Mains overvoltage transients and inductance of the mains wiring?

[ocset]

Hello
How do we assess which value of mains wiring inductance to use for our investigation of mains transients on our outdoor electric lights?
The schematic and LTspice simulation of our situation is as attached.

The following says that impedance of the mains in UK is 0.25+j0.23 ohms
http://www.acoustica.org.uk/other/mains_Z.html
That means that the mains inductance depends on the frequency.  So which value do  we use?

[Zero999]

That article looks like audiophool material to me.

So what the mains has inductance and resistance? That will get reduced by the turns ratio of the transformer and the massive filter capacitors reduce the impedance down to the milliOhm level at audio frequencies.

As far as your lights are concerned, just use thick enough cable to keep the voltage drop low and ensure the breaker will trip in time, if there's a short circuit. If you're having problems with voltage drop, due to the impedance of the mains being too high, then contact the power grid company.

[ejeffrey]

The following says that impedance of the mains in UK is 0.25+j0.23 ohms
http://www.acoustica.org.uk/other/mains_Z.html

Other than as a typical value, this doesn't mean much.  The impedance seen at a particular outlet will vary depending on the building wiring, utility transformer, etc.  There will be an effective upper bound of course -- your local electrical code will specify minimum gauges of wire and maximum lengths for various amperage circuits.  An outlet right next to the utility panel will have less impedance than one on the far side of a large building, and a 60 amp hard wired circuit for an industrial welding robot will have less impedance than a 20 amp convenience outlet you use to run a desk lamp.

And it doesn't matter at all.  The resistance of any outlet (the real part of impedance) is low enough to not interfere with delivering the circuits rated ampacity.  The inductance (reactance) is even less important.  As you have suggested, power supply inductance describes how much the voltage drops during a load transient.  The reactance is small at the mains frequency of 50 or 60 Hz, and your power supply has to take over at higher frequencies anyway.  Remember, the voltage supplied by the mains drops to zero twice each cycle.  At that point, the power delivered is zero and your device is operating solely on power stored in the filter caps.  With a linear or non-PFC power supply, the rectifier diodes will actually be non-conducting for a substantial fraction of each cycle.  If the power supply can do that it can respond to whatever load transients you throw at it.

[T3sl4co1l]

IEC 61000-4-5 and related IEC and CISPR references contain the specifications you are looking for.

Tim

[ocset]

Thanks,
Though does anybody know what might be typical  mains wiring inductance values for say 10 metres of mains cable?
We need to know  because it is the wiring inductance which causes mains overvoltage transients......the sudden breaking of current flow in mains wiring inductance is the cause of mains overvoltage transients which can destroy equipments connected to the mains.

[T3sl4co1l]

Huh, must be pretty crappy equipment if it's dying from mains EFT?  Lightning induced surge is the bigger one (but still not a big deal, even with pretty average circuit designs).

Anyway, in either case, 61000-4 has everything you need: line models, recommended surge waveforms and networks, all that stuff.

Tim

[SeanB]

If 10m of cable is your problem then you probably want to install a mains filter and a lot of protective MOV devices in there. You get them as either an IEC inlet socket with integrated filter and fuse, and add the MOV devices after the protection so that when the MOV fails the fuse provides protection against fire.

The better solution is to use a better power supply design, so that it is not affected by small mains transients, a 10m cable is a pretty common thing to have as power connection in a consumer or commercial usage.

[ejeffrey]

Inductance of a pair of circular wires is basically mu_0 * l/pi, with a prefactor that is logarithmic in the geometry.  Logarithmic functions grow extremely slowly, so that factor can just be considered order 1.  The magnetic constant is ~1.26 microhenry/meter.  So 10 meters of cable of almost any kind will have an inductance roughly 4-10 microhenry depending on the conductor thickness and separation.

Energy stored in an inductor is 0.5 * LI^2.  So if you are drawing 20 amps, the energy in a 5 uH inductor is about 1 *millijoule*.  Utterly insignificant, and in the case that the wiring is disconnected, that energy will easily be absorbed by the tiniest input filter or supply capacitors.  Whatever is causing the problems you are seeing, it is not caused by energy stored in 10 meter of mains wiring.

Many devices connected to the mains, such as relays, transformers, and inductors have much, much higher inductance, potentially several henries, or a million times higher than the inductance of your 10 meter cable.

You can run into a problem here.  If your device has such a relay or transformer in it and you switch it off, you can generate a large surge as the energy stored in the coil has to go somewhere.  You *can* just dissipate it into the mains, in which case the mains inductance does play a role in how effectively it can clamp that spike you generated.  But really, that is just because your device is broken.  It shouldn't dump surges of power back into the mains, it should use a snubber network to absorb that energy safely.

[Zero999]

What sort of lights are they? Compact fluorescent, fluorescent or HID with a magnetic ballast or electronic ballast or LED?

Magnetic ballasts have inductances ranging from around 0.5H to 3H, depending on the power rating, much higher than the cable inductance.

[floobydust]

Calculating mains surges from scratch and working backwards with those numbers to design your equipment- it's just wrong 

At one site (restaurant/bar) after multiple failures of equipment, brought in a line-disturbance analyzer showed regular 3-5kV hits.
Traced it back to the compressor motor on the walk-in refrigerator. 10HP 3-phase motor, it made huge transients.

No amount of math or cable inductance calculations will cover that.
UL/IEC has decades of field measurements which they use for their transient-overvoltage scenarios, why not use those numbers?

[MK]

You could hire a mains disturbance analyser for a bit to find out how much the mains in bouncing around in your area. might even find one on ebay at a reasonable price. The drantez 808 was often used to track down severe transients like the one above. It has a 6KV rated measuring input, and we even had that die occasionally with the mains borne spikes.