DIY 600V AC reference voltage for calibration

[lostvoid]

I need a way to get 600V AC at 60hz for calibration on a Fluke 179 meter. I have considered using a step up transformer to do this but it appears that the only ones I can find are industrial transformers. Has anyone else faced this issue, and if so how have you handled it?

[JS]

Maybe you can wind your own transformer, you can mesure the losses and make an estimation of the error, I don't know if you can get to 0.X% uncertainty. You can wind exactly say 100 turns on one side and exactly 1000 turns on the other, no problem with that, you just do the  winding and go out twisting the leads (with some extra insulation between). Then you can mesure the primary voltage and current due to core losses, applying that current to the primary resistance would give the primary voltage, then the unloaded secondary voltage will be 10 times that.

The uncertainty then is in your measurements and the construction, how well you know the source voltage, the resistance of the primary, the primary current, etc.

JS

[SeanB]

Take 3 small 230V primary to 6VAC secondary 2VA potted transformers, and use them in reverse. Feed the parallel secondaries with your adjustable 4-6VAC, and wire the primaries ( now your secondaries) in series, so you have an adjustable AC source, with a low power capacity, but able to drive a meter input. Just make sure you wire the secondaries so they add up in voltage.

Supply can be an audio amplifier fed from a signal generator, or simply a 10VA 8VAC transformer with the primary fed with a variac. Your transformer basic insulation for the potted types is easily capable of withstanding 600VAC, as they have to be able to withstand 2kV for flash over testing in manufacture.

I did a similar thing to make some vacuum tube power supplies for a Marconi moisture meter, using what I had on hand, 220VAC to 12VAC 2VA transformers, using 5 transformers to make a 1V4 filament supply ( LM317 regulator and 2 resistors) and a 60V supply, using a string of 12V 1W zener diodes and a resistor. It worked for years, and was a lot cheaper than the weekly buying of PM3 batteries and D alkaline cells.

[Jester]

A small contribution to anyone considering the audio amp driving a transformer approach.

A while back before I had a regulated ac source I needed mains level voltage that was stable enough to allow calibration of some analog circuits measuring MAINS. So I cooked up a cheap and dirty audio amp driving a step-up transformer and it worked reasonably well. It does not include feedback from the transformer secondary so it does drift with temperature.  I crammed it all in a tiny metal box that I had laying around. This is low power probably a few VA or so.

Below is a plot using HKJ's Test Controller program (thanks again) connected to three, 34401A meters measuring: 1) Raw MAINS, 2) output from the beige box, and 3) A Clarke-Hess 828 calibrator.

So  the audio amp approach is not perfect, but a lot better than trying to use the EVER fluctuating MAINS

Note, this was from cold start, the red line drifts less over time, plot is 0-20 minutes

[fcb]

Calibrating from the mains is close to impossible, the harmonic distortion, voltage variation and noise is horrendous. 

Sinewave>audio amplifier>mains transformer in reverse is pretty good.  A couple of notes:

1. Using transformers in series to 'boost' the output voltage is fine BUT you will reach a limit where the primary<>secondary insulation voltage is insufficient for the output swing. Make sure you protect your output driver.
2. Put some resistance in series with the audio amplifier and some across your new primary, makes stabilising the amp stage easier. Also try and understand what a Zobel network is.
3. Use feedback to compensate for losses in the output if you can, or at least calibrate over a series of load conditions. Probably the best way to do it would be building your own high-res converter and digital feedback path, an RMS converter IC might also work if you can live with a really slow settling time.

If you need a 'stiff' & fast-responding AC supply then you'll really struggle with this approach.