Varistors Question

Carlinux

Hi,

I have two AC Relays (from air conditioning unit) 24vac and I would like to minimize the voltage spikes off the coils when closing. The voltage is about 25vac rms on the coil and the spikes are about ~95v.

I would like to try a varistor but after reading many articles I'm more confused on which one I should use to keep the spikes as minimal as possible.

Thanks,
Carl

Carlinux

... my question was more... what voltage and latching voltage the varistor should be?

Carl

Erik Friesen

I would use a tvs. http://www.mouser.com/ProductDetail/Vishay-Semiconductors/P6KE30A-E3-54/?qs=sGAEpiMZZMsYiK5PgaDog840l4f5fhE9PFBGPB0e%2fc8%3d That is rated at 30v. That spike is on the field collapsing, which is when the coil is released.

LoopyByteloose

There are really two parameter to consider with varistors - The voltage at which they kick in AND the amount of power they can dissipate.

Since Watts = Amps x Volts, the higher the voltage, the more overall watts in a similar spike.

I'd divide spikes into two, maybe three groups = automotive systems, A/C mains, and potential lightning strikes.

You may be able to use a lower wattage on the automotive system. A/C mains get spikes both from inductive loads (like a refrigerator) and occasional lightning strikes. Lightning strikes are generally the highest wattage and may make an MOV useless after one event.

If you are considering protection against lightning, it may help to have a 'crowbar' circuit that blows a fuse instead of just being dependent upon an MOV. The 'crowbar' is created by a special diode, somewhat similar to a pair of back to back zener diodes, that shorts the high voltage to ground and blows the fuse. You can still have an MOV in the same circuit for noise AND it is protected as well.

A 95VAC spike on a 24VAC relay coil would likely be best controlled with an MOV nearer to the 95VAC rating.
Why so?

Well you have two wear factors - how many times the MOV is triggered AND how powerful the spike. I'd say an 80VAC MOV would be rated for higher power spikes than a 26VAC MOV. So it would both be used less and would handle more. All relay coils use copper wire with enamel insulation. The insulation's voltage rating far exceeds the working voltage of the relay, so it should easily handle transisents into a few hundred volts. And the transients in this case are originating from the coil, so they should be well below the rating of the insultation.

stamptrol

The best guidance I ever got on this subject was from the Toshiba corporation as it related to protecting industrial control circuits from spikes and arcing.

1. on DC circuits, a reverse connected diode ("free-wheeling diode") across the coil. Diode withstand voltage at least 3 times power supply voltage. Diode current rating larger than load current.

2. On AC or DC circuits with varistor, varistor voltage is twice the peak value of the power supply.

3. On AC or DC circuits, snubber circuit (capacitor and resistor in series) around coil. Resistor is 0.5 to 1 ohm per volt of coil voltage. Capacitor is 0.5 to 1 uF per amp of coil current. Use non-polarity capacitor.

These 3 rules have been proven to just work in any case I've applied them.

Carlinux

Thanks all for your answers and suggestions. I'll do some work on my side and tell you what are my results.

Erik Friesen:
Are TVS Diodes only for DC or do they also work on AC current?

Carl

Erik Friesen

Bidirectional TVS work with ac.

LoopyByteloose

Bidirectional diodes for surge supression usually have a symbol that looks like two zener diodes back to back. The above link referred to a 'unidirectional diode' that is DC only (unless you use two back to back).
''

Carlinux

Wooohooo!!! The TVS Diodes made my day! No more spikes! ..see the attached file.
Thanks to all.
Carl