2
\$\begingroup\$

I'm looking to measure AC voltage (just the voltage, not the phase etc.) and power a raspberry pi 3 with a lora concentrator (so there would be a power draw up to 2A @ 5V) from the same ac-ac adaptor, like in the schematic: enter image description here

The main reason I don't want to use a separate power supply and sense adapter is installation inconvenience.

Would this work, if I chose a beefed up enough wall wart? I'm thinking that a 12V 24VA should be more than sufficient. And since it's using a transformer, the voltage at the adc divider should be linearly proportional to the mains.

The required resolution is 1V, so using a 10bit should be sufficient. 1V @ 240V translates to 12.5mV @ 3V and the ADC has a resolution of 3.2mV @ 3.3VRef

Edit: The 5V regulator before the raspi would most probably be a switching one

\$\endgroup\$
2
\$\begingroup\$

Yes using any calibrated “instrument transformer” can measure primary AC voltage with known verified ratio.

You may or not be interested in doing all types of measurements, peak and average then convert both to rms for a sine wave. This depends on signal conditioning in both cases for bandwidth and then sample rate. You can also use S&H and trigger events that exceed a preset threshold or a a ratio differenc and then choose to compress sampled data with statistics such as % of time below -10% of nominal.

Just a few ideas

\$\endgroup\$
  • \$\begingroup\$ Do standard wallwart ac-ac adapters count as instrument transformers? Also just to confirm, if the transformer is used to measure voltage, and at the same time it is used power the device, there should be no problem as long as it is within power limits right? E.g. if the transformer is rated 24VA and we take 20VA \$\endgroup\$ – Tomas Svitil Feb 9 '18 at 6:45
  • \$\begingroup\$ A linear voltage transformer has a defined voltage ratio for a VA rating when loaded to cause a certain tolerated temp rise. depending on insulation class. so using 20/24= 83% of the rated VA is ok . Usually winding loss is <=10% so it means your voltage will sag <=8.3% at that load. If that's too much error, you need load a much bigger better transformer or a separate one \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Nov 21 '19 at 10:25
1
\$\begingroup\$

A normal off-the-shelf transformer isn't going to work accurately for what you want. The transformer secondary will have a small resistance. So the more current you draw from it, the lower the output voltage will be, even if the input voltage remains unchanged. The problem is that the Raspberry Pi isn't guaranteed to draw a constant current.

You'd be better off using separate power supplies, so that there is no load on the transformer you use for measurements. Even then, you will need to calibrate it, as the transformer designer will have allowed for the resistance when calculating its nominal voltage at the rated current. That's unless you use a purpose-designed transformer (as suggested by Sunnyskyguy EE75).

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.