Cleaning Relay contacts by Electrical Current.

ian field

Joined Oct 27, 2012
6,536
One thing we didn't talk about was the environment where the failed relays come from and is that a factor?

I agree that you should be looking into correctly chosen alternatives.

I worked in a lab and when there were recurring failures with a known cause, I would fix them forever.

My predecessors would replace a drive transistor every year or so. I took me longer, but once fixed, the next failure was a circuit breaker that died of old age.
Relays with completely exposed contacts are obviously vulnerable to any environmental contaminants - tar particulates from smoking are probably about as bad as it gets.

Contacts with a bit of wetting current usually give little trouble, when switching low level audio; it doesn't take much contact resistance to introduce crackles and hum.

At the other extreme; contacts operated close to their limits tend to pit and erode. Usually that's the end of life condition - rigging a way to exercise the contacts without load may polish the contact areas, but its only a temporary fix in most cases.

Many relay contacts are case hardened - once that's worn away, the metal underneath is like cheese.
 

cmartinez

Joined Jan 17, 2007
8,219
One small tidbit, for what it's worth. When I was a teen in a shop repairing pinball machines, sometimes a relay would get its contacts fouled, and stop working. It was invariably a relay that was switching DC. Relays that had AC running through the contacts never seemed to fail. So maybe AC, or reverse bias DC, will clean the contacts?
You might be on to something... in a car battery, rust and corrosion normally builds up on the cathode, and not the anode. Maybe only one side of the contact is being affected by this sort of phenomena?
 

#12

Joined Nov 30, 2010
18,224
This is a common problem. Consider 'pilot duty' contacts. They are gold plated. Use too much current and the gold disappears. You are left with the backing material which is more suitable for current that will destroy gold plating. Consider the defrost cut-off switch in a Kenmore refrigerator. It used to switch a 3 amp, 120 VAC heater. Now it switches a 5 volt signal for a microprocessor. It fails because they didn't replace it with gold plated contacts. The contactor on an air conditioning compressor has a predictable lifetime because you have to know that throwing a 100 amp start surge into a compressor, 3 times an hour, for 10 years, adds up to over 100,000 operations if the air conditioner only runs for 6 months per year.

Contacts go bad.:(
 

cmartinez

Joined Jan 17, 2007
8,219
This is a common problem
It's a pain in the backside for me too. Sometimes I make machines whose controllers operate at 5V. Many times over I had to keep replacing malfunctioning microswitches that served as position sensors, not knowing what was going on. It wasn't until I learned about contact corrosion that I decided to use 12V signals at the switches instead, and install a simple resistor voltage divider at the controller to bring it back down to TTL (I normally use a 1k and a 680Ω resistor as dividers). That not only took care of the problem, but it also minimized spurious noise arriving from the long wire connected to the switch.

Still, I wonder how long that arrangement is going to last before the problem reappears. The switch now works at 12V, but is only conducting 7mA.
 

#12

Joined Nov 30, 2010
18,224
10 mA is typically a good number to aim for. 7 mA is pretty close.
10 ma is the lower limit for a 317 chip. To convert that to 5 volt logic, install ~ 470 ohms on the receiving end. Any voltage lost in the wires and connectors will be compensated by the constant current configured LM317L. And I wouldn't go for the full 40 volts the chip is rated at, but 30 volts will kick the fungus off the contacts better than 12 volts will.
 

Attachments

Last edited:
Top