Archived Forum

Is it a good idea to use a 9-0-9 transformer instead of the readymade AC-AC adapter?

If you have your own enclosure that is properly and safely made, it is protected and fused appropriately, it will be OK. Otherwise, not.

If it is ok, is it good to use the same transformer for both voltage measurement and power supply(for arduino clone board) purpose?

It might be. When the reservoir capacitor charges near the peak of the mains cycle, the large current pulse drops the voltage and so the peak value you read will be lower than the true peak voltage. So if your transformer shows a fairly low impedance to the output, it might be OK. You must decide whether the error is acceptable. Many people here think it is not.

You cannot use a bridge rectifier, you must use the centre-tapped secondary and two diodes only to give you full-wave rectification, and tap off the a.c. to measure upstream of the rectifiers of course. If you use a single 9 V winding and a bridge rectifier, you cannot measure the a.c. voltage because there is no common reference point on the a.c. side of the rectifier to measure the voltage from. And if you do try to link the ground reference to the a.c. side, as some have suggested in the past, you short out one of the rectifier diodes in the bridge, and the voltage you monitor is nothing like the mains waveform.

Wouldn't it be better to generate the DC by half-wave rectifying one winding and the a/c from the other winding?

That way the a/c winding won't be loaded by the reservoir capacitor.

edited to say: either way you'd have to a/c couple the voltage signal because you won't be able apply the d/c offset to centre the a/c signal.

Maybe.

But you have twice the current at half the frequency drawn from the loaded winding, and you still have the primary impedance reflected into the "unloaded" winding, so it might or might not be less bad. (That simulation isn't accurate, it was only intended to illustrate the effect.)

My experience has been that using a transformer with two independent windings and using one for the power supply and the other for the voltage sense works OK.  I did try using the same windings on a centre tapped transformer, as you've shown, and it didn't work at all well in my application (which was a PV diverter).  What happened was that I got a voltage drop of around 6% with a load change of around 15mA on the supply, which caused an unacceptable  error in the measured power.

The circuit I used is here: http://openenergymonitor.org/emon/sites/default/files/PV%20power%20diver... (but note that it was for a different application)

JSHarris: My experience has been that using a transformer with two independent windings and using one for the power supply and the other for the voltage sense works OK. 

Yep, I would agree with that (providing that all necessary safety precautions are in place)

I still think the best solution is to use a separate switch-mode supply for the 5V, then you can be sure that the load is not distorting the waveform and you don't need to worry about how much current you take.

This is my PSU for a portable GLCD based meter  I built...

The transformer has a 400V primary so it is well away from saturation.

It's just a pity that there isn't a commercial CE approved supply like this.

As MartinR pointed, I am inclined to use a separate PSU (may be those transformer-less ones). But I am willing to know about the safety precautions that you people are mentioning.  

San

I don't know what part of the world you are from, but in the UK particularly and Europe in general, we have British and European standards that relate to the ability to touch live parts, the standards for insulation - not just at the operating voltages but in the face of high voltage transients such as lightning, protection from overcurrent in the event of internal faults and such matters.

It is illegal to sell in Europe a unit (such as the one above by MartinR) that does not carry the official CE mark. But you can make one for your own use.

Ok. Got it!

Here is my similar setup:

Three boards inside - one is power supply, the othe voltage+current sampling and the third is an Arduino clone

mharizanov, your image is not showing up, mind reposting?

Looks fine here.

And here (Firefox under Windows).

And here (Firefox on PowerPC Mac)