This was his reply:
Dear Sir: we resolved the èroblem with a variac, which confirmed,
substantially, that our measurements were right,but I repeat that meny
more measurements will be made by the validators. A final report will be
published only after the end of the validation. At that point we will
have no more aprototype, but a certified product and the measurements
will be made directly in operation by the Customers.
Warm Regards,
Andrea Rossi
On 09/10/2012 12:33 PM, Andre Blum wrote:
And, unless I understand wrong, (and depending on the algorithm used
to give your every half-second-or-so display update), there may have
been an accidental correlation between the PWM duty cycle of the
resistive heater and the measurement cycle of the true RMS meter. This
is more likely when the PWM signal is somehow synced with the AC
cycle, (which would not be such a bad idea).
Especially now we are talking a factor 2-3 in measurements, we could
well be seeing effects like these.
Rossi confirms to me over mail that they have completed the test
without PWM and using a variac, and that they stand by their own
measurements.
Andre
On 09/10/2012 12:24 PM, Jones Beene wrote:
*From:*Andre Blum
Can anyone on this forum comment on the True RMS meter that was used?
The link Frank supplied seems to suggest that one of its applications
is to find unexpected high currents.
A true RMS meter of any kind is NOT sufficient in this situation.
A dedicated power analyzer must be used, if we are dealing with a
duty-cycle correction or spiky input, as appears to be the case. We
saw this problem clearly back years ago with Naudin's MAHG, which is
actually a very similar device to e-cat, except in the use of
tungsten instead of nickel.
Naudin, who is quite experienced with prototypes and actually worked
for EDF (French grid utility) at the time -- nevertheless measured
input power with a systemic 20:1 error. (gives one confidence in your
power bill, if you are French, n'est pas?).
How did it happen? George or Terry may have a better recollection but
IIRC Naudin was pulsing the input power at low duty. He measured
voltage and current, but the current was across a shunt and the
voltage was seen on the PS meter. The duty cycle was 5%, so to make
the duty cycle correction, Naudin then multiplied voltage x current x
20, when he should have corrected only the voltage -- as the current
was actual. Thus, he saw a most remarkable COP of 20, when it was
actually a COP of one; with a systemic error of twenty.
Actually it is not that simple -- but had Naudin used a dedicated
power analyzer, there would have been no doubt in the results, which
would have been far less remarkable. AFAIK -- despite years of
pleading that error still appears on Naudin's site.
Is Rossi (or his "expert" colonel) doing something similar? Probably.
Jones
