Using a low-voltage switch in a high voltage circuit

Question

I have an application that requires switching a circuit between the following two configurations. Circuit Lab is infuriating as it does not allow you to draw diagonal wires, so I used arrows instead. They are not connected in the middle.

^{simulate this circuit – Schematic created using CircuitLab}

I have thought about using a DPDT NO NO switch to swap between the states but the switch is only rated for 12 volts dc, and I want my circuit to be operable up to 170 volts.

Because the resistors will absorb almost 100% of the voltage drop (assuming the resistance of the switch is negligible), I am assuming that I will not have power or current-rated problems considering that the whole setup will be shipped with a 1/4W power rating. The problem I will likely encounter has to do with potential breakdown of the switch from arcing and the like.

I would not hesitate to insert a SPST switch in series with two large resistors at any voltage, as the voltage drop across it would be almost nothing. However, in my setup the DPDT switch would have the full 170v potential across the two poles. If I use a small switch in this configuration, would I be safe or would my switch arc/breakdown across the poles?

Answer 1

The voltage rating of a switch is precisely about what how much of an arc it can break while it opens, and how much voltage it can withstand when open. Your switch must be rated for the maximum voltage that may appear across it.

This is true of any single pole of a switch, so it doesn't matter whether the switches you are using are SPST or DPDT. In principle, the pole-to-pole isolation could have a different rating than the contacts, but I've never heard of that being done for ordinary switches. Another place the voltage rating matters is the isolation between the contacts and the case and handle/button.

In your case, the maximum voltage appears to be 2/3 of the voltage applied to the ends of the circuit, or 114 V, since R1, R2 & R3 operate as a voltage divider, but I'd want to choose a switch that can handle the full voltage in case something goes wrong elsewhere in the circuit. (Presumably your "R"s are actually loads more complex than a resistor on a circuit board, and so there's more opportunity for failures such as short circuits across any given R, increasing the voltage applied to the switch.)

There's another complication: switches have different ratings for DC and AC. If you apply DC, a switch with an inadequate DC rating is not guaranteed to successfully break the arc, even though it will withstand the voltage while open. High voltage DC switches are a somewhat unusual item (though becoming more relevant in this day of e.g. solar power systems). Apparently there might be electronic strategies to be able to use an only-AC-rated switch with DC.

Answer 2

You can use a relay rated for 240Vac and 30Vdc if you use an RC snubber design to limit V=LdI/dt and Ic=CdV/dt within the linear operating range.

• Since we do not know your circuit or cable impedances , I cannot specify what will work.
• But I am sure a design can be made to work that is well damped and limits AC voltages with 170V transients absorbed in an optimum RC snubber.
• The inductance of the cable must be modelled with the load and your circuit and switch for this design to be verified as an air gap can withstand 1kV/mm.

• If you wanted a snubberless switch , then you could buy ($$) a vacuum sealed relay rated for >= 240Vdc.

• Typical values may be some film cap <= 100 nF and 10 Ω to 1 kΩ in series.

• If current is <=2A, then the contacts should be gold plated, otherwise if >2A then 10% rating transition wetting current. Eg. telephony relays are rated for high voltage , low current and thus are gold plated.

What does a 250Vdc relay look like ?

MKS2PI DC24. OMRON. $16.41 CDN