By: andrea.s on Mercoledì 18 Giugno 2014 20:47
[Scusate l'Inglese ma ci siamo allargati ultimamente]
Wally:
while we wait for the Independent Third Party's Second Report, hopefully First Class this time, here is some food for thought.
It buzzed in my mind driving home tonight, and I spent some spare time on this instead of watching the World Cup, so please let me know what you think.
I will recapitulate our respective positions.
You and other nicks (e.g. Cosmonauta and Franco) argue that the PCE-830 plots show I2 in phase with V2n (i.e. a resistive load) and that the fact that I1 appears about in phase with I2 is proof (or a strong case in favor) of scam, since this can only happen with I1 clamp on L2 instead of L1 which will result in reading one quarter of the actual average power consumption.
I prefer an explanation less offensive to the authors, like: the plot of I1 was missing for whatever reason, and it was retrieved in another instance, not synchronous with V1, and added to the report. The authors declaredly only meant to illustrate that they checked for any significant harmonic content (and found none) so phase was not significant.
Now the new part.
You posted a link to a circuit diagram
[ http://www.cobraf.com/forum/immagini/thumbs/R_123555574_1.jpg ]
that shows L1 and L2 connected via TRIACs to a single resistive load, with TRIACs regulated so that short circuit never occurs, and a diagram of currents vs. time . I think it makes sense, except I don't believe the I1 clamp was misplaced.
I will now show that your circuit diagram cannot consume more than declared by the authors.
The I2 plot (which you trust) shows a current peak around 5A (4.79 to 5.12 A on the display) at the closing of the TRIAC. The effective current is 1.47A on the display.
Neglecting TRIAC drop, this is well modeled by a 25 ohm resistive load and a TRIAC regulated to be closed at 150 degrees delay from each zero-crossing, and overlapping the plots one can see a good agreement on I2 (and on I1 in magnitude). Resulting power consumption over the period is 55W. Accounting for two active phases, consumption is 110W.
This is too low to be the average condition during the 116h test: the declared consumption during ON time intervals was declared by the authors to be 910W to 930W.
We must thus assume that TRIAC (automatic) regulation during the ON intervals was allowing much more power. However, to prevent shortcircuit between V1 and V2, at most TRIACs can be closed at 120 degrees delay from each zero-crossing. In this limit condition, with the same 25 ohm load, the power consumption becomes 830W as sum of the two phases, slightly below the one declared by the authors.
In a 116 hour test with 35% duty cycle, the circuit you have posted cannot consume more than 34 kWh, to be compared to 37.6 kWh declared by the authors. Thus the report is consistent, and there are no grounds to argue that actual consumed energy was four times higher.
Don't get me wrong, I won't deny there are plenty of possibilities to cheat elsewhere. But allow me to retain some interest in this story and some trust in the people involved in the testing.