Is joule thief circuit gets overunity?

Void · May 13, 2013, 01:59:22 PM

Quote from: ltseung888 on May 13, 2013, 01:08:25 AM
It looks like the use of rechargeable batteries for Board 116 or similar makes a difference in battery life (or LED lighting time).

After the batterry hits the 0.4V mark, the LED with rechargeable battery continued for at least 3 days. The LED with normal batteries died within 12 hours. I now have 8 rechargeable AA batteries to try.

That may possibly just be a characteristic of the discharge pattern of the type of rechargeable battery you are using. To see if that might be the case, you could set up a control test where you put maybe two rechargeable batteries in series with a LED and a resistor, and monitor the discharge pattern of the batteries this way as a comparison. You would want to choose a resistor value that limits the LED current to a comparable level to what your joule thief circuit runs at when the battery is fully charged. It won't be an exact comparison as in this case you would have two rechargeable batteries in series, but it might still be a valid comparison. Just an idea.

TinselKoala · May 13, 2013, 04:15:31 PM

Yes, pretty much, if your time slices are equal in duration.
Really, you are using the spreadsheet to perform a numerical approximation of an integral. The instantaneous power value is multiplied by the "duration" or the interval to the next pair of samples. This gives a tall skinny rectangular area. Then all the areas are added together to approximate the actual area "under" the smooth instantaneous power curve. This is a numerical integration and the time durations of the individual rectangles can be variable, but should obviously be as small as possible. This corresponds to the total energy in Joules (Watts x Seconds) that pass your measuring point in the total time interval. Then you divide this value by the time in seconds and you get an average power value in Watts (Joules / second) for that entire interval.

Void · May 13, 2013, 04:48:13 PM

Quote from: TinselKoala on May 13, 2013, 04:15:31 PM
Yes, pretty much, if your time slices are equal in duration.
Really, you are using the spreadsheet to perform a numerical approximation of an integral. The instantaneous power value is multiplied by the "duration" or the interval to the next pair of samples. This gives a tall skinny rectangular area. Then all the areas are added together to approximate the actual area "under" the smooth instantaneous power curve. This is a numerical integration and the time durations of the individual rectangles can be variable, but should obviously be as small as possible. This corresponds to the total energy in Joules (Watts x Seconds) that pass your measuring point in the total time interval. Then you divide this value by the time in seconds and you get an average power value in Watts (Joules / second) for that entire interval.

Ok, thanks. I will see if I can figure out how to do this with my scope, but the user manual for my scope is almost useless. I think I can figure out how to get the scope to log a channel's waveform data to a data file, but the manual doesn't explain how to set up to log both channels to a data file at the same time, nor how to start and stop data logging on demand. I will play around with it and see if I can figure it out.

ltseung888 · May 13, 2013, 05:31:13 PM

Quote from: Void on May 13, 2013, 04:48:13 PM
Ok, thanks. I will see if I can figure out how to do this with my scope, but the user manual for my scope is almost useless. I think I can figure out how to get the scope to log a channel's waveform data to a data file, but the manual doesn't explain how to set up to log both channels to a data file at the same time, nor how to start and stop data logging on demand. I will play around with it and see if I can figure it out.

Does your scope have a USB that allows the capture/save of the CSV file? If so, that CSV file should contain the Instantaneous Voltage and Current at the many sample points.

ltseung888 · May 13, 2013, 05:56:39 PM

Day 7 of the experiment. LED on Board 116 OFF totally when checked at 8:30pm on Day 6.

LED on Board 124 still ON but dimly. The experiment will be continued until the LED is OFF totally.

@TK

I am aware of my poor DSO skills. That is why I sow seeds and let others do the "proper job". One good happening is the poynt99 scope shots and the subsequent interest in getting the 4-CH Tektronics in Hong Kong. I shall NOT be the owner but may have access to it with experts helping me. So just consider my Atten Scope shots in the past, present and future as trash if you like. For me, the lighting of the LED for 3 days with a rechargeable AA battery reading 0.352V and the spike waveforms is worth further research (especially by others with the resources and skills).