Archived Forum

I think the only concern is that your DC power supply design ("5V rectifier and regulator") might distort the AC waveform, which would in turn make the power calculations inaccurate.  You want your low voltage AC swing to be an exact scaled replica of what's happening on the mains.   If you DC power supply only charges up at the peaks, and briefly draws significant current when it does, it may well cause a distortion on the low voltage sine wave that doesn't exist on the mains.

OK so in short, not a great idea then :-) Thanks. I was hoping I could free up a socket being used to power the emontx. Maybe I'll need to think of another way. Perhaps just replacing the socket faceplate with power and usb, may be the easiest solution. 

You could copy the way it's done in the emonTx V3, but you'll need to do some careful optimisation of the power supply components to get enough current to run the V2 whilst not putting too big a dent into the ac wave and invalidating the voltage measurement.

Actually my main reason for wanting to do this was to know if a lower noise DV input would result in more accurate readings. And that's when I thought I could easily take the stepped down AC already coming in. 

"I thought I could easily...". Wrong! It turns out to be not easy.

The problem is the regulation of the AC adapter, and needing to use either half-wave rectification, or a second transformer or an op-amp to maintain the required isolation between the ac reference and the dc side. [Somewhere on this site there's an ill-thought through circuit by somebody who, when you re-arranged his circuit diagram, had shorted one arm of his full-wave diode bridge rectifier with a non-polarised electrolytic, then inserted a potentiometer to calibrate the voltage because the sums didn't work!] Opting for half-wave rectification, it means that there is a large current pulse which, by virtue of the adapter's regulation, dips the voltage that should be representing the ac supply. Either of the others would mean you could use full-wave rectification and halve the magnitude of the current pulse and hence the dip, but at extra cost or complexity.