Mains switch makes trip device go poof

[eamoex]

Hi all, I'm not an engineer, I just happen to like Dave's work and have no big knowledge in electr(on)ics. But I'm sure you guys know the answer. I have a wall outlet at my home that has a power strip connected. The strip has a switch and a few devices hooked up to it, including a PC with a beefy supply, its monitor and a desk lamp. The trip device in the garage often goes poof when I switch the strip on. I know what causes this (or I think I know –– "transients", is it?) and I know that I can switch the devices on one by one to avoid making the trip device flip.

Question: is there something I can make, or a device I could add to the circuit, to retain the "one button switches all" functionality without the trip to go "whaaaaaa"?

Note: I can solder and I know basic electronics parts (I've fooled around modding Commodore Amigas back in the day).

Edit: I know it's high voltage, high current. I will not put myself in danger. I don't muck around with mains. I'm just curious if a solution, theoretical or ready-made, exists.

[richard.cs]

Assuming the trip that goes is an RCD / GFCI then this is usually caused by a combination of two things:
1) One device plugged into the strip has a significant leakage between earth and neutral (or a leak in the power strip).
2) The switch is double pole, but (due to mechanical tolerances, etc.) usually makes the live before the neutral.

So, if for example one device had a 5 kOhm leak between N and E, in normal operation the N-E voltage is low, perhaps 10 Volts. A leakage current of 10/5000 = 0.001 = 2 mA flows but nothing happens. Instead consider the moment of switch-on, say that the live contact in the switch makes first then the neutral wire of all the connected devices flies up to nearly full mains voltage through the low-impedance of the loads.  Lets say that gets up to 220 V, now the 5 kOhm leak passes 220/5000=44 mA, the RCD sees a 4 mA difference in current between L and N and trips.

The fix is to do any of the following:
1) Identify the device with the earth/neutral link and fix it or throw it away.
2) Use single-pole switching in the live only.
3) Use double pole switching but ensures the neutral makes first and breaks last.

There are other possible causes, but this is very common.

[capt bullshot]

This also can happen with your normal circuit breaker, especially the modern "B" characteristic types. No cure but either replacing the breaker with a "C" characteristic or turning on the devices sequentially, or some kind of inrush current limiter:
https://wunderkis.de/esb/index.html (German text only)

Edit: Sometimes a long enough extension cord also helps. Inrush current will be smaller due to the increased resistance.

[eamoex]

Thanks to both of you.

@richard.cs: okay I think I understand. Can I test the N-E resistance with my multimeter on the corresponding pins on the switched off devices? Will this reading be of value here?

@capt bullshot: nice text, I can only understand 50% German (shame as this is my country's main language...) but I think the guy has the exact same problem as I! Thanks for the link... and there's the schematics to his contraption, very nice.

[richard.cs]

Measuring the N-E resistance with a multimeter would help, but it might be something that appears open-circuit at low voltages (like your multimeter tests with) but starts to leak at higher voltages so it's not a definitive test.

The two problems described are different, mine relates to earth leakage, the other to inrush current. A few questions to help us to work out which it is:

What kind of trip device is in your garage? (is it tripping on overcurrent or imbalance?) If you're not sure then tell us what's written on it or post a photo.
Does it only ever trip at the moment of switching on?
If you switch on another way does it still trip? i.e. with a switched outlet, or just by plugging the power strip in.
Is there any one appliance that when unplugged from the power strip means it never trips?

[hermit]

Sometimes circuit breakers just go bad.  They are usually cheap enough to replace.  If that doesn't do the trick you have a spare.

[eamoex]

@richard.cs: yes I get the 2 problems are different. To your questions:

• attached is a photo of the popping sucker; I have no clue what this thing is
• it only ever trips when switching on
• it has tripped when using the physical rocker switch on the power strip AND a radio controlled outlet switching thingy I just bought (some device you plug to the wall outlet and then your device to — in my case, I put it between the wall outlet and the "always-switched-on" power strip)
• it never seems to trip if the PC power supply is off (not just the PC of course, I mean the rocker switch on the power supply unit itself)

@hermit: well, seeing the apparent age of the thing, you might have a point here...

[capt bullshot]

This is an real crusty old circuit breaker. It's S characteristic should allow for a pretty high inrush current, more than a modern "B" one, so I wouldn't expect the power supply to trip it normally. My recommendation: Have it replaced - but then you'll get a modern "B" one and may run into the same problem  Ask the Electrician for a "C" characteristic and hope he doesn't refuse to install it (as they sometimes do here) - tell him you need the "C" for the high inrush current.

BTW - the guy with the same problem was me (so I pointed you to my homepage)

Edit: sorry, it's an "L" ("Lichtstromkreise") one, the old version of the "B", so I believe it is still OK, no need to replace it, it isn't faulty, forget what I wrote about the "S". Still the same advice: get an "K" ("Kraftstromkreise", old standard) or "C" (modern standard).

FYI: https://de.wikipedia.org/wiki/Leitungsschutzschalter (sorry, it's German again)

[eamoex]

Hey capt bullshot, okay, I went back and I don't think I could deduce it was your page (a page search for "bullshot" returned zilch).

Very interesting Wikipedia page too.

Why is the trip constructed with three levers?

[fubar.gr]

@richard.cs: yes I get the 2 problems are different. To your questions:

• attached is a photo of the popping sucker; I have no clue what this thing is
• it only ever trips when switching on
• it has tripped when using the physical rocker switch on the power strip AND a radio controlled outlet switching thingy I just bought (some device you plug to the wall outlet and then your device to — in my case, I put it between the wall outlet and the "always-switched-on" power strip)
• it never seems to trip if the PC power supply is off (not just the PC of course, I mean the rocker switch on the power supply unit itself)

@hermit: well, seeing the apparent age of the thing, you might have a point here...

This is a three phase MCB. probably made sometime between late 70s to early 80s.

Doesn't have the markings you expect to find on modern MCBs, like trip curve or max fault current and uses the older symbols. The letters U,V,W and R,S,T are seldom used anymore to denote the three phases, we use L1,L2,L3 instead.

Big PC power supplies can draw a significant inrush current due to their big input filter capacitors, and this is probably what trips the MCB. But I don't think the problem is the inrush current per se. As already noted, these can get sensitive with age and trip when no real trip conditions exist.

Have an electrician replace it with a modern 16 Amp C curve MCB and this will most probably fix the problem.

[eamoex]

Will do. Thanks for your replies all. I'll also meter the N-E resistance of stuff around the house, just for kicks. If only to get O.L. messages. 

[drussell]

You'll have to have the power off completely to get any kind of meaningful resistance readings from N-E.  If there is any power flow at all in the circuit it is going to make your resistance readings completely invalid.