Is joule thief circuit gets overunity?

[ltseung888]

Input Waveform with capacitor providing energy to the LED.


Board 55 was used for the demonstration.  It has a built-in capacitor (2.3V 10F).  The Board was initially started by DC Power Supply.  When the Input Voltage as indicated by CH1 Vrms = 600mV, the DC Power was turned OFF.


Three different waveforms were captured.  They were at CH1 Vrms = 560, 500 and 480mV.  These waveforms clearly showed the crossing of the 0 axis by the Input Current. (CH2 Vavg).  The CH2 Vavg value changed from negative to positive.  Such pictures were expected from all the testers.

@Poynt99
Did you get such pictures?

[picowatt]

Lawrence, why do suppose the current wave form trace is both above and below thr 0 ref line?

Is it because of an offset in your scope? I wouldnt think that much offset would be possible. How about inductive reactance. Could that be playing a part. Let's assume for a moment that it is because current really is going in both directions. In your input current trace, wouldnt that indicate that the battery should last almost indefinitely since the average current is almost zero, or I assume could be tuned to appear as though it was zero?

.99,

That does seem to be a lot of offset, but I believe that is likely the problem.   The DC accuracy specs for these scopes is less than stellar to begin with.

At < 100mv, DC accuracy ifor the ADS1102CAL is given as:

±[3%X（|reading|+|offset|）+1% of |offset|+0.2div+2mv]

Although not specifically stated in the manual, I believe "offset" in the above relates to the setting of the the zero V baseline distance from the graticule center.

Also, in previous data dumps, the output current is given as 12ma when only 80mv is across the LED.  That is more likely an indication of +12mv of offset in the scope channel used than actual current flow thru the LED at that applied voltage.  As well, the more recent waveforms posted also tend to indicate +10 to +12mv of offset.

As the subject of scope input offset has been brought up before by you and others (as well as the wierd gain error in the compensation shots) with no response given, I doubt Lawrence will ever accept the challenge to verify his scopes with some easy to perform offset and DC accuracy measurements to prove to everyone that his scopes are reading properly.

PW

[poynt99]

PW,

I'm going to crack open the new Tek scope today so hopefully I'll have an idea if that wave form is real or not.

The average current seems to vary about the 0-ref line as Lawrence varies the supply voltage. Do you think that is consistent with a fixed scope offset?

[picowatt]

PW,

I'm going to crack open the new Tek scope today so hopefully I'll have an idea if that wave form is real or not.

The average current seems to vary about the 0-ref line as Lawrence varies the supply voltage. Do you think that is consistent with a fixed scope offset?

.99,

Look at the first waveform in post 661.  With the oscillator apparently not even running, the scope says there is 12ma flowing into the battery.  More likely, that indicated +12mv (+12ma) is the actual zero volts line.

Again, in the bottom most trace of post 665, the output current appears to be "clipping" at +12mv (+12ma) when it is more likely that the indicated +12mv is again the actual zero current point.

Combine that with the fact that the data dump for board 80 indicates that +12ma is flowing thru the LED when only 80mv is applied to it and yes, I would say that both scopes are demonstrating a consistently positive offset of about 10 to 12mv.

Possibly his LED's are something way different than anything I have around here, but I cannot find one LED, white, blue, green, red, or yellow, that draws 12ma at 80mv applied.  However, his data dump for output power for board 80 shows just that, 12ma flowing thru the LED with 80mv applied (actually, it would be 68mv applied if the indicated 12mv across the CSR is accounted for).   It is more likely that the measured 12ma is an indication of 12mv of offset in the scope channel.

As both scopes appear to demonstrate positive offset, that works out quite well for "demonstrating" OU with the way the measurements are being made (ie, the tendency to indicate less power drawn from or some actually returning to the battery, and an indication of more power being output than there really is).

Posting some shorted lead shots showing the true zero ref line and a measurement of 10 to 20mv using a known V source and couple resistors with the scopes at their typically used settings would clear all this up.  But for some reason little response regarding this is given. 

PW

.

[poynt99]

PW.

Agreed. Stay tuned for some results today.