Is joule thief circuit gets overunity?

[TinselKoala]

Yes, Lawrence, there is. But first you must realize that POWER is not the conserved quantity of interest when making excess ENERGY claims of overunity or COP>1.

The true energy INPUT to your system is the energy used to charge the capacitor. But the energy used by the system is what is on the capacitor, not what was used to charge it (there will be losses in the charging process that have nothing to do with the energy balance of the JT itself.)

Since the energy on the capacitor at any time is easily calculated from knowing the capacitance (fixed) and the voltage (varying with time), you can simply measure the voltage on the cap at the beginning of the run, and at the end of the run.  The energy in Joules is E=(CV²)/2, so you subtract the energy remaining on the cap at the end, from the energy on the cap at the beginning, and this will give you the number of Joules the cap supplied during the run.

Your scope provides the current and voltage output data; the time integration of the instantaneous power curve calculated from these values will give you the total output energy in Joules over the time of the test. These numbers may be legitimately compared with the energy supplied by the cap as calculated above.

A refinement will then be to measure accurately the energy used to charge the cap in the first place, which can also be done with your scope by monitoring the voltage on the cap and the current as it is charged to the desired voltage, multiplying and integrating as usual to get the number of Joules you applied to charge the cap to your test voltage.

So there are two different "input" energy measurements: that which the cap supplies to the circuit during the run, and that which was used to charge the cap in the first place. The first can be easily known by measuring the voltage before and after the run and computing the energy using the formula. The second can be known by scope monitoring, and since it is done _before_ the cap is used to run the circuit, you have time to change the scope hookup and take output power (current and voltage) readings that will be multiplied and integrated by your spreadsheet and legitimately compared to the energy used to charge the cap right before the test run.

[plengo]

That is a very interesting way of measuring the input/output. So with the Cap we would have  a definite buffer of input energy while the output we measure with the scope.


Quick question, would not the dropping voltage of the cap as we use it in the JT change the whole behavior of the JT and therefore its performance?


I had the impression that JT needs a stable input power to perform to its peak performance, no?


Great tip.

Fausto.

[TinselKoala]

That is a very interesting way of measuring the input/output. So with the Cap we would have  a definite buffer of input energy while the output we measure with the scope.


Quick question, would not the dropping voltage of the cap as we use it in the JT change the whole behavior of the JT and therefore its performance?

Yes, as the voltage decays the frequency of the JT oscillation changes. As long as you are within a certain range though, you don't notice this by looking at the light output. The light comes from short peaks in the oscillations, so as the frequency changes the duration and spacing of the peaks also changes. As I understand Lawrence's scheme, the timer allows one to set the interval between cap recharges. This is inefficient as I have pointed out because the cap is charged fully as soon as it is at the desired voltage; leaving it connected to the main power source any longer than that is wasteful. I think it should be done like this: the cap voltage should be compared to some reference voltage by a comparator. When the cap voltage reaches the preset value and the comparator flips state, the cap is disconnected from the supply and allowed to run the JT. When the cap voltage then drops to another preselected value, after say 20 minutes but this time going not by time but by energy, the cap is reconnected to the charging source for another "jolt" of energy.

I had the impression that JT needs a stable input power to perform to its peak performance, no?


Great tip.

Fausto.

Yes, that's right. Since the frequency changes with the voltage change, one would like to keep the input voltage stable at or near the optimum value. This can be done while still running primarily on the cap by tightening the "interval" using the preset comparison voltage. Keep the cap within, say, 1.2 and 1.3 volts by recharging it whenever it drops below the preset 1.2 volts. Or even tighter: 1.25 to 1.30 volts, say. Of course the tighter your voltage control, the more often you will have to recharge, but the smaller the energy input from each recharge.

[ltseung888]

Yes, Lawrence, there is. But first you must realize that POWER is not the conserved quantity of interest when making excess ENERGY claims of overunity or COP>1..............

If we have two 2 channel oscilloscopes, can we use one to measure the Input Power and one to measure the Output Power at the same time?   We can easily calculate the average power values.  Will that be a legit comparison?

[TinselKoala]

If we have two 2 channel oscilloscopes, can we use one to measure the Input Power and one to measure the Output Power at the same time?   We can easily calculate the average power values.  Will that be a legit comparison?

Yes and no. In the first place you are not concerned with POWER, but energy. Please try to remember that. The instantaneous power curve is the intermediate step to obtaining energy flow values, and it is the energy that is important.

Yes, you can take one run, monitored with two oscilloscopes as you describe, and use the integral of the instantaneous power curves for your comparison. However you will have to use statistical procedures to assure that there are no differences in the scopes or probes etc. that affect your values. This means that you will have to do multiple runs (required anyway), switching the positions of the scopes between runs, and average the values across runs, then compare the average values, including a measure of variance like standard deviation or standard error of the mean. So you might do 10 runs, with 5 having Scope A monitoring the input, and 5 having Scope B monitoring the input. This will cancel out any influence of the scopes themselves on the data. This won't allow you to do moment-by-moment comparisons, though, only the final energy numbers will have meaning.

And No, because you will not be able to synchronize the measurements between the two scopes. If you want to display graphs of the rundown as you did in your spreadsheet, moment by moment, the moments must be known precisely. Each data point for input and output power  must be made simultaneously for a true moment-by-moment comparison (one voltage and one current data point for each input and output, 4 values simultaneously). If one scope is, say, half a second behind the other one then the momentary comparison will be inaccurate, even though the final computed energy integral values may be correct for the two scopes (if you are comparing the exact same time interval, which requires synch).  Really, the 2 channel scope will make the measurements on each channel consecutively, separated by its minimum sample interval, but there is nothing to be done about that. Fortunately the error from this inherent lack of simultaneity will be small if your sample rate is high. It is important to remember that the DSO is a digital Sampling oscilloscope and it only takes one sample at a time, no matter how many channels it has or what its sample rate is. This will inevitably introduce some error into computations requiring simultaneous values on two variables like instantaneous power computations.

It is possible to synch separate oscilloscopes but take my word for it.... it is beyond your pay grade, and would only introduce unnecessary complexity in what _should_ be a relatively simple process.

It is much simpler and even more accurate to use a DMM of high impedance and accuracy to make precise readings of the capacitor voltages and use that to compute the input ENERGY to your system. Of course you must also know accurately your true capacitance value, which can also be easily determined in various ways. The best way is to set up a resonant circuit with a known inductance, find the resonant frequency, and compute the capacitance from that. Or you can let an RLC meter do the same thing and give you a number in a nice box.