help driving relay

Many SSR's don't like highly inductive loads (like your relay coil). High dv/dt (back emf) can make the SSR self-trigger. RC snubber networks are relatively common in this application. Your SSR's maker should have some application notes to help you sort it out.

You'd do better with an SCR-based SSR (or possibly an alternistor type). These types are more tolerant of switching inductive loads, particularly the "off" part of switching.

Kudos for mentioning that it's a high power relay (or contactor) and thus we can pretty much eliminate the possibility of the few mA of leakage through the snubber holding the relay on.

Best regards, Spehro Pefhany

I think the cause of the problem is the highly inductive nature of the load. Because of that, the load current is not in phase with the voltage, so at the moment of zero current (when the SSR tries to switch off) there is considerable line voltage that is being bucked by the inductive rate of change of current produced voltage. But at switch off, the rate of change of current snaps to zero, allowing the line voltage to appear, essentially instantaneously across the SSR. This exceeds its dv/dt limit, and the voltage swing capacitively couples into the SSR device gates, switching it back on.

Adding a series RC across the SSR output will provide a path for a little current at that moment, to greatly reduce the dv/dt. You may get by with something like 100 to 470 ohms 1/2 watt and .1uF, but the capacitor needs to be rated for across the 240 volt line, since it sees this all the time the SSR is off. E.g.

Prefab (resistor potted with capacitor in single case) snubber networks like this are also available.

How are you measuring the off state, with a high impedance meter and no load? The SSR leakage will show the full voltage across the meter. Almost any kind of SSR will handle a relay coil drive, why don't you measure it with the relay in place, the meter should show much less than

This method does potentially aggravate the issue of transients on the line. You can always put supression in both places (across the load and across the line), but you have to deal with the unknown (and possibly low) impedance of whatever is creating those spikes, rather than the usually known (and relatively high) impedance of the load.

No, but if it's triggered by the dv/dt or voltage of a spike it won't likely be at the zero crossing.

On the plus side, for this method is that a failure of the snubber tends to put the load in a safer condition. Depending.

Best regards, Spehro Pefhany

Question of the week:

Why don't people put the snubber acroos the load to solve the snubber leakage problem? The turnoff dV/dT would be the same.

I did it - solved most problems with small solenoids, and I never had any problem with powerspikes triggering the SSR's. When zero-voltage switching at turnon the inrush current is no problem either.

Regards, Arie de Muynck