[RESOLVED] Relays sticking?



  • Ok, I had a chat with my father, who is an old engineer and used to work in BT in the 50's, he told me of a trick they had back then. Apparently they let the relay weld and would "unstick" it, and do this a few times. He claims this would even out the surface and create a bigger surface of contact so less micro-welding.

    Odd no?

    Well, I tried this, and after 5 or 6 times the relays no longer locked!

    All is well now :)

    Thanks guys.



  • @Samuel-Mathieson said in Relays sticking?:

    ...Apparently they let the relay weld and would "unstick" it, and do this a few times...

    I'm not sure if I understood right, you took the solder Iron and made the 8legs of the relays "warm" that the solder ran back to the hole of the circuit board from the extension? Without adding new solder?



  • No, he let the relay stick a few times and it worked itself out. Interesting!

    Glad you're on your way. BTW, not sure your level of solder skills, but I like two of the suggestions above and would try either of them if you get more problems:

    1. Replace the relay with a pin-compatible heavier-duty one (or)
    2. Have the onboard relays drive the coil of heavier-duty relays

  • administrators

    Amazing, glad to hear to worked out!
    @Kit-Bishop I had no idea that was even possible!

    @Samuel-Mathieson make sure to post your project when it's done, I would love to see the final result.



  • @Samuel-Mathieson said in Relays sticking?:

    Should I use a diode on the contacts?

    Thanks for any help.

    Try throwing a diode across the coil of the relay eliminating contact bounce as the culprit.



  • There are already rectifier diodes on the Relay Expansion, one for each relay coil. They can be seen in the schematics here, see components D3 and D5.



  • @Gabriel-Ongpauco Nice schematics



  • @Gabriel-Ongpauco @Guest @Samuel-Mathieson Yes, the Relay Expansion does already have diodes across the coil. This will reduce bounce due to surges in the coil current but won't necessarily eliminate it. The way to reduce arcing when the contacts make/break is to use capacitors across the points themselves - see: http://www.industrologic.com/mechrela.htm for an example.



  • Its funny how things worked out. I did think about getting another relay to activate with the relay expansion relay but in the end I decided that would beat the purpose of a relay.

    I did not think of replacing the relay expansion relays. I think that would be a good way to go if I see more problems in the future.

    Thanks again everyone.



  • Some background information about the reason for the "arcing" and "micro-welding" described above:

    It is usually caused by inductive loads, like electromagnets, motors, transformers - pretty much anything that is based on a coil of some kind.

    Coils have a surprising property - the current that flows through them cannot suddenly change. So if you have a motor running on, say, 200mA, and your relay breaks the circuit, the current needs some time to ebb, it cannot go from 200mA to zero at once.

    But how is this possible with the circuit open? There's only one way - the voltage between the now open contacts will rise so high that the current can flow through the air between the contacts! High voltage can ionize air to make it conductive. In small scale this is called an electric arc, in large scale it's called lightning...

    In both cases, it's very hot, and happening between relay contacts causes contact material to melt and micro-weld. Even if the relay spring is strong enough to rip the contacts apart eventually, and the relay does not stick, this ruins the contacts over time, and should be avoided.

    But how?

    If the circuit is DC, it's simple. You just need a so called flyback diode over the inductive load which allows the coil to discharge its energy without generating high voltage.
    Note that this is also essential when driving a relay (a coil, too!) from an electronic circuit. If you forget the flyback diode, the high voltage will disturb or even destroy the electronics. See D3 and D5 on the Onion Relay Expansion schematics.

    If the circuit switched by the relay is AC, a diode does not work (would short-cicuit half of the time and destroy itself). Instead, use a RC snubber, a capacitor in series with a resistor. For use with 230V AC, ~20..100nF/400V capacitor and a 100Ω/2W resistor works. The snubber consumes some energy of it's own, that's why the resistor should be a 2W type, and will produce some heat. There are also RC snubbers in a single part, like this one.

    Hope this help to tame your relay circuits ;-)

    As a side note: there are circuits where the high voltage caused by breaking the circuit of a coil is the desired result - ignition in combustion engines, electric cattle fences...



  • @Lukas-Zeller Some very good information particularly in relation to inductive loads. :-)

    I also have some concern in relation to the resistive load of incandescent light bulbs at turn on time.

    The original poster (@Samuel-Mathieson ) indicated that he was driving 20W @ 120Vac - this (when the bulb is warm) equates to about 166mA - this should be within the capabilities of the relay.

    However, if the light bulbs are typical tungsten filament bulbs, they can have a cold resistance of around 1/15 th of the warm resistance. So the current at switch on from cold could be as high as 2.5A (2500mA) briefly - this could potentially briefly exceed the relays rating



  • If it is possible and does not affect the operation of the circuit, place a delay in the opening of the circuit allowing the inductive loads to subside then opening the contacts under non inductive load.


Log in to reply
 

Looks like your connection to Community was lost, please wait while we try to reconnect.