I hope it is alright to continue a topic that has been started in another discussion (ADE7758 | Gain Calibration (xWG, xVAG and xVARG)) in this one to link the topic better to the corresponding headline.
I'm currently facing problems with the xWATTHR readings at low power factor. I performed the whole calibration process (inculding power offset calibration) but still see errors of about 2% at a power factor of about 0.2. From power factors of about 0.5 the errors are much lower and tend towards 0.1%. Let me show you an example of the errors I recognized at different power factors while performing measurements over 1 minute:
PF=0.2 | accumulation time 1 second
Phase A: 0.3280 Wh (+0.70%)
Phase B: 0.4269 Wh (+2.01%)
Phase C: 0.4157 Wh (+1.83%)
PF=0.5 | accumulation time 1 second
Phase A: 1.0984 Wh (+0.20%)
Phase B: 1.1009 Wh (+0.57%)
Phase C: 1.0620 Wh (+0.69%)
PF=0.9 | accumulation time 1 second
Phase A: 5.6317 Wh (+0.12%)
Phase B: 5.6040 Wh (+0.13%)
Phase C: 5.5639 Wh (+0.02%)
First I thought the higher error might be caused by too less data within the xWATTHR registers at a low power factor, but increasing accumulation time (even up to 30 seconds) did not solve the problem.
My second idea was that the phase error of my CTs might increase intense for currents under 1 A (PF 0.9 measurement is at ~1.5 A) and this could be the source of the larger errors, but the current is nearly the same for PF0.2 and PF0.5 measurements (0.5 - 0.6 A). So I don't think this can cause the difference of +1.4% (Phase B) in error between the two measurements, can it?
My last thought was, that something is going wrong within the active power calibration process, but I can't find out what it could be. I do power calibration at the current levels Itest = 1.70 A accumulating for 500 halfcycles and Imin = 0.05 A accumulating for 23500 halfcycles. As my source as well as the load is not ideal, current and voltage fluctuate during the measurement, so I'm using averaged values for calculation but pay attention that the fluctuation is at a low percentage. Here is my calculation of the active power offset values:
Itest = 1.7387 A (+/- 0.012%) @ 233.74 V (+/- 0.013%) --> AWATTHR = 23874d
Imin = 0.0502 A (+/- 0.14%) @ 234.21 V (+/- 0.14%) --> AWATTHR = 32580d
Offset = -189d
AWATTOS = -172d = 0xFF54
Itest = 1.7497 A (+/- 0.016%) @ 234.47 V (+/- 0.016%) --> BWATTHR = 24103d
Imin = 0.0502 A (+/- 0.18%) @ 234.19 V (+/- 0.18%) --> BWATTHR = 32757d
Offset = -263d
BWATTOS = -240d = 0xFF10
Itest = 1.7623 A (+/- 0.009%) @ 233.81 V (+/- 0.006%) --> CWATTHR = 24284d
Imin = 0.0500 A (+/- 0.15%) @ 233.55 V (+/- 0.14%) --> CWATTHR = 32734d
Offset = -362d
CWATTOS = -331d = 0xFEB5
I appreciate any hints about what might cause the "higher" errors at low PF and how to get rid of them. I'm looking forward to hearing from you guys...