dividing with R1 or R1+R2, depending on the range selection. I've scribbled down a picture of the relevant part of the circuit below.

I'm planning to use a toggle switch for SW1, but notice that they have maximum AC and DC voltage ratings. Now, I understand the difference between the two ratings (arcs self-quench more easily with AC), but I'm not sure what properties I need from SW1 for my example.

Specifically, R1 is a series chain of resistors adding to about 450kOhm and R2 is much larger. I want to be able to measure 700V pk-to-pk AC (British line voltage if the probes are out of phase). I'm more ambivalent about DC ratings: I don't think I have a 500V DC source lying around... (Incidentally, I'll be using a panel-mount switch for SW1, so the terminals on the PCB can be well separated.)

The question (finally) is what ratings I need for SW1. I assume I'll need ≥1kV insulation strength to protect the user, but how do I characterise the AC/DC voltage rating? I mean, if SW1 is fully open, it will have almost the entire input voltage across it (because R2 >> R1), but R1 = 450k means that, even if V is 1kV, the maximum current that can flow is about 2mA: I assume that arcing won't be a huge problem...

However, I'm not sure how to express that in terms of switch ratings: can anyone tell me what I'm missing?


simulate this circuit – Schematic created using CircuitLab

  • \$\begingroup\$ Arcing is voltage-dependent, regardless of current. \$\endgroup\$ – Ignacio Vazquez-Abrams Feb 10 '15 at 22:59
  • \$\begingroup\$ Can you expand on that? I mean, I don't think I can sustain an arc with zero current... \$\endgroup\$ – Rupert Swarbrick Feb 10 '15 at 23:01
  • \$\begingroup\$ He means that an arc will develop if there's enough voltage difference. The fact that R1 can provide little current has no effect. An arc occurs if electrons find a path of lesser resistance in the air, the max current you can provide to the arc once it's formed is not important. \$\endgroup\$ – Vladimir Cravero Feb 10 '15 at 23:06
  • \$\begingroup\$ Oh, ok that makes sense. So do people make switches with high DC ratings that aren't designed to switch multiple amps of current? (I can't find them on Farnell, but maybe I'm not looking for the right thing) \$\endgroup\$ – Rupert Swarbrick Feb 10 '15 at 23:09
  • \$\begingroup\$ Make sure your resistors are rated for that voltage also. \$\endgroup\$ – George Herold Feb 11 '15 at 14:26

You should use a switch rated for at least the max voltage and current you expect to see (700VAC and 500VDC in this case). Using a relay is probably the way to go to isolate from the higher voltage. Having a higher current rating shouldn't matter, but here are some options that look like they meet your spec w/ relatively low current ratings:

http://www.digikey.com/product-detail/en/DAR70510/725-1081-ND/751985 http://www.digikey.com/product-detail/en/DAR70510S/725-1202-ND/2811118

| improve this answer | |
  • \$\begingroup\$ I'm confused. Why should a relay provide the user more protection than a switch? Aren't they fundamentally the same thing? \$\endgroup\$ – Rupert Swarbrick Feb 11 '15 at 21:31
  • \$\begingroup\$ A relay electrically isolates the switching circuit from the higher voltage circuit. link \$\endgroup\$ – shoul25 Feb 12 '15 at 7:25
  • \$\begingroup\$ Ok, so I understand that if I use a switch to control a relay then I have two sets of isolation between the high voltage and my fingers. But I contend that this is the only difference - I don't propose to use a switch whose actuator is electrically connected to the input! \$\endgroup\$ – Rupert Swarbrick Feb 12 '15 at 19:18
  • \$\begingroup\$ It's an added layer of protection and allows you to avoid running high voltage wires out to a panel or some other place (which could increase risk of arcing if not done properly). This is why it tends to be easier to find HV switches that are relays. Did the above help clarify your question about specs? \$\endgroup\$ – shoul25 Feb 12 '15 at 23:09
  • \$\begingroup\$ That makes sense, thanks: I think I'll drive a relay from a switch as you suggest, but much smaller and more feeble than the ones you linked to. If it fails, it won't be a safety hazard: it just... won't work. So I'll admonish users (mostly me) not to switch it when there's a high voltage input connected. \$\endgroup\$ – Rupert Swarbrick Feb 13 '15 at 0:00

At these voltages, using a panel-mounted switch is suicide. The likelihood of voltage getting into the toggle is just too high. Use either a high-voltage relay, or a remotely-mounted switch with the toggle activated by a non-conductive link such as plexiglass.

| improve this answer | |
  • \$\begingroup\$ Well, "these voltages" to which you refer are just British line voltage. I have many home appliances devoid of plexiglass in their switches... And I take issue with the "suicide". If a switch fails short and the human has zero resistance then... 2mA. Similar to walking into a cattle fence, I think. \$\endgroup\$ – Rupert Swarbrick Feb 11 '15 at 21:34

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.