Archived Forum

Try to reverse either the current transformer or the voltage sensor (wall transformer). Current and voltage seem to have opposite direction in your current setup.

 I've reversed the current transformer before reading your message, and now it seems correct. :)

For now, I will continue to calibrate the other two phases..

 EDIT: In order to obtain PF: 1.00 on Phase 2 I need to put 3.54 on phasecal2 variable, it's correct such high number??

Thank you so much!

 EDIT: In order to obtain PF: 1.00 on Phase 2 I need to put 3.54 on phasecal2 variable, it's correct such high number??

Even from my limited experience with the CTs and the abovementioned sketch it seems very unlikely that the PHASECAL values should be  differing by more than ~0.5. 

The individual CT phase shifts should be rather identical (identical construction), so the biggest influence comes from the voltage sensors phase shift. Then there is also the different sampling point (the samples are taken in ~100us intervals) which has to be taken into consideration.

You need to understand that it is not possible to obtain accurately timed voltage samples for the second and third phases - to start with, if the mains frequency shifts, then the timing changes. You need to interpolate between samples that are closest to 120° and that is exactly what phasecal does. In the standard sketch, it corrects not only for the phase difference between c.t and p.t. but also the timing difference because the two quantities are sampled consecutively. In this sketch, for the second and third phases the coarse adjustment is which sample to use, phasecal is the fine adjustment.

This entire sketch can only deliver an approximation - the limitations are clearly stated in the comment at the top of the file. If you need more precision, you must sample all three voltages and currents.