I've build a simple set up with a Raspberry Pi and this relay board to switch some lights. I have removed the jumper from the relay board, connected 12V to the VCC/GND pins, connected the Pi GND to the relay board's COM pin, and the GPIO pins to the relay board inputs. I connected the load to the relay's NO (normal open) connectors. When I use the gpio utility I can switch the pins on the Pi on and off; the relay board follows this, as confirmed by the LED and the load on the relays switching on and off. So far so good.

But then, after a little bit of playing around, the relays stopped switching off. The LED for the relay does switch off, but the contact stays closed; the lights connected to that relay stay on. Not all relays have the problem, and every now and then the ones who do have the problem work once or twice; but they don't reliably switch off.

So, if we ignore mechanical failure of the relays for a second (I doubt all relays on a board would have the same defect...), what could cause this behaviour? I mean, when I remove the signal from the relay board, the optocoupler on the board would stop conducting, the signal would disappear on the 'other side' of the optocoupler, and all relays should switch back to 'open', right? But even if I unplug the Pi, the relays stay closed! And when I unplug everything (the Pi, the power to the relay board, the power to the load) and replug, the relays that were stuck still are closed.

What further steps can I take to debug this? Thanks.

  • 2
    \$\begingroup\$ What load were you switching and what is the spec of the relays used? \$\endgroup\$
    – Andy aka
    Commented Feb 10, 2017 at 22:53
  • \$\begingroup\$ The loads are 2 to 4 e27 sockets into which are screwed led lights of about 5w each. (These are 'filament bulb replacement' lamps, so working on 230v ac). So between 10 and 20w per relay @ 230v ac. The data sheet for the relay is on the link I posted, under the second tab. It seems to be a mechanical relay, I imderstood ssd relays are tricky wrt inductive loads but I think that doesn't apply to my situation. Also, when everything is unplugged and I measure across the contacts with a multimeter, the relay is still in the closed position. Thanks. \$\endgroup\$
    – Roel
    Commented Feb 10, 2017 at 23:15
  • \$\begingroup\$ So, off the top of your head you don't know the contact ratings then? \$\endgroup\$
    – Andy aka
    Commented Feb 10, 2017 at 23:36
  • 1
    \$\begingroup\$ I have the same board (unused) marked ver:8R1A 12/03/14. The relays are marked "Songle 10A SRD-12VDC-SL-C" A datasheet says the relays need 12V but the same PCB design is used for both 5V and 12V relay modules. The vendor description says "It only requires a voltage of approx 1.0V to switch the inputs on but can handle input voltages up to 12V. This makes it ideal for both 5V and 3.3V devices." \$\endgroup\$ Commented Feb 11, 2017 at 0:09
  • \$\begingroup\$ Mechanical failure could mean manufacturing defect or in this case user operation failure. \$\endgroup\$
    – Passerby
    Commented Feb 11, 2017 at 17:31

1 Answer 1


what could cause this behaviour?

Welding the relay contacts together due to a high initial current.

If your lights on the load side are incandescents, they have a low resistance until the filaments heat up. So a relatively high current can pass for several milliseconds.

enter image description here
Image from Lamptech

Note the 500W lamp has a normal operating current of around 2A but during the first five milliseconds the current is greater than 15A. Your relays are rated 10A.

What further steps can I take to debug this?

Measure the current flowing in the first few milliseconds.

Hit the recalcitrant relay with a blunt object so that the mechanical shock can break the contacts free?

Take a relay apart and inspect the contacts?

See also


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.