.Snubbers are placed across the contacts in situations where producer/designer has no influence on what the customer will do. Naturally it's better when they're placed across the inductive load - thus limiting RF emission by connecting wires. Malfunction risk is not any greater for the near-contacts placement and modern mains voltage X2 class caps have overvoltage specs at 2.5kV. Varistors limiting back EMF protect the caps, as well.ncc1502 wrote:Putting a VDR of RC cirquit parallel to the contact does (in my opinion) not help, it might even introduce a risk. When VDR or RC cirquit mailfunction you might bridge the contact and turn on the big relais unintentionaly.
rainer wrote:Controlling mechanical parts this way is the craziest idea I ever heared of.
The idea isn't mine but I used it for single relay protection (the controller circuit was immune to arcing) as customer tended to overburden the relay - and this solution worked - no more burned contacts. I've checked and indeed, I've used voltage zero-crossing with ca. 7ms advance switching (obviously adjusted for specific inductive load). Unfortunately, the solution with advance switching is not applicable for cascade of relays.Malcolm_M wrote:My controller has two internal relays, driving two external slave relays which drive two switchgear
Solid state relays are indeed used for switching larger currents but there are optocouplers for lower loads (as I've sugested before), like Fairchild's MOC3163M (zero-crossing turn-on of inductive load is not dangerous for triac at small (<100mA) currents). There are also Toshiba's TLP3023 or TLP3063 but with worse dV/dt parameters.Malcolm_M wrote:The external slave relay coil current is only 8mA which is less than the Minimum Load Current of any solid-state relays I have looked at ? :
What Rainer didn't mention is that triacs also need snubbers for inductive loads (though for different reason which is sensitivity to commutating dV/dt). Most ac solid state relays have the snubber built-in, but not small-current switching optocouplers. A snubber consisting of 1nF cap in series with 100 Ohm resistor should be enough for your small relay. MOC3163M may even work properly without the snubber at 8mA load (a 400V varistor should be left for protection, though).
Still, from your description, it doesn't seem that controlling the intermediate (slave) relays is the main problem. If you cannot introduce changes to the rest of the circuitry, then the only solution is to make your controller immune to external disturbance (like the old controllers that worked in these conditions
).
