Is joule thief circuit gets overunity?

[Void]

Void,

Yes, this is the way Lawrence and I measured his input power.

I already pointed out this issue some time ago in this thread, and Lawrence simply compensates for the CSR in the spreadsheet calculations. It makes the input power appear to be a little lower than it actually is.

There is no need to worry however, as all measurements have shown far less than unity, even without compensating for the CSR error.

The way you are measuring the LED power, you have the same problem there too.

Hi poynt99. Ok, I wasn't aware that Lawrence was taking steps to try to compensate for that in his calculations. As long as he is compensating for that correctly then I guess it should work out ok in the wash. Thanks for the info on that. This is not a problem with the way I measure the output power, since we actually want to measure the total output power of the load (power dissipated by both the LED and the 1 ohm series resistor). In the case of the total load (LED and the one ohm resistor) the voltage drops across the LED and resistor are additive, which is what we want. So the power measurements for the load are correct.

[Void]

Here is where I made that post:
http://www.overunity.com/12686/is-joule-thief-circuit-gets-overunity/msg356492/#msg356492

Void, Why are you OK with your Pout measurement when the very same issue is present?

It's not the same issue. See my reply above...

[poynt99]

Sure,

If you want to include the output CSR as part of the load. But none of us have done that to this point. We have considered the LED as the load.

If you want to go along that line of thinking (where the CSR is taken as part of the load), then you should be measuring and accounting for ALL the components in the circuit and consider them as an "output", excluding the battery of course.

Technically speaking (I know you hate that phrase), ALL the components excluding the battery are "output" components. The idea and hope is that one of these output devices gets most of the power. In this case, that component is the LED (which btw DOES get most of the power).

If anything, you should be more concerned about including the power in the input CSR vs. the output CSR, as the input CSR dissipates far more power than the output CSR. 

And don't forget the transistor, it has the second highest power dissipation in the circuit.

So you have a choice, either you call the LED the "load", or every component except the battery, the combined total "load".

[Void]

Sure,

If you want to include the output CSR as part of the load. But none of us have done that to this point. We have considered the LED as the load.

If you want to go along that line of thinking (where the CSR is taken as part of the load), then you should be measuring and accounting for ALL the components in the circuit and consider them as an "output", excluding the battery of course.

Technically speaking (I know you hate that phrase), ALL the components excluding the battery are "output" components. The idea and hope is that one of these output devices gets most of the power. In this case, that component is the LED (which btw DOES get most of the power).

If anything, you should be more concerned with the power in the input CSR vs. the output CSR, as the input CSR dissipates far more power than the output CSR.

From the diagram I have of Lawrence's scope probe connection method (from Feb 2013), he seems to be measuring the output power the same way I am, unless he changed the way he is measuring it since February 2013.  At any rate, measuring power dissipation across the entire load is what we want here. Regarding the efficiency measurements, what is being measured is the overall circuit efficiency, and so it is understood that power dissipation by circuit components impacts the overall circuit efficiency.  Sure, the power dissipation by the one ohm current sense resistor in the input circuit skews measurements a bit as well, but unless someone has a really good quality current probe to use to measure the input current, then we don't really have a choice to not use this resistor. It's a good point though. It might actually be more accurate then not to compensate for the voltage drop across this input current sense resistor, since it would be more accurate to subtract the power dissipated by this resistor from the input power for circuit efficiency calculation. However, as you mentioned, if the measured circuit efficiency is not within a few percent of 100% then this measurement difference is probably not too critical. If I do ever measure efficiencies close to 100%, I would certainly want to move to a much higher quality scope to improve on measurement accuracy, as the error with the scope I have now may possibly be as much as +/- 5% at these low voltage and current magnitudes (hopefully the error in my scope measurements is not more than that) .

[poynt99]

As a rule of thumb when measuring the input and "output" power of a device, unless the chosen output component (LED in this case) power is easily exceeding the measured input power, one must individually-measure ALL the circuit components (excluding the source) and sum them all to compare to the input power.

If the sum total of all the output components' power is greater than the input power, then you may have something.

In the case of Lawrence's JT, the input power "Pin" is the source or battery power.

Each "output" component to be measured for power is listed below:

1) Input CSR resistor
2) 1k base resistor
3) base inductor resistance
4) collector inductor resistance
5) transistor
6) LED
7) output CSR resistor

If (P1+P2+P3+P4+P5+P6+P7) > Pin, then 

If (P1+P2+P3+P4+P5+P6+P7) </= Pin, then