Mains power puzzle

------------------------------

** The electro filter bank stores 125 joules so is the major contributor to total inrush surge current - but the saturating toroidal tranny produces the bigger current spike. ** Yep, it gets larger of the AC supply is very good.

** In theory maybe.

Typical household outlets in Sydney have about 0.5 to 1 ohm impedance - so a 10amp load drops the voltage at that outlet by 5 to 10 volts rms.

And a 120 amp surge drops it by 60 to 120 volts peak.

Incandescent lamps do visibly dip.

It's enough to sometimes trip a 10 amp thermal/magnetic breaker.

.... Phil

------------------------

** Both those links are simplistic drivel full of errors too.

FFS take a look at some schems for REAL bench ohm meters.

.... Phil

Carbon on the IEC inlet from a lighting strike? My first guess was the Toroid, but I see that was discussed.

I had POTs RJ11 wall plates with carbon tracks from a lighting strike cause havoc with the old DSL line.

Cheers

He said "I think pissed off users uprate the fuse to T8 or T10 amps to make it work." that seems to imply the issue affects all units, from the factory.

----------------------

** Once again:

The unit has no faults or damage.

The behaviour is 100% normal.

A quick look inside revealed what was going on.

Too simple for any of you ??

... Phil

Loose connection ?

--
Rheilly

Phil Allison wrote in news: snipped-for-privacy@googlegroups.com:

Even our DC fed HVPS had to have soft start front ends to the main oscillator or it would 'hard start' and fail to begin the oscillation.

Phil Allison wrote in news:6c2715ed-fcaa-4a9e- snipped-for-privacy@googlegroups.com:

Dried out caps. Bloated. "a quick look inside"...

Assuming 240 V at the distribution transformer, the 120 A current means a 2 ohm total loop impedance, which includes the distribution and domestic wiring as well as the amplifier itself.

So a dead short at the main socket will cause a 240 to 480 A peak current, which should blow a domestic fuse quite quickly.

Assuming total loop impedance of 2 ohms, the amplifier and it's power cord represents a 1 to 1.5 ohm impedance.

Sounds about right with that mains impedance.

Especially if you mix different types of fuses and fast I2t magnetic breakers.

However, your DC measurements with several ohms doesn't make sense. The DC measurement should just show the amplifier transformer primary copper resistance, which should be lower than the 1-1.5 ohm AC impedance.

Either there is something wrong with the AC peak measurement or the DC measurements.

-----------------------------

** 120A peak implies 2.83 ohms ( 340 / 120) The peak value of the inrush current pulse viewed on a scope, with time base stopped - or as captured on a DSO. I have both. ** Peak = "peak value" NOT max rms value.

** To you, maybe.

Cos you thinking has false assumptions.

..... Phil

---------------

** The only likely way a wound or laminated steel core can become magnetised is if a fuse blows at peak inrush current at switch on.

Even then it is a damn rare event.

..... Phil

Phil Allison wrote in news: snipped-for-privacy@googlegroups.com:

Nearly impossible to magnetize with a single stroke.

Damn rare indeed.

Well, it does happen due to hysteresis in the core. Especially large transformers. Can cause breakers to trip when plugging them in. Happens quite a bit in our shop.

I don't think you said how often this surge happened ? Is it every time the amp is plugged in ? Or is it a random happening ?

Think hysteresis. It happens.

Read original again. Repeated 120 amp surges.

Obviously, a DC current is necessary to read the real DC resistance cuz an Ohm Meter reads too high. Hmmmmm....

120A = 230V rms / 2R (roughly) so that would agree with the DC 1/2 amp measurent and 1 amp measurement (roughly) so after starting up the impedance rises. An AC resistance test might show more.

That looks to be a correct assesment.

I'm trying to guess a motive for putting semiconductors in series with the transformer primary. but with 120A inrush it sounds like a fairly hostile environment for semiconductors.

--------------------

** LOL !!

** You alone are on the right track, just one step more.

One you figure out the likely reason, you have the answer.

.... Phil

Could it be a thermistor in series with primary to limit in-rush. The DMM ohmmeter is too low current to heat it enough hence the high reading, the VOM ohmmeter current partially heats it hence the slightly lower reading. The difference between VOM readings with probes reversed is hard to explain, could the inrush thermistor partly cooled while reversing the prods?

The in-rush limiter mostly works in usage but is fooled by short duration power off blips where it can't cool down fast enough before power is restored - that might explain the sometimes blowing fuse?

piglet

I thought you made clear there ware no extra items in the path, no in-rush limiter, timer, relay, etc. Just power cord, switch, fuse and transformer primary. No defective AC indicator, defective output delay relay, etc. No complicated IEC AC-voltage-switching connector, No smps filter stuff. Yes indeed, diodes or a triac, etc., in the AC input path, could explain your weird measurements, but didn't you say there weren't any?

Can we assume you found and fixed the problem? If so, what did you replace? The power switch?

--
 Thanks, 
    - Win

Hmmm, did it use that "DC-killing" circuit with inverse parallel diodes in series with the transformer primary?

These were sometimes used because there can be a slight DC imbalance in mains (due to nasty asymmetric loads elsewhere) and DC is really bad for torroidal transformers.

Allan