I recently purchased one of these relay boards ...

enter image description here

to use along with my Raspberry Pi to switch an AC outlet on/off to control a fountain pump. Since the relay board requires 5V for the control lines and needs to pull more current (IIRC) than the RPi can supply on 5V, I put together this little circuit to test out my system:


simulate this circuit – Schematic created using CircuitLab

Relay Schematic: enter image description here

I know that my "To Relay Control" line is only going to pull down to ~.3V when the RPi GPIO is high, but it was enough to toggle the relay, so I went with it. I have software that turns the pump on 15 minutes out of every hour, and the system was functioning fine for a few hours. However, after a few (~8) hours of operating the relay is no longer switching -- even if I pull the "To Relay Control" line directly to ground the relay will not flip into the "on" position. I thought this might have just been a random bad part, but I switched to another relay on that board and had the same problem after running fine for a few hours. The LED on the relay board lights up when the line is pulled low, but that's it, I don't hear a "click" out of the relays anymore when I toggle them.

Is there something about my test circuit that is causing these relays to fail? In general, what would cause a relay to fail in the manner that I've described?

  • \$\begingroup\$ The unlabelled arm of the relay coil is connected to +5 V? \$\endgroup\$
    – The Photon
    Commented Dec 22, 2015 at 17:03
  • 1
    \$\begingroup\$ What's the point of connecting the normally closed on the relay to anything? Maybe you're frying (eg. welding) the relay contacts. How much watts/horsepower does the fountain pump draw? Motor loads can be unfriendly to cheap relay contacts. \$\endgroup\$ Commented Dec 22, 2015 at 17:06
  • \$\begingroup\$ What voltages are connected to the pump? The schematic only says 'ACin' and 'ACout'. How much power is the pump? Also, why are both positions of the relay connected? For 'on/off' it only needs one side connected. \$\endgroup\$
    – gbulmer
    Commented Dec 22, 2015 at 17:08
  • 1
    \$\begingroup\$ Do you have a snubber in the system? Does the damage occur if you drive an unloaded relay for, say, 24h cycles? \$\endgroup\$
    – pjc50
    Commented Dec 22, 2015 at 17:18
  • 2
    \$\begingroup\$ If your pump provides power right after disconnection from mains (i.e. it turns from a motor to a generator powered by its inertia), connecting its two terminals via relay contacts may "weld" them in that position. Even if you power the coil, the contacts will not change anymore. Can you check the coil's continuity to verify if it is open? If it is not, then probably is some type of contacts being "glued" problem. \$\endgroup\$ Commented Dec 22, 2015 at 17:35

1 Answer 1


enter image description here

  1. Get rid of the extra AC connection on the relay. This contact set could weld themselves if an inductive load were present and the contacts were shorted like that.

  2. For inductive load add a snubber across the relay contacts to help protect the contacts from welding.

  3. I took the liberty if showing the AC source correctly and not as + and -.

  • \$\begingroup\$ point 1, that extra circuit also means an arc is directly across the supply. \$\endgroup\$ Commented Dec 23, 2015 at 4:02
  • 1
    \$\begingroup\$ Four months later and my relays are still running! This answer solved my problem. \$\endgroup\$
    – Adam
    Commented Apr 28, 2016 at 6:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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