Help interpreting a design for power management unit

[jendrix]

Hello,

I have been looking to build a pmu similar to this device, it will be for domestic use.

https://www.digikey.com/en/articles/techzone/2014/jan/low-cost-microcontroller-based-phasor-measurement-units-improve-smart-grid-reliability

I am having a bit of trouble with the sampling part and was unsure whether it was a typo. It says "AC signals are measured 30 times per cycle and 1200 times per second"

If this is from a 60Hz supply then wouldn't it end up 1800 times per second?

There is also this part later on

"The microcontroller receives a filtered analog signal as input to the A/D. These signals are converted at a minimum sample rate of 1200 ksps"

Wouldn't 1200ksps be too high, surely that would be a sampling rate of 1.2MHz, I have seen other devices which sample in the KHz range.Regards

 

[CWatters]

I agree with you. They also refer to figure 1 as showing how sampling at 1200 samples per second helps show up oscillation, yet the time axis on figure 1 is in hours.

 

[Rive]

Wouldn't 1200ksps be too high, surely that would be a sampling rate of 1.2MHz, I have seen other devices which sample in the KHz range.

As far as I know digital postprocessing (especially filtering) requires quite an oversampling. But 1.2MHz indeed is a bit too much. Guess somebody there strongly believes in marketing...

They also refer to figure 1 as showing how sampling at 1200 samples per second helps show up oscillation, yet the time axis on figure 1 is in hours.

In the context of the graph it's OK.

 

[jendrix]

Thanks, so was probably a typo?

So where does the 1200 come from? Wouldn't it be 1800 samples per second?

 

[anorlunda]

I agree, 30*60=1800, not 1200.

I have been looking to build a pmu similar to this device, it will be for domestic use.

Now that's interesting. Do you have any plans? any hopes? for what to do with it when it is complete?

 

[jendrix]

I agree, 30*60=1800, not 1200.
Now that's interesting. Do you have any plans? any hopes? for what to do with it when it is complete?

It's actually for a fun project believe it or not, I'm finding it quite challenging as I have mainly done theory in the past, so transitioning that knowledge to a real life application should be fun.

If the aim were 30 samples per minute does that mean I would need to sample at 30x the mains frequency or would I be able to sample lower, reconstruct the wave and get the data from there? I was confused whether I would need to even reconstruct the waveform or could I merely just take the output from the ADC and read the binary values on a microcontroller? Regards