0
\$\begingroup\$

I refer to the diagram from this thread: What is the use of the capacitor in this relay circuit?, which is also shown below. If I were to use a capacitor for my relay coil (12v latching type) which already has a flyback diode installed, what voltage rating, capacitance and type (e.g. electrolytic, ceramic) should I be looking at? Also, if I were to use multiple capacitors for my set-up, would combining different types of capacitors be ok (e.g. electrolytic, ceramic), or not recommended? enter image description here

\$\endgroup\$
2
\$\begingroup\$

Paralleling the coil with a capacitor can work the driver quite a bit harder (possibly causing it to fail) and may cause a brief dip in the 24V supply- which could cause glitches. If the driver is relatively slow or the current is limited it may be useful to reduce EMI from the coil, however the contacts usually dominate the noise and in any case the driver circuit will likely determine how big you can safely make the capacitor.

Chances are a reasonable value would be some nF and best served by a ceramic capacitor of adequate voltage rating.

\$\endgroup\$
  • \$\begingroup\$ The relay I am using is this: DSP2a-L2-DC9V; au.mouser.com/ProductDetail/Panasonic-Industrial-Devices/…. And just for my own learning, why would you recommend a ceramic capacitor over the other types? \$\endgroup\$ – Craver2000 Feb 24 '18 at 5:55
  • \$\begingroup\$ It's cheaper and smaller, and generally has as-good or better characteristics than a film capacitor in that capacitance and voltage range. Electrolytics are not in the running (too low capacitance for them to be required, and you only use them when they are required because their life is more limited and they are relatively bulky). \$\endgroup\$ – Spehro Pefhany Feb 24 '18 at 13:52
0
\$\begingroup\$

Most people don't use one.

the Diode is going to catch most of the energy when the relay switches off, so the capacitor is only needed for the short period before the diode starts conducting, if that's a problem, use a slower switch.

\$\endgroup\$
  • \$\begingroup\$ The reason why I want to implement the capacitor is that I suspect the flyback diode may not be enough for my circuit, as other components seem to be malfunctioning in the circuit, however I am just speculating that it might be have to do with some residual inductive spike from the relays. A couple of I2C devices failed over the past several days when used with a raspberry pi device. They are implemented in a circuit where the power supply is controlled by a few latching relays which switches between two types of power supply of same voltage. \$\endgroup\$ – Craver2000 Feb 24 '18 at 10:27
0
\$\begingroup\$

The capacitor is not needed. The diode may also not be needed even if the relay did not have one. The drivers I've used in the past had the diode built in. See ULN5801 etc.

\$\endgroup\$
  • \$\begingroup\$ I once tried a relay setup with my raspberry pi device without using flyback diode to the transistor, and it eventually started malfunctioning, got warmer, before it smoked. So I probably would continue using the diode. The reason why I want to implement the capacitor is that I suspect the flyback diode may not be enough for my circuit, as other components seem to be malfunctioning in the circuit, however I am just speculating at the moment. \$\endgroup\$ – Craver2000 Feb 24 '18 at 10:24
  • \$\begingroup\$ Are you sure a diode is built in to the relay? I've used very similar relays and don't remember a diode. The schematic does not show one. \$\endgroup\$ – Robert Endl Feb 24 '18 at 11:41
  • \$\begingroup\$ There is no diode in the relay. So I've connected my own flyback diode. \$\endgroup\$ – Craver2000 Feb 24 '18 at 11:43
  • \$\begingroup\$ @Craver2000 If the relay does not have a diode it is almost always necessary to have some form of energy recovery/dissipation and voltage limiting circuitry in place. Without a formal sink for the stored inductor energy it WILL find an informal sink. Without a formal voltabe limiter the voltage WILL rise to whatever level is required to transfer the energy from the coil to the sink. \$\endgroup\$ – Russell McMahon Feb 24 '18 at 20:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.