Coil trips RCD/GFCI

entrepiece. 240v 240 ohms measured, 10w (according to rating plate). It wor ks flawlessly, but often trips the 30mA RCD when switched off. 500v inuslat ion testing shows no problem. I assume it has no built in snubber. I'm not working out why it trips the RCD. Any ideas?

Presumably the turn-off voltage spike is higher than 500V and something spa rks over to earth. That could well trip the RCD.

Back to back avalanche diodes to 0V rather than ground might be enough to c ure the problem.

snipped-for-privacy@gmail.com wrote in news: snipped-for-privacy@googlegroups.com:

It is called "back EMF".

centrepiece. 240v 240 ohms measured, 10w (according to rating plate). It w orks flawlessly, but often trips the 30mA RCD when switched off. 500v inusl ation testing shows no problem. I assume it has no built in snubber. I'm no t working out why it trips the RCD. Any ideas?

parks over to earth. That could well trip the RCD.

cure the problem.

obviously being a 1946 appliance there is no earth connection, and no user shocks occur, and the insulation is sound. If it were arcing it would have degraded.

NT

Back emf is certainly called back emf. Would you like to explain how that trips an RCD?

NT

The switchoff transient could pump a current spike through the winding-to-ground capacitance. Try adding a MOV somewhere. Or reversing the coil connections.

Some GFDs have arc detectors. An MOV could help that too.

With respect to what? How was the 'insulation' being tested. Is there an earth?

I might run in through a filtered trailing socket.

I had a set of hair clippers that did that. An RC snubber across the coil fixed it. I think 100R 100nF. MOV should be good option also.

My guess is the residual current detector HF response is wrong, somewhere I once saw reference to response should fall off above 2.5kHz

- RCD/GCFI should not trip on a high energy burst of RF.

piglet

Those diodes have an avalanche voltage between a bit above the rated value and 1600V. He needs one that's guaranteed to be below 350V, etc. That would be a TVS type. Like a 1.5KE200A for 190 to 210V peak, or a 1.5KE440A for 418 to 462V, etc.

entrepiece. 240v 240 ohms measured, 10w (according to rating plate). It wor ks flawlessly, but often trips the 30mA RCD when switched off. 500v inuslat ion testing shows no problem. I assume it has no built in snubber. I'm not working out why it trips the RCD. Any ideas?

It doesn't like big arc voltages that possibly exceed the insulation rating of the mains wiring. In U.S. they're called *arc* fault circuit interrupte rs, that should be a clue. One way to kill it that's nearly as old the vibr ator is to hang a full wave bridge in parallel with the vibrator, the brid ge DC output drives a big electrolytic, say 500WVDC, with enough capacity t o take the hit with negligible voltage rise. That should limit it to 350V, it shouldn't arc across the mains rated on/off switch, and the interrupter should let it pass.

centrepiece. 240v 240 ohms measured, 10w (according to rating plate). It wo rks flawlessly, but often trips the 30mA RCD when switched off. 500v inusla tion testing shows no problem. I assume it has no built in snubber. I'm not working out why it trips the RCD. Any ideas?

The anything to ground capacitance has to be tiny. It's plastic cased & has no ground/earth connection. There is a metal shaft from the coil but it ha s a big bakelite end on it, so very little C. Which doesn't stop it happeni ng of course, just that there won't be much i involved.

NT

no, the case screws & the metal rod from the coil were treated as earthed to test it.

NT

centrepiece. 240v 240 ohms measured, 10w (according to rating plate). It w orks flawlessly, but often trips the 30mA RCD when switched off. 500v inusl ation testing shows no problem. I assume it has no built in snubber. I'm no t working out why it trips the RCD. Any ideas?

I think a snubber is the way to go, if there's space for one.

My money's on that too.

NT

centrepiece. 240v 240 ohms measured, 10w (according to rating plate). It w orks flawlessly, but often trips the 30mA RCD when switched off. 500v inusl ation testing shows no problem. I assume it has no built in snubber. I'm no t working out why it trips the RCD. Any ideas?

ng of the mains wiring. In U.S. they're called *arc* fault circuit interrup ters, that should be a clue.

I've heard of AFCIs, but wasn't aware they were RCDs.

ve bridge in parallel with the vibrator, the bridge DC output drives a big electrolytic, say 500WVDC, with enough capacity to take the hit with negli gible voltage rise. That should limit it to 350V, it shouldn't arc across t he mains rated on/off switch, and the interrupter should let it pass.

Time to look for a small selenium BR & a wet lytic :)

NT

ng centrepiece. 240v 240 ohms measured, 10w (according to rating plate). It works flawlessly, but often trips the 30mA RCD when switched off. 500v inu slation testing shows no problem. I assume it has no built in snubber. I'm not working out why it trips the RCD. Any ideas?

ting of the mains wiring. In U.S. they're called *arc* fault circuit interr upters, that should be a clue.

wave bridge in parallel with the vibrator, the bridge DC output drives a b ig electrolytic, say 500WVDC, with enough capacity to take the hit with neg ligible voltage rise. That should limit it to 350V, it shouldn't arc across the mains rated on/off switch, and the interrupter should let it pass.

I'm making the assumption yours are the same as ours. Here the AFCI is prot ection primarily against arcs, but they also trip on what they call a machi ne ground fault of 30mA or more. A dedicated GFCI has no arc protection (un less the circuit arcs to GND) and it's ground fault trip threshold is much more sensitive at 5mA.

I think he means "backlash".

ote:

trips an RCD?

Residual current detectors compare the current going into a device on one w ire with the current coming out of the deive on another wire. If they don't match, some current is finding another path back to the power supply, whic h could pass through somebody. The RCD turns off the power to the device wh en this happens.

If your coil generates a high voltage spark which finds some path back to t he other end of the coil that doesn't pass through the two power leads, th at could - if big enough, and sustained for long enough - trigger the RCD ( also know as an earth leakage trip).

The brakdown path could be through air, rather than the enamel on on the co il, and wouldn't necessarily degrade or wreck the insulation.

ving centrepiece. 240v 240 ohms measured, 10w (according to rating plate). It works flawlessly, but often trips the 30mA RCD when switched off. 500v i nuslation testing shows no problem. I assume it has no built in snubber. I' m not working out why it trips the RCD. Any ideas?

rating of the mains wiring. In U.S. they're called *arc* fault circuit inte rrupters, that should be a clue.

l wave bridge in parallel with the vibrator, the bridge DC output drives a big electrolytic, say 500WVDC, with enough capacity to take the hit with n egligible voltage rise. That should limit it to 350V, it shouldn't arc acro ss the mains rated on/off switch, and the interrupter should let it pass.

otection primarily against arcs, but they also trip on what they call a mac hine ground fault of 30mA or more. A dedicated GFCI has no arc protection ( unless the circuit arcs to GND) and it's ground fault trip threshold is muc h more sensitive at 5mA.

We have RCDs & RCBOs with 30mA threshold that protect entire ring circuits. AFCIs are being talked about but I've not seen one yet.

NT

??

NT