In that hypothetical situation, you have an overunity device, because some power must be dissipated in the DUT. There are several methods to arrive at total power dissipated by the DUT, not including load resistor.
Can you guess how this can be done using the same test setup with a slight re-arrangement?
I designed this test strictly for black box devices as a first cut trial. Most DUT's will be quickly vetted as underunity and not require more time be spent. I'm sure this will occur with this rather poorly designed blocking oscillator.
The black box itself is exactly that, a small enclosure that will contain the circuit and allow input / output connections.
I chose the open air, resistor with taped thermocouple because it is quick responding and has fairly linear losses to air compared to the insulated water bath technique, which is more of a power integration method.
I prefer the simplicity of the open-air method of ION, especially for a quick test. But the water-bath method of MH is rather good -- I may have access to a precision calorimeter at the university... that would be fun. Note below, though -- the output power is small... for this prototype.
Thanks for posting the photo, MH. Your questions:
For starters I think it would be prudent if PhysicsProf checked with Lawrence to find out what the claimed COP is for "Prototype A." This device consists of a series of components. So the other piece of information that PhysicsProf would need to get from Lawrence is precisely which components constitute the "output." Note that there are two LEDs, some resistors, etc. If PhysicsProf could get those two questions answered by Lawrence and then post his reply in this thread, then we have a basis for doing a thermal evaluation of the energy in and energy out of this device.
I think Lawrence posted the results on the OU forum, but anyway I have this from him:
The output is between the bottom two copper wires in the photo (he numbers them 5 and 6), connecting to the output coil. I checked -- there is no direct coupling between this coil and the other wires in the prototype. The components on the output, starting at the bottom in the photo, are: 10 ohm resistor, 100 ohm resistor, and LED.
The claimed COP -- is done two ways by him, peak-to-peak (PP) and RMS. He measures the voltage output of the battery and the voltage drop (using an oscilloscope, as the DC is pulsed) across the 1.o-ohm resistor (connected to the black lead from the battery), both PP and RMS.
Then he measures the voltage across the output leads (5 and 6, described above) and to get the current, the voltage drop across the 10 ohm resistor (bottom of photo).
OK -- Input power PP is 0.117 W, RMS is 0.0023 W.
Output power PP is 0.37 W, RMS is 0.023 W.
Thus he calculates COP to be 3.2 = 0.37/0.117 from PP measurements, and from the RMS measurements, 10 = 0.023/0.0023.
As I said above, I don't like the multiplying of V*I when the voltage trace is so non-uniform, nothing like sinusoidal. And Lawrence intrinsically admits such when he says "the more accurate value from the integration of the power curves should be used." So let's be fair on that point.
Later, in a 20 Dec 2010 posting at OU, he says, "in the Workshop at Hong Kong University on Oct 9, 2010, ..one student, Felix, produced a prototype with COP > 280."
Probably calculated in a crude way, but that is not stated.
Hope that helps MH-- you see the output power is very small with this prototype A, would not warm water very fast at all.