Is joule thief circuit gets overunity?

[Neo-X]

@itsung888

I hate to say this but im also agree with conrad. Im not convinced in your circuit that it was overunity exept you made it self looped.

Imho, magnacoaster has more potential of getting overunity and replicated than joule thief.

[NickZ]

  If the circuit were truly OU the battery would not be discharging, due to it being recharged by the circuit itself. But, as you know the battery does discharge, so at best it would only be a slightly more efficient circuit. Not a self runner. So, what is the big deal?
There are already many very efficient JT circuits out there. But, no self runners that can be replicated, as yet. One that never needs to have its battery replaced for months, or years.
  The solar garden lights are very efficient, lighting not just one but several leds (4 or 5 leds), and recharging their own source battery.
Is what you are showing any better?
 

[conradelektro]

I believe the Forum Adminstrators have or have access to high end oscilloscopes that are better than the Atten Oscilloscopes in my possession.  They should be able to provide more conclusive results. 

At least 30 of such "triple checked boards" will be sent (or have been sent) to the top Universities worldwide.  They all have top end oscilloscopes.  It will take them time to come out with "conclusive statements".  Confirming Output Power > Input Power with their reputation at stake is not to be taken likely.

@Itseung888: It is possible that I make an error. Unfortunately your scope shots do not show enough detail.

Your Atten oscilloscope is good enough, you just have to set a shorter time base. The screen of your scope will then show only 2 or 3 spikes instead of the many in your posted pictures.

You should also indicate the zero line, which will enable me to do a rough average calculation. (Yes, it is important where the measurement is below or above the zero line.)

Greetings, Conrad

[conradelektro]

@Itseung888: in your  Reply #230 on: February 20, 2013, 10:06:35 PM in this thread you show your circuit and in the circuit two 1 Ohm resistors.

It is very important that these 1 Ohm resistors are precision resistors with a tolerance of e.g. 0.1 %. A Joule Thief is very efficient, near 100%, may be 98%. In case you use 1 Ohm resistors with a 5% tolerance or even with only 1% tolerance your measurements will not be right.

I have some ordinary 1 Ohm resistors and they are very imprecise. I had to get better ones with a tolerance of 0.5% (the ones with a tolerance of 0.1% are expensive, about $20.-- ).

Measuring a Joule Thief correctly is not easy and one has to be very precise and careful. I am not criticising your difficulties measuring a Joule Thief correctly (I could not do it either). What I do not like is that you claim OU without very precise and careful measurements.

You say yourself that your Joule Thieves connected to a capacitor (without battery) stop after 30 or 60 minutes. Well, that clearly indicates an efficiency below 100%.

Once you connect a battery to a good Joule Thief (and I assume that your Joule Thieves are good in the sense that their efficiency is near 100%) the run time will be in the months. And it is very difficult to measure the power stored in a battery precisely and over many days you will not see a drop in the Voltage of the battery. So, once you have a battery connected to a Joule Thief you are entering the field of almost impossible measurements. Tests and measurements which have to be conducted over months are almost impossible to do and you will not find many scientist willing to do that for you.

I would say your tests with the capacitors (instead of a battery) are about the best which can be done practically and they clearly rule out OU.

Greetings, Conrad

[ltseung888]

 It looks like many people are not aware of the full DSO analysis.  I shall use the file board38.xls in reply 224 to provide a full explanation here.

(1)    In the page ads00002out, A1 to C3 gives information about the parameter setting of the oscilloscope.  That is useful for information purposes but are not important in our analysis.
(2)    Column D provides the time of the sample from -0.0015 to +0.0015 or a total of 0.003 seconds.  This snap shot was for a duration of 0.003 seconds.
(3)    CH1 measures the Instantaneous Voltage across the LED Plus the 1 ohm resistor.  The particular 1 ohm resistor is not precision (+ or – 5%).  For our comparison purposes, such a variation is not significant.  It is also the Output Voltage across the load (LED + 1 ohm resistor).
(4)    CH2 measures the Instantaneous Current.  It is the voltage across the 1 ohm resistor.  The CH1 and CH2 values are provided by the oscilloscope in the CSV file.
(5)    CH1*CH2 provides the Instantaneous Power.  This is the most important calculation and curve indication.  In the Atten Oscilloscope, we cannot display the Instantaneous Power Curve easily but this EXCEL analysis can provide the full details.
(6)    For comparison purposes, the Input Power values were copied into column H.
(7)    The average Output Power and the average Input Power can be obtained from the average function in EXCEL.  Their ratio provides the COP.
( 8)    In examining the actual waveforms, the graph plotting of EXCEL was used.  In this case 14 peaks in the voltage curve was seen.  This means that we have at least 13 complete waves (or cycles) contained in the 0.003 seconds.  The cycles are not perfect waves but the average value is good enough for our comparison purposes.
(9)    The same discussion can be applied to the current curve (14 peaks and not perfect waves).
(10)The Output Power curve showed 14 spikes with all area in the positive region.
(11)The Input Power curve on this ads00002out page is for comparison and showed negative areas (below the zero line).
(12)We should then look at the inputadjusted page and confirm that the Input Power Curve indeed contained all negative areas.  This is the significant result of this FLEET board.  Not all JT circuits exhibit this characteristic.
(13)The BMP files are also shown in the spreadsheet for comparison purposes.  They show the same shape as the EXCEL analysis.  They did not show 14 peaks because the values displays on the RHS of the screen obstructed some peaks.

There may be minor improvements such as using higher precision resistors, better resolution oscilloscopes, more exact analysis etc.  However, the fact that the Atten Oscilloscope can show the waveform of the Instantaneous Power with the EXCEL analysis is undoubted.  Once we have these Instantaneous Power sample points, we can do the comparison.

Average Output Power > Average Input Power in this case is clearly shown.