Curious experiment with a relay
Antonio, Fri Dec 19 2014, 02:04PMI have observed a curious phenomenon while looking at a relay circuit. The assembly is very simple: A voltage source, a mechanical switch, and a relay coil in series. A 12 V supply, a micro-switch, and a 12 V relay, for example. Observe in an oscilloscope the voltage over the relay coil.
It's well known that when the switch opens a large voltage pulse appears over the relay coil, caused by the sudden interruption of the relay current, and a spark occurs at the switch.
The oscilloscope shows this pulse clearly, or several pulses in series, negative if a grounded positive supply is used.
The curious observation is that the pulses never exceed ~300 V.
The reason appears to be "Paschen's law", that says that in air there is a minimum voltage that can cause a spark, at about 300 V. So, when the switch opens interrupting the current the voltage over the open switch, that is still opening, and with the contacts at a small distance, must get close to 300 V before a spark limits the voltage.
In most cases (I tried several relays) what is observed is a series of ~300 V pulses, but in others, when the current is larger, a long pulse limited at ~300 V appears. Smaller pulses sometimes appear too, but they are probably due to contact bounce.
A capacitor in parallel with the switch reduces the voltage rise rate, but up to the point where there is no more spark, the voltage limit appears to be the same ~300 V in the tested relays.
Re: Curious experiment with a relay
Sulaiman, Fri Dec 19 2014, 02:38PM
assuming that your 'scope probe is not significantly loading the circuit,
this is the reason for R.C snubbers across contacts
to limit the voltage gradient to allow time for contact separation distance vs. time
capacitor sets dV/dt (=Icoil/C), resistor limits cap pulse current when contacts close.
(and absorbs some of the back-emf, and during contact bounce at closing)
(I've been considering reed switches in a long life motor)
any switch has a maximum dV/dt on turn-off.
I see what you mean about the 300v, a good guide,
the opening contacts get an equivalent 300V 'head start' in the race of distance vs. voltage rise.
Sulaiman, Fri Dec 19 2014, 02:38PM
assuming that your 'scope probe is not significantly loading the circuit,
this is the reason for R.C snubbers across contacts
to limit the voltage gradient to allow time for contact separation distance vs. time
capacitor sets dV/dt (=Icoil/C), resistor limits cap pulse current when contacts close.
(and absorbs some of the back-emf, and during contact bounce at closing)
(I've been considering reed switches in a long life motor)
any switch has a maximum dV/dt on turn-off.
I see what you mean about the 300v, a good guide,
the opening contacts get an equivalent 300V 'head start' in the race of distance vs. voltage rise.
Re: Curious experiment with a relay
Ash Small, Fri Dec 19 2014, 09:00PM
there is also a 'minimum distance' at which the gap can artc. Too close and the voltage required is higher. The points must open to this minimum distance before the arc occurs'
There are cases, due to Paschen's law, where the discharge path is not the shortest distance, due to this phenomenon.
Ash Small, Fri Dec 19 2014, 09:00PM
there is also a 'minimum distance' at which the gap can artc. Too close and the voltage required is higher. The points must open to this minimum distance before the arc occurs'
There are cases, due to Paschen's law, where the discharge path is not the shortest distance, due to this phenomenon.
Re: Curious experiment with a relay
doctor electrons, Sun Dec 21 2014, 03:05PM
This may not be the case here but, some scopes can only measure 300v before the signal is internally clipped.
If your scope does do that, try a scope with a 600v input and see if your voltage goes over 300.
I watched an engineer try to take measurements on a large dc bus with a portable scope meter for 3 days before he realized
it was not rated to read what he was looking for.
doctor electrons, Sun Dec 21 2014, 03:05PM
This may not be the case here but, some scopes can only measure 300v before the signal is internally clipped.
If your scope does do that, try a scope with a 600v input and see if your voltage goes over 300.
I watched an engineer try to take measurements on a large dc bus with a portable scope meter for 3 days before he realized
it was not rated to read what he was looking for.
Re: Curious experiment with a relay
Dr. Dark Current, Sun Dec 21 2014, 03:20PM
Try a capacitor in parallel to the switch. Anyway this is how mechanical car ignition or fluorescent starters work
Yes you're probably right that the voltage is limited by sparking - but of course also by how quickly can you separate the switch contacts.
Dr. Dark Current, Sun Dec 21 2014, 03:20PM
Try a capacitor in parallel to the switch. Anyway this is how mechanical car ignition or fluorescent starters work
Yes you're probably right that the voltage is limited by sparking - but of course also by how quickly can you separate the switch contacts.
Re: Curious experiment with a relay
Ash Small, Sun Dec 21 2014, 05:15PM
Minimum breakdown voltage is 327 V in air at standard atmospheric pressure at a distance of 7.5 µm
Ash Small, Sun Dec 21 2014, 05:15PM
Minimum breakdown voltage is 327 V in air at standard atmospheric pressure at a distance of 7.5 µm
Re: Curious experiment with a relay
Antonio, Mon Dec 22 2014, 12:24AM
Really, I have to verify if the 10x probe used was rated for more than 300 V. But I looked at the current coming from the power supply using a current transformer, and the spark current could always be seen, indicating that the sparking was at the switch.
Antonio, Mon Dec 22 2014, 12:24AM
Really, I have to verify if the 10x probe used was rated for more than 300 V. But I looked at the current coming from the power supply using a current transformer, and the spark current could always be seen, indicating that the sparking was at the switch.
Print this page