Resistance of mains cable up to the plug socket in the wall?

[treez]

3 Amp mains cable (16/0.2mm) has a resistance of 39 milliohms per metre, what do you usually take as your figure for minimum cable resistance from the service entrance to the plug of a offline switch mode power supply?.....its got to be at least 10 metres?....ie 0.78 ohms?

(as you know, its important when thinking about inrush limitation for offline smps’s which must last for a long time)

Mains cable
https://www.farnell.com/datasheets/1829275.pdf

 

[D.A.(Tony)Stewart]

Line distribution losses are generally designed for 5% loss at rated current and 5% for feeder regulation error.

Additional drop cable resistance is your choice to meet your Vmin requirements and startup requirements and power dissipation implications on all surge passing components.

A soft start is recommended.

 

[treez]

Thanks, the subject which follows on from this is inrush current into offline smps's through hot NTC's. This inrush causes damage by

1....Surging into the electrolytic capacitor, which for a (boost) PFC'd supply means the output cap of the boost PFC stage.
2.....The overvoltage ringing that such a current surge causes when it surges through the input emi filter of the offline smps.

Which of the above would you say is worse, and are both 1 and 2 pretty much made harmless by the damping effect of the mains cable resistance in the walls leading up to the power socket into which the offline smps is plugged?

 

[Dan Mills]

In an office environment, on the floor right above the rather large transformer down from 33KV, and measuring at a socket near to the floors DB, I have seen PSC measure well over 6KA.

230V/6KA = 38 milli ohms.

I would note that there is another issue with 3A wire, you may find that the I^2T adiabatic rating means the supply MUST be fused at an overly low value. This is why 0.75mm^2 is normally the smallest size seen in kit intended for international sales where the fusing arrangements for cable protection are highly variable.

I don't think you can use the mains cable for this, but a pulse rated resistor in the power supply may be possible depending on your efficiency constraints.

As ever some modelling will revel much.

Regards, Dan.

 

[treez]

Thanks, Dan, thats a great way to look at it, -ive just seen a new house getting built right next to a transformer substation, - i bet they constantly blow fuses in any equipment they have which has an offline smps in it -due to the very high inrush currents.
I also bet they blow up the movs in such equipment too...regularly, and overvoltage the caps in the primary side of the smps due to the enormous ringing when switching on at l mains peak.

 

[Dan Mills]

I actually rather doubt the fuse thing is much of an issue, plug fuses are stupidly slow, 40A or so for a half cycle will not stress a 13A fuse.

Remember also that the common mode choke you seem so concerned about is just that, a COMMON MODE choke, the startup surge is purely differential mode so the effective inductance is **much** smaller then the nameplate one, being effectively the leakage inductance).

Finally, remember that at high frequency you are getting into transmission line effects in the network, and line Z up there will be determined by geometry not the near DC loop impedance figure.

By all means spice it, but I don't think it is a fraction of the issue you believe it to be.

At work we have however had issues with relay switched PDUs burning the contacts due to inrush (Somebody had bussed 20 laptop chargers onto one 10A rated output).

Regards, Dan.

 

[treez]

Thanks, but sorry i did not explain.... the common mode choke that we are using in our 100W offline flyback is this one....

epcos B82734R2132B30

datasheet of common mode choke
https://en.tdk.eu/inf/30/db/ind_2008/b82734r_w.pdf


..it has 250uH of differential mode (stray) inductance in it.

So there is a lot of differential mode inductance. I find this is usually the case with common mode chokes,,,they are loosely coupled.

Its a boost pfc followed by the 100w flyback....and the input stage has two of the above 47mH common mode chokes, and another 16uH common mode choke.......so theres a lot of differential mode inductance in there to ring with

 

[Dan Mills]

Have you modeled it?

The thought occurs that the PFC inductor and DC bus cap form a lowpass pole at a relatively low frequency, so the high voltage ring down may be held to the input side of the boost inductor by the action of this pole.

You may even be able to cheat a bit and place a diode and series resistance between the output of the rectifier and the DC bus cap in parallel with the boost stage which will effectively place that resistor in series with the line at switch on until the boost takes over, a few ohms rated for surge duty should help.

Time for some spice models.

Regards, Dan.

 

[treez]

Thanks, yes i have simulated it, and the answer to how high the ringing voltage goes all depends on the resistance of the mains wiring...take the attached simulation in ltspice with 300 milliohms of mains wiring resistance...this rings up to 515V...seriously overvoltaging the PFC capacitor even with the parallel bypass diode to it.

Also, the mov would be brought into action which adds to the aging of the mov. As you know, you want movs to conduct when theres a mains transient, but not every time the psu is switched on.

LTspice simualtion attached, this simulaiton shows all the inductances and resistances in the path, as well as the capacitances....the mov in this psu is directly before the mains rectifier, so the ringing magnitude there is of great concern.

 

[Dan Mills]

That thing has a few problems, and they are related.

Firstly: The enormous inrush current, best part of 100A in a 100W supply?
From this stems most of the other difficulties, thermistors while not perfect, are reasonably effective for this, or do the opto triac, mosfet or relay soft start thing.

Secondly (And this is problematic mostly because of that 100A inrush), you have the ringing in the common mode chokes, easily cured just place a 50 ohm resistor in parallel with each winding to kill the Q (Experiment with values, but fix the inrush first), given the very low impedance of the mains you should still get sufficient HF attenuation.

I would note that the excessive inrush current is probably a far more serious reliability issue then the momentary overvoltage on the main cap.

Regards, Dan.

 

[treez]

Thanks, yes I often lament the fact that electrolytic capacitor datasheets never give any indication of how inrush events affect their lifetime...I mean 100A inrush into a 82uF, 450v capacitor happening once per day, how would that affect the lifetime of that capacitor?