Electronic – Relay – Doesn’t work…at first

relayreliability

I have a latching small signal <2A relay (V23079E1201B301) that switches between two audio signals. The relay has worked in about 10k boards over 7-8 years without much issue. We recently got a run done and 4 of 6 boards that we built the relay doesn't work at first – it passes audio in one position but not the other. But then after it switches on/off 15-25 times it starts working sometimes, and then after 30-40 switches back and forth it passes signal fine. If I power it off and check it a few hours later it's still fine.

I hear it clicking every time so I don't feel like it's sticking. It's almost like there's oxidation on one set of contacts and it eventually burns off. I'm a little nervous because I haven't experienced it before, and not sure if there's going to be long term reliability ramifications. I ran the issue by our repair guy and he says the relay never an issue with units coming back from the field.

Anyone else experienced this?

UPDATE: I broke apart a good relay (date code 2014), and one of the recent ones that had been 'exercised' a bit and now conducts (date code 2019). Comparing them they're super similar. A few differences, but nothing that screams counterfeit.

The obvious issue though is that I can see black stuff on one set of contacts (the moving contacts). It's worn off and gold again in the middle, likely from the exercise, but you can see black bands along the outside. Interestingly there doesn't seem to be any other decay or oxidation on any other part internally – or the stationary contacts. Maybe this suggests that it didn't become un-sealed, but that the moving arm has a defect with the finish. I'm not sure.

Going back and forth with the CM and manufacturer to try to get to the bottom of it. Thanks for all the advice.

Best Answer

Make sure the relays are not counterfeit. I assume you've specified gold-plated contacts. Might be worth dissecting one from current production, though it would be difficult to test if the plating (assuming there is some) was thick enough.

But my fancy ratio tran$$$former box "exercises" the relays upon each power-up, so they are expecting trouble.

Generally gold-plated contacts are okay if you pass more than a few uA (10uA, say) "wetting current".