Blitzwolf SHP6 with no load reports apparent power value > 0 #6797
Comments
|
This might be solved with some kind of non linear function, being able to template within the GUI. |
|
Please, update. Your Tasmota version is old and this has been fixed. Thanks |
|
Despite upgrading to 6.7.1, the issue remains: On which version was this corrected? |
|
Have you performed a power calibration? |
Yes, I have. Both with a 60 W incandescent type bulb and with a 2000 W heater. They give a similar result, with the 60W bulb seeming to give a more accurate measure for the low end values of current and apparent power. I believe that would be awesome to let the user to template a calibrating curve with 3 or 4 points to fit better the non linearly from this direct measurement type plug. |
|
Please read #4727 (comment) or better the complete issue. |
I managed to re-do the calibration with the method 2 setup. Using both the 2000 W heater and the 60 W incandescent bulb, again the results seems to be quite close to the ones achieved with the method 1. The issue remains for low loads, the reactive power is quite high if calibrated with a high load (as the 2000W heater). On the opposite site, calibrating with the 60 W bulb, the 2000W heater is way off the chart, specially the reactive power measured (almost a power factor of 0.5!). |
|
The issue seems to be related to a single point calibration, instead of a feature for fitting the non-linearity with multiple points (3, e.g.). |
|
SHP 6 with no load connected |
|
What parameters have you right now for power, current and voltage? |
|
After method 2 calibration, with a 60W bulb ON: The reactive power is waving all over the place... |
|
And, for some weird reason, the PowerCal command is not working on 7.0.0.1: |
|
Radiator |
|
I remember it took me long to get it calibrated.... |
|
Although the voltage, current and active power readings are not that far away, the reactive power is. The reactive power is proportional not so much to current and voltage, but to their phase difference. @arendst, could you help explaining how reactive power is estimated? |
|
The SHP6 uses the Formulas in Now if you use command |
|
@ascillato + @arendst, there is definitely a non linearity on this plug, as for a 60W bulb I get a currentcal value of 2500 and the oil heater of 3097, for reasonably accurate active power readings. There is any way to add a template feature for best fitting the calibration for each device characteristics? Like a currentcal_low and currentcal_high? |
It seems that comes from a high current reading of nearly 130 mAr. That is weird, maybe some new board revision? I recall from the flashing steps the board has some extra hardware for better voltage insulation and some other version written, different from the images in the wiki. |
|
Is this syntax correct, please? |
|
Have you tried it? Should work. Anyway, could be simplier as Rule1
ON energy#power<=100 DO Currentcal 2500 BREAK
ON energy#power<=500 DO Currentcal 2635 BREAK
ON energy#power<=1000 DO Currentcal 2788 BREAK
ON energy#power<=1500 DO Currentcal 2942 BREAK
ON energy#power<=2000 DO Currentcal 3095 BREAK
ON energy#power<=2500 DO Currentcal 3249 BREAK
ON energy#power>2500 DO Currentcal 3402 ENDONAll together for pasting it into the console: Rule1 1
Rule1 ON energy#power<=100 DO Currentcal 2500 BREAK ON energy#power<=500 DO Currentcal 2635 BREAK ON energy#power<=1000 DO Currentcal 2788 BREAK ON energy#power<=1500 DO Currentcal 2942 BREAK ON energy#power<=2000 DO Currentcal 3095 BREAK ON energy#power<=2500 DO Currentcal 3249 BREAK ON energy#power>2500 DO Currentcal 3402 ENDON |
|
Closing this issue as you have made a workaround for the nonlinearity of your own device. Thanks for sharing your report and your work. 👍 |
ascillato2
added
good tip
BUG DESCRIPTION
Blitzwolf SHP6 with no load reports apparent power value > 0
REQUESTED INFORMATION
Backlog Template; Module; GPIO:Status 0:TO REPRODUCE
No load used.
Calibrated as the Wiki states, but with a 2000W resistive load.
EXPECTED BEHAVIOUR
Apparent power and current values should be as much close to zero as possible.
The text was updated successfully, but these errors were encountered: