AC/DC Power/Efficiency Meter

[jaxbird]

I'm sure we are all used to the 4 meters and a calculator drill to measure efficiency. So being tired of that, especially when trying to get an idea of efficiency under various conditions like e.g. a dc/dc converter under different loads, input voltages etc, I thought it would be useful to construct a device to do all the work.

Originally I thought DC only, 0-30V 0-5A, but then it wouldn't require much to include AC, as that allows all sorts of interesting calculations based on voltage vs current phase, real vs apparent power with a power factor.

But if doing AC it makes most sense if it will do mains voltage, then we are talking 120/230Vac, so essentially 0-260Vac or
+/-370V peaks. That will of course require some more attention to safety.

Anyway, quick overview of what I have in mind:



This is of course just a rough sketch. Also I'd probably like to add a USB connection for both charging and data collection, but to keep it safe I think best to make this double insulated, as in another DC/DC plus isolator for data, then if enclosed in a plastic box that should satisfy the requirements for a proper double insulated device.

For ADC I think 16 bits should be sufficient and maybe a decent 5milliohm shunt to achieve 10A current capability.

[jaxbird]

Looking at components, I think the AD7682 would do well http://www.analog.com/media/en/technical-documentation/data-sheets/AD7682_7689.pdf 16 bit quad SPI ADC (or dual diff) with build in reference.

And for data isolation probably something like Si8441, high speed digital isolator https://www.silabs.com/Support%20Documents/TechnicalDocs/Si844x.pdf 2.5kVrms isolation.

For DC/DC this seems a good match, Murata CME 5V/5V DC/DC isolated converter: http://power.murata.com/data/power/ncl/kdc_cme.pdf should be good for 3kV.

Current shunt, the Vishay WSL Series 5mR should do the job.

Time to order some components and do some testing.

[senso]

Why not use one of the dedicated IC's made for power measurement?
Like the ADE's from Analog Devices:
http://www.analog.com/en/products/analog-to-digital-converters/integrated-special-purpose-converters/energy-metering-ics.html

Or the ADS130E08, or a smaller brother, they implement almost everything in hardware, simultaneous sampling, filtering, calculations, etc.

[jaxbird]

Why not use one of the dedicated IC's made for power measurement?
Like the ADE's from Analog Devices:
http://www.analog.com/en/products/analog-to-digital-converters/integrated-special-purpose-converters/energy-metering-ics.html

Or the ADS130E08, or a smaller brother, they implement almost everything in hardware, simultaneous sampling, filtering, calculations, etc.

But are they not primarily AC only and fairly fixed 50/60 Hz?

I'd like to do a fairly large Hz range, maybe up to 1kHz plus pure DC of course.

[jaxbird]

Why not use one of the dedicated IC's made for power measurement?
Like the ADE's from Analog Devices:
http://www.analog.com/en/products/analog-to-digital-converters/integrated-special-purpose-converters/energy-metering-ics.html

Or the ADS130E08, or a smaller brother, they implement almost everything in hardware, simultaneous sampling, filtering, calculations, etc.

Had a closer look at your suggestions, thanks.

So for the dedicated solutions, they appear only targeted at AC, and really I would prefer to do calculations in mcu, just to avoid limitations from slightly misusing highly specialized solutions.

The ADS130E08 is a very nice (and cheap) 16 bit ADC, on the up side it's got PGA (x1, x2, x8) and simultaneous sampling plus more inputs than I need, but on the downside it's only 8ksps and a slightly large 64 pin package.

Compared to the AD7682 I had in mind, it does come in a smaller 4x4mm package and it's got 250ksps (multiplexed, not simultaneous) but it does not have PGA.

Ideally I'd want an ADC with 16 bit resolution, built in reference, 2 differential inputs (or 4/3 single ended) 20ksps simultaneous or 80ksps multiplexed, plus a PGA and serial control like SPI.