torroidal transformer vendor

I think you're right John.

Graham

I did finally find the article I mentioned, April 2004 Power Electronics Technology, "Solving Inrush Problems at the Source",

Paul Mathews

I did a similar thing on one big amp: AC power series resistor to limit surge, then shorted by a triac later on. We even cheated and used it to improve transformer regulation, sort of a 1-step AC switching regulator.

John

It's a pulsed-field gradient amplifier for NMR. The TTs were to allow us to use the smallest possible fuse rating (to pass the transformer overload test) and still not blow on high peak pulse loads or toroid+capacitor powerup surges. I didn't do this personally, but I do recall that the choice of the relatively hard-to-get TTs wasn't casual.

John

Apparently it's IEC/EN 61010 2001 Edition, Clause 4.4.2.6.2 and UL 3101, 1993 Edition, Clause 14.7.2

John

Hi John:

I hit this problem (too) many years ago on a switcher - no toroid, but serious issues on inrush current "burning" switch contacts. Even back then the NTC thermistor was an often used solution, but had a major problem with time constant - there always seemed to be some timing of the user "mis-hitting" the power switch that would bypass the protection.

What I did was to put a high voltage power MOSFET inside a bridge rectifier in series with the input, arranged to always turn on, with a real simple R/transistor current sense to turn it off when the instantaneous current exceeded a set point (together with gate protection). Worked a treat. No idea if this would now fall foul of CE harmonics legislation, ask John Woodgate. There may now be easier ways to do this, but it was pretty cheap even back then.

I read in sci.electronics.design that John Larkin wrote (in ) about 'torroidal transformer vendor', on Wed, 9 Feb 2005:

Is this transformer driving a motor or something? You should never need TT fuses for an AC electronic load or a normal rectifier. Unless there is something very unusual going on, the inrush phase is, for fusing purposes, over in one cycle of the supply, and T fuses are OK for this application.

I read in sci.electronics.design that Ian wrote (in ) about 'torroidal transformer vendor', on Wed, 9 Feb 2005:

No. First 10 s after switch-on is not measured for harmonics. Clause

This is an interesting idea for a high-power toroid, where NTCs and other fixes are unattractive. You'd have to watch for spikes caused by leakage reactance when the current is cut off suddenly.

I read in sci.electronics.design that Paul Mathews wrote (in ) about 'torroidal transformer vendor', on Wed, 9 Feb

I recommend a good deal of caution when considering this paper.

I read in sci.electronics.design that Paul Mathews wrote (in ) about 'torroidal transformer vendor', on Wed, 9 Feb

I recommend a good deal of caution when considering this paper.

The fact that it attemps to encompass the issues of harmonics and EMI in one overall broad sweep shows the author's basic lack of understanding.

Graham

Bear in mind that the author's 'specially wound transformers' can be wound by anyone. It looks like they're simply offering transformers with a lower than usual working flux.

Graham

I read in sci.electronics.design that John Larkin wrote (in ) about 'torroidal transformer vendor', on Wed, 9 Feb 2005:

Unfortunately, I don't have either of those standards. Maybe someone else can comment. You are sure that the product falls within the scope of IEC/EN 61010?

I'm not sure of anything. My customer wanted our boxes tested to CE and UL, so we told them to test them themselves, and they did. They told us that the loaded-transformer-smoke-test was required.

But if the only function of a fuse is to clear shorts, why can't I just slap in 10 amp fuses and be done with it?

And if the transformer overload test isn't required, then I should be able to use a cheap transformer that will smoke and catch fire if, say, a fet in the output stage failed.

Right?

John

I read in sci.electronics.design that John Larkin wrote (in ) about 'torroidal transformer vendor', on Thu, 10 Feb 2005:

The 'smoke test' is indeed designed to prevent that.

That's not quite the point. IEC/EN 61010-1 is unusual in requiring that test, which some consider too stringent, insofar as it assumes a fault with just the critical worst-case resistance. Other safety standards use a different test.

No need; I can do the math. More like 50 amps.

Which is why, presumably, they did the test.

John

I explained it in my post of 2/8/05.

John

I can *still* do the math!

John

I'm not familiar with the specific EN you mentioned. Seems to be rather more application specific than the ones I'm used to.

Can you explain the 'transformer overload test ' ? Sounds most unusual to me.

Graham