Electronic transformers

[ppppenguin]

In my rather limited experience of monitoring the current input to SMPS (including LED lighting) the PF correction aligns the current and voltage peaks pretty well. The current peak is also pretty symmetrical, unlike with uncorrected SPMS and linear PSUs. Unfortuantely there's still a lot of harmonics as the current tends to be taken in a big gulp around the voltage peak. I suspect this meets regulations but breaches their spirit.

[Murphyv310]

I think the confusion has come about as I certainly took it Mike was trying to measure the load current from the supply which to all intents and purposes is a high frequency square wave with varying MS ratio.
There should be no issues measuring the 50Hz AC input current with a moving iron meter even with a SMPS type of supply. I've done it many times and so long as inefficiency is taken into account it's fairly accurate.

[ppppenguin]

If the SMPS output is AC at 20kHz+ then many types of meter will measure the current (and the voltage) inaccurately. Forget about DMMs, especially true RMS ones, and moving iron. Ordinary rectifier analogue meters (AVO8 etc) may make a half decent stab at it if the output is sinusoidal which is a bit unlikely. A VTVM should be OK-ish but with the same warning about wave shape. Current is harder than voltage here as it needs a shunt which will need to be very low resistance at these currents, which are beyond a standard AVO anyway.

I've got Crompton "Tongtest" clamp meters which work on both AC and DC and probably do something close to true RMS. No idea how they'd respond to 20kHz+. Modern electronic clamp meters? No idea.

If I needed accurate measurement I'd use a scope, either with a current probe or a 4 terminal shunt. Then there's the faff of doing RMS calcs from the waveform. Easy enough if it's sine or square, much harder if it's anything else.

PS: Rather to my surprise, you can still get new Tongtest style meters: http://www.weschler.com/_upload/sitepdfs...ampons.pdf

[Mark Hennessy]

Mike was clear about it being the input to the transformer, not the output from the transformer.

Jeffrey, your Rigol scope will do a pretty good RMS conversion of an arbitrary waveform, providing you've got a sensible display on the screen.

Many DMMs extend to 100kHz or more, with a crest factor of 3 or better. The Fluke 8920A goes to 20MHz and is as true-RMS as it gets.

Back to the original question, I suspect that it's the meter that's the problem. Mike says the reading is half the expected value. Which, in round numbers, should be about 1A. But the meter linked to has a 20A FSD. Moving iron meters tend to be a bit inaccurate at the bottom end of their scale. That's probably all it is...

[ppppenguin]

I'd forgotten about the computing capabilities of modern scopes. I also didn't realise that some DMMs had such extended response. I looked up the 8920A and discovered it had a thermal converter - that really is true RMS.

Not only are MI meters less accurate at the bottom of the scale but the bottom end of the scale is also cramped which reduces the resolution. i certainly wouldn't trust measuring 1A on a 20A FSD MI meter.

[Murphyv310]

Mike was clear about it being the input to the transformer, not the output from the transformer.

Not what I see on post 5 Mark, also to an extent on post one.

Quote: Thanks for your replies, not sure if I made it clear but in this case the demonstration was setup with the current meter on the output of the electronic transformer.
I cant find the one where I saw the frequency as 22kHz but here is one which states the output as 42kHz.
http://www.electricsandlighting.co.uk/pr...ransformer
Mike

[Kalee20]

A moving iron meter made to work on 50Hz will not give a useful reading at any other frequency.
You will need something else to measure it.

It should do, actually. It may be problematic at 10Hz and below, due to needle vibration -  coming Ok again at DC, obviously.

However, 22kHz is pushing it, and I can see winding shunt capacitance, and eddy currents, all conspiring to make it read less than it should. It depends what you mean by 'useful reading'. Can you use an audio amplifier, signal generator, and a light-bulb load, to give an idea of calibration change with frequency?

[Refugee]

A moving iron meter made to work on 50Hz will not give a useful reading at any other frequency.
You will need something else to measure it.

It should do, actually. It may be problematic at 10Hz and below, due to needle vibration -  coming Ok again at DC, obviously.

However, 22kHz is pushing it, and I can see winding shunt capacitance, and eddy currents, all conspiring to make it read less than it should. It depends what you mean by 'useful reading'. Can you use an audio amplifier, signal generator, and a light-bulb load, to give an idea of calibration change with frequency?

I have a feeling that the amount of time and trouble would be beyond the scope of a quick measurement.
You would need to use a pretty hefty square wave generator to make up a calibration chart.
An audio amp would work for sine waves and would be a lot less useful for calibrating a meter that is going to measure square waves.

[Crackle]

This setup was put together for a C&G training course. They received a second meter from RS because they thought the first was faulty but the problem still existed.
The "training centre" seemed to not know why, but they got around the issue by recalibrating the meter, by drawing a new scale and inserted it on top of the old scale.

Mike

[Murphyv310]

That's very poor that a trainer wasn't either aware of the result or didn't understand the theory...... How things have changed.