Anybody ever attempt to repair a relay?

[uoficowboy]

I have a ~35 year old SPDT relay that isn't working very well. The NO side closes and opens just fine, but the NC side never closes. Oddly enough, I suspect that during this relay's life it was never switched, or at most was switched very few times.

It is completely sealed. Is this fixable at all? I am able to order a replacement for it, but it's more fun to fix rather than replace

My guess is that the contacts are corroded and are beyond repair - but I figured I'd ask!

[jahonen]

Try connecting problematic contacts to a lab power supply, set output voltage and current limit so that contacts are not overloaded. Switch it few times and that should do the trick. Resulting spark should clean the contacts if at all possible.

Regards,
Janne

[uoficowboy]

Try connecting problematic contacts to a lab power supply, set output voltage and current limit so that contacts are not overloaded. Switch it few times and that should do the trick. Resulting spark should clean the contacts if at all possible.

Regards,
Janne

Wouldn't a spark oxidize it even more? Also, what sort of voltage/current limits do you mean? I have a 50V, 0.5A supply I use for everything - think that would be reasonable?

[David_AVD]

The switching action can blow away the built up surface oxidation.

[Skimask]

Wouldn't a spark oxidize it even more? Also, what sort of voltage/current limits do you mean? I have a 50V, 0.5A supply I use for everything - think that would be reasonable?

You would think it would wreck it even worse, but 'taint the case...
Back in my C-130 days, we had a relay panel, lots of relays, called the 'NIRP' (navigation instrument relay panel), contacts would fail all the time, low voltage contacts.
The NIRP tester was nothing more than a momentary relay coil energizer and would shoot AC across the contacts for awhile until either you gave up or the relay started working.
Seems that generally if you put DC across relay contacts, one side will lose material, and the other side will gain it, and in the process, both sides would be full of garbage and not make contact.  Shoot the current in the other direction basically blasts the crud off the contacts and 'fixes' it, at least for awhile.

[bookaboo]

Interesting stuff on the ac.
I guess if you knew what metal the contacts are then you could google on how to de-oxidise them (like using coca cola or brown sauce).

That aside I'm all for repairs but unless the cost is horrendous I'd buy a new relay just for reliability.

[SeanB]

Go buy some relay contact cleaner files ( diamond coated thin metal blades) and open it and clean. I use needle files on larger industrial relays and this works well, just often there is a load large enough and inductive so the contacts wear out electrically before the mechanical parts fail.

With those aircraft relays we just took them out and tapped them on top and put back, this often was enough to clean them. If not they were replaced without a second thought. I did change a fair number of EFD26/23 5V, 12V and 28V relays though, in a very tiny 8 pin symmetrical can about as big as a 47n 250V polyester cap. Those were a pain to unsolder from the boards as the traces were thin, the relays were spaced at 0.1in and you had to make sure of the voltage of the coil as well as there were 3 different ones. Those mostly were either burnt coils or mechanical failure, not contact problems as the sealed contacts were very reliable.

[Neilm]

Go buy some relay contact cleaner files ( diamond coated thin metal blades) and open it and clean. I use needle files on larger industrial relays and this works well, just often there is a load large enough and inductive so the contacts wear out electrically before the mechanical parts fail.

It depends how it was sealed. I quite often see relays that are potted. If the relay the OP has is potted you can't do anything without digging the potting compound out and trying to re-pot it. Those relays can also be vacuum sealed so that would a non-starter - although it is older than I am so if there was a vacuum there probably isn't one now.

Neil

[madires]

Relay contacts are somewhat soft, not as solid as most would think. And the contact alloy needs some current to be kept clean. It's not just overloading which "bakes" the contacts together. Missing switching operation or very low currents can cause the same. For such situations gold plated contacts should be used.

The best deal to fix the relay is to hit a little bit (with the handle of a screwdriver for example) to break the bonding, switch it a few times and pass some current through the contacts.
 

[JoeyP]

I used to "fix" reed relays routinely by connecting a function generator to the coil and slapping the contacts at around 30Hz for a few minutes. I did this for relays that were either no longer available or extremely expensive. It would typically give them at least another year or two of life.

For your sealed SPDT relay, it sounds like disuse is probably the problem. The mechanism might just be gummed up from sitting. If you have a switchable source which can do it, I'd try driving the coil with something switching at around 10Hz for a while. Couldn't hurt to have some AC current available across the NO contacts during the process so as to clean the contacts as well.

[Clear as mud]

I fixed a relay for a power driver's seat in a car one time by carefully opening it up, cleaning/sanding the contacts, and bending the contact slightly so the spring would press it together again.  It had gotten worn and the springiness had gone out of it.  That was on a 1980 Oldsmobile, exactly like this car.

More recently, my furnace fan relay quit working, and I took it out, to take it to the furnace parts store and ask for a new one, but as I was walking to class, I accidentally dropped it on the sidewalk.  I reinstalled it, and it worked for three more years.  After that, dropping it on the sidewalk again did not fix it and I had to buy a new one.