Feed high power psu from limited mains

[najrao]

Having repaired this 60V 50A CVCC (quote link=topic=412861.msg5304994#msg5304994 date=1706521961), I have more questions before putting it to use.

First, it has no PFC to feed its 2800uF 320V** bulk storage, so needs up to 25A from my 230V mains. Too much for my home wiring.  Yes, mercifully it has inrush limiting by 3×12R-28dia NTC's and a 30A relay shorting them. I have to rustle up my own PFC to bring down the feed, to under 15A.

 I dissected a 3000W telecom rectifier I had, and tapped out its PFC output, still leaving the converter downstage connected. This should be good for the PSU. But new questions arise. Can one simply add 2800uF to the internal 800 the rectifier uses? What about its own inrush limiting components? Would they not vaporize on first power up? Or would the NTC's in the CVCC keep down the inrush well enough?
I also have a more generously rated 4000W rectifier.
** Any PFC will push these 400V rated caps to full stretch.
 
I have available 3-phase 415V good for at least 20A, and so considered a half-wave rectifier instead of PFC. Much work required to realize this. And how would the distribution transformer @250kVA like 15A dc up its star secondary's neutral? The smart energy meter may also have things to say. So, kept in abeyance for now.

Opinions please.

[najrao]

The telecom rectifier 3000W proved to be perfect. Puts out exactly 400V dc, and the CVCC PS takes this without immediate issue. I have fitted a 16A 2-pole MCB at the mains input, just sufficient for this. The standard British standard 13A socket for the dc output allows using the CVCC without PFC, i.e., directly plugged in to mains.

The inrush protection components in both the rectifier and CVCC showed no distress on power on.

[Ian.M]

I wouldn't want to be anywhere near that the first time someone flips the socket switch under load.  There wont be a second time!

High current 400V DC is extremely unforgiving and requires special switchgear and circuit breakers.  You may find something suitable made for the solar industry.

[wasedadoc]

The telecom rectifier 3000W proved to be perfect. Puts out exactly 400V dc, and the CVCC PS takes this without immediate issue. I have fitted a 16A 2-pole MCB at the mains input, just sufficient for this. The standard British standard 13A socket for the dc output allows using the CVCC without PFC, i.e., directly plugged in to mains.

The inrush protection components in both the rectifier and CVCC showed no distress on power on.

Is that UK 13A socket providing 400V DC?  Bad idea.  At least put a huge warning notice beside it.

[najrao]

Nobody other than me will ever have any access to this unit. But I shall put up a warning sticker nevertheless.

The socket switch is redundant, but I could not get rid of it. I checked what is inside, and chose an old design MK which has a fast snap action switch with over 5mm contact separation. It is not required to turn it off (or on) in usage at any time. May be I can short it out altogether, by some surgery.

Still, this switch separates an 800uF capacitor from a 2800uF one, both at exactly the  same  voltage, and full contact separation takes place in say, 1ms. There would not be any arc strike at all. True, this is my conjecture, and needs to be proved.

[Ian.M]

If you *MUST* abuse a BS1363 socket, at least fit an unswitched one!

[najrao]

True, but alas, no stock. I will see about freezing the switch closed permanently.

[Ian.M]

Tape the switch closed with a rigid cover over it, while you are trying to source an unswitched one.   

[Jwillis]

Any time you use an AC socket, switch or breaker for DC applications, you must derate from the AC rating. In many cases, switches and breakers have to be de-rated down to 10% of the AC ratings for DC applications. The DC arc on contact points will burn and in many cases weld those contacts together. Even at low voltage and current. It's a really bad idea to use AC power connections and switches for DC applications, especially at the voltage and current your demanding.