Another odd fault, you decide..

[flyingsparks]

Got called to look at a TIG welder at a local engineering factory. The welder was plugged into an extension lead, 3phase 32A SY flex. Nearly every morning when it was turned on it would trip the local c32 mcb and the c40 feeding the submain to the local DB. Then you could reset the mcb's and try again and the welder would be fine, running all day with no problem.

They tried plugging the welder directly into the 32A socket on the wall, and this never had any problems with tripping.

The lead had all tight connections, meggered and meggered at 500M.

I believe I have now rectified the problem, as it has been fine now for a few days.

Any ideas as to what I think caused the problem?

 

[Jackrae]

So what did you do to cure the problem ?

Inrush current charging "flat" capacitors ??

 

[JohnW2]

Questions, not answers ... did the extension lead have a neutral? was it wired correctly pin-to-pin?

Kind Regards, John

 

[RF Lighting]

One of the phases open circuit on the SY cable

 

[flyingsparks]

No, all phases on lead correct. No neutral in lead, not required.

I will admit I dont fully understand why this may be a cause to take out two MCB's, but the armouring on the SY was not earthed. Earthing it cured the problem. Could it be a similar kind of phenominan as to the problem I had the other day with the DP switch taking out the RCD, with some form of capacitive coupling in the flex? The flex was quite long probabily about 20m, so capacitance could be an issue?

 

[RF Lighting]

They were just standard C type circuit breakers and no RCD protection or anything like that?

What's the input rating of the welder?

 

[flyingsparks]

Yes standard C types. There is 30mA RCD to, however I am not sure if this was tripping or not. RCD main switch on the sub board.

Unsure of the welder rating off top of my head at the moment.

 

[JohnW2]

I will admit I dont fully understand why this may be a cause to take out two MCB's, but the armouring on the SY was not earthed. Earthing it cured the problem. Could it be a similar kind of phenominan as to the problem I had the other day with the DP switch taking out the RCD, with some form of capacitive coupling in the flex? The flex was quite long probabily about 20m, so capacitance could be an issue?

I would have said 'no way'. To get a handful of milliamps through stray capacitances in cable etc. to trip an RCD is one thing, but to get an MCB operating by the same mechanism is a totally different kettle of fish. If the current being taken (including inrush currents etc.) was so marginal in terms of MCB operation that adding a tiny bit of extra current through stray capacitances resulted in trips, then I would have expected sporadic trips (resulting from random variations in actual current) when it was plugged straight into the socket (without extension lead). However, as always, I obviously could be wrong!

Kind Regards, John

 

[RF Lighting]

I think it's just inrush current.

 

[Spark123]

Yep, inrush current. Don't be surprised if you get a call tomorrow to say it is doing it again. Type D breakers (if Zs allows) or fuses are more tolerant.

 

[flyingsparks]

Doesn't explain why it never tripped when not using the extension lead tho?

Yes I will check the zs if it happens again and see if a D type will be ok..

 

[RF Lighting]

Maybe the inrush duration is shorter due to the reduced VD from removing the extension cable.

 

[JohnW2]

Doesn't explain why it never tripped when not using the extension lead tho? Yes I will check the zs if it happens again and see if a D type will be ok..

It does seem very odd, but I really think it would be stretching things to come up with an explanation based on cable capacitance, unless there were some large and extremely high frequency components of the inrush current.

A D40 is going to need a Zs of ≤0.23Ω, which may be asking a bit! Of course, if the Zs is low enough, then changing to a Type D will probably cure the problem, regardless of the mechanism that is causing it.

Kind Regards, John.

 

[JohnW2]

Maybe the inrush duration is shorter due to the reduced VD from removing the extension cable.

Possibly, but the inrush magnitude would then be even greater, at least partially cancelling the effects of any reduction in duration.

Kind Regards, John

 

[flyingsparks]

A D40 is going to need a Zs of ≤0.23Ω, which may be asking a bit! Of course, if the Zs is low enough, then changing to a Type D will probably cure the problem, regardless of the mechanism that is causing it.

Kind Regards, John.

Ze was 0.09 when measured today, and this sub board and socket isn't very far from mains, so may be in luck!