Snubber capacitor suggestion

[Gokkul55]

I have a relay design, where the relay switches AC 240V/10A (Phase Line). Sometimes during switching the LCD on the board goes blank. Similar problems have been reported on various other forums too.

Suggestion was that the switching of the relay produces spikes and a snubber cap of 0.1uF/1KV will arrest the spike. I haven't had the opportunity to check if it works as the circuit where this happens is not with me right now it is in field and i don't have direct access to it. Will test it when i get my hands on it.

I have attached a pdf of the circuit. (Just the relay part)

Any suggestion or improvement in design is welcome so that i can implement it in future design.

Thanks
Regards
Gokkul

 

[VEBill]

You've already got the diode across the coil. Not the fastest diode option, but it'll have junction capacitance to bridge the gap anyway.

I'm not sure that the added capacitor across the contacts would solve whatever the problem actually is. If the contacts are arcing due to an inductive load, then maybe.

I assume that the load wires are dressed away (routed in accordance with EMI best practice) from the uC signals?

 

[Gokkul55]

Yes the load is inductive and the wires are routed away from uC signals.

What is the best snubber solution to prevent contact arcing. It is this arcing that generates spikes that corrupt the LCD module.

Thanks
Regards
Gokkul

 

[ScottyUK]

Solid state relay. Guaranteed no sparks.

 

[Skogsgurra]

The .1 uF is OK. But you should add a 47 or 100 ohms series resistor to prevent welding of the relay contacts. Do that, and you will have no problems. At least if the inductive load isn't more than about one hundred mA.

If more, then use a varistor instead of the RC combination.
Select a varistor with 275 V rated voltage (230 V is too tight) and large enough to carry the coil current for a few milliseconds. I use Siemens S10K275 for most applications. It can handle most contactor coils easily.

Question: Does the relay really switch 10 A? Or is that the contact rating? If it really switches 10 A, then I would use an S20K275 instead.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[VEBill]

In addition to the other advice, twist the load wiring. Maybe even shield it as a last resort.

If the EMI design is correct, arcs and uC can coexist.

I don't think it's the wee little arc itself, it's probably the noise (generated by the wee little arc) on the load wiring.

 

[Comcokid]

Don't use a ceramic for the capacitor. The dielectrics of non NPO ceramics can overheat when connected to AC line voltages. A proproplyene or metallized polyproplyene film capacitor will be the best.

 

[Gokkul55]

Question: Does the relay really switch 10 A? Or is that the contact rating? If it really switches 10 A, then I would use an S20K275 instead.

Yes the relay really does switch 10A. The relay is rated for 30A. And also the Voltage is AC (just reminding )
In the choice of varistor instead of 275V rating can i go for anything even higher like S14K1000 or S20K1000?

I don't think it's the wee little arc itself, it's probably the noise (generated by the wee little arc) on the load wiring.

Yes you are right its the noise from the arc that causes problem. But to eliminate the noise i have to eliminate the arc at least minimize it. Also the arcing reduces the life of contacts (that is what most say in forums).

Thanks
Regards
Gokkul

 

[iop995]

If it's not too complicate (and/or expensive), try to add a zero current detection circuit and made AND logic with control signal to switch off relay. Depending of frequency current and relay speed maybe need to move "zero" window slightly before effective zero current transition. Depending of current waveform may be more precise to detect derivative of current (that is maxim at zero transition for sinus waveform).

 

[OperaHouse]

Capacitor selection is important. Use a X rated (across the line) capacitor like a MPX. I worked for a company that made RC snubbers. I tell this cause it is funny. The boss wanted to get an old line CE approved. Told him it couldn't because it didn't use a capacitor that had agency approvals. Told to fill out the paperwork anyway. He went to the land of Skoggs and thought he could smooooze them with drinks I guess. Spent 40K and they stopped the test early with 70% of the samples failing. Most of these capacitors failed open. The "welds" to the foil are the weakest part. Most capacitors can recover from a puncture. Long term many capacitors will fail on AC from corona, little air pockets that light up in a changing field. X capacitors have an intermediate foil that halves the field. So if you want your product to last long term use an agency approved cap.

 

[Gokkul55]

f it's not too complicate (and/or expensive), try to add a zero current detection circuit

Sorry mate, but yes that will be expensive for this particular design, on a tight budget. Also i might need to completely reroute the PCB. So it may not work here. But do appreciate the suggestion

Capacitor selection is important. Use a X rated (across the line) capacitor like a MPX.

Will these varistor do the trick S14K1000 or S20K1000.

Thanks
Regards
Gokkul