PhysicsProf Steven E. Jones circuit shows 8x overunity ?

[albertouno]

Good Morning JouleSeeker:

Per your request, I did a quick test of my copy of your circuit (and my coil) with two different values for Rb (the base resistor).  Cb was left at 150 uf for each test. Same breadboard wiring model. The following are the results:

A. Rb=55k.  Vin=2.70v, Iin = 4ma DC, Vout across 1000 ohm load=7 v pulse peak, period=5.2us, Iin AC= pulse=10mv p-p, 0.2us duration.  LED= bright.  These are the same values I obtained with my resistor test before.  GAIN est=.47 for circuit.

B.  Rb=470K ohms.  Vin=2.68 v, Iin=0.42ma, Vout across 1000 ohm load=7 v peak, Period=60us, Iin AC=2mv peak.  LED= DIM.
    Notice, while the input power (DC based) did drop almost by a factor of 10, the output pulse PERIOD increased by a factor of 10, which means the output power should have dropped proportionally.  Hence no apparent GAIN increase for the circuit.  The longer output pulse period would also result in a lower charge on a battery, if my observations are correct.

I did not have time to do more, but not an encouraging result.  If you did get overunity with 55k before , why are you trying these higher values for Rb now?  It gives us a moving target to replicate and test.  If I got an overunity result, I would  measure each component, and freeze the circuit and test data until it was thouroughly understood and replicated, etc.  You may have already done that.

Happy testing,
Alberto

[JouleSeeker]

  Alberto --   You asked "If you did get overunity with 55k before , why are you trying these higher values for Rb now? "   As I have said before to you, I did not claim overunity, but I did show "evidence for" ou based on the Tektronix 3032 P(t) = V(t)*I(t) waveform method.  (Which method I have explained in some detail.)  Would you comment on that method compared to your method for measuring Power in and Power out?

Frankly --  have your read all of this thread?  If so, perhaps you would have a better understanding why I am using the circuit now under testing.  You will note that I am looking at other approaches and finding similarities between them which provides me/us with ideas for testing.  I expect to continue in this process, with or without your approval -- my motive being that I would like to see SOME approach  achieve repeatable and readily-measurable overunity.

Would you kindly provide a photo of your set-up, so I can see your toroid and your build?  Note that I provided a photo (as I have done with earlier versions in general).

Also, I provided the Initial and Final voltages for 1) Input batteries and 2) Output batteries -- and I asked you to do the same.  Would you?

Also,  I asked you to provide the results for the case where Rb is 4.6 megaohms (as I did) -- how bright is your LED under this condition?

I'm trying to help, but without direct responses to these straightforward requests, I'm finding it difficult. 

[albertouno]

JouleSeeker:
Here is some of what you requested, picture attached.  The LED is lit but doesn’t show up well.  A battery pack, 2.67 volts, is connected to the input, with  a resistor load of 1000 ohms.
Hard to comment on the Tektronix 3032 gain measurement without being able to replicate it myself.  I assume others have repeated it, or otherwise witnessed the actual scope measurements.  However, If that is the only data showing gain for the circuit, I would obviously be  skeptical.  I assume that the scope can handle any waveform shapes including a mix of DC and AC? 
I have read all 62 pages of the discussion following your work, and found it interesting.  You have done an excellent job describing your tests.  However, I was hoping to see some scope pictures of the output waveform  on a different time scale over one period (assuming it is periodic).  That would help me compare  to my observations and estimates of output power.
So far without capability of the tektronix, I have tried three types of tests, capacitor to capacitor, visual scope estimates using the output waveform , and the battery to battery tests.  All have their pitfalls as you describe.  No gain greater than 60% obtained with any of these.  Your quest to find other approaches is great,  I certainly wasn’t trying to be critical, but didn’t see where the Rb resistor variation test was going, particularly since I havn’t seen your output waveform shape over one period.  I don’t happen to have a resistor value of 4.6 megohms in the house, so can’t give you that info.  My LED was very dim with less than 1 Megohm.
The input battery voltages for my 1.5 hour charging period was nearly constant, 2.68 volts start, and 2.66 ending voltage.  The charging battery  voltage was 0.2V when discharged the night before the test.  It was up to 1.140 when starting the charging test, an increase similar to your experience.  That is a concern with the use of batteries. However, it was certainly depleted.  The discharge values were given before. 
Alberto

[JouleSeeker]

JouleSeeker:
Here is some of what you requested, picture attached.  The LED is lit but doesn’t show up well.  A battery pack, 2.67 volts, is connected to the input, with  a resistor load of 1000 ohms.
Hard to comment on the Tektronix 3032 gain measurement without being able to replicate it myself.  I assume others have repeated it, or otherwise witnessed the actual scope measurements.  However, If that is the only data showing gain for the circuit, I would obviously be  skeptical.  I assume that the scope can handle any waveform shapes including a mix of DC and AC? 
I have read all 62 pages of the discussion following your work, and found it interesting.  You have done an excellent job describing your tests.  However, I was hoping to see some scope pictures of the output waveform  on a different time scale over one period (assuming it is periodic).  That would help me compare  to my observations and estimates of output power.
So far without capability of the tektronix, I have tried three types of tests, capacitor to capacitor, visual scope estimates using the output waveform , and the battery to battery tests.  All have their pitfalls as you describe.  No gain greater than 60% obtained with any of these.  Your quest to find other approaches is great,  I certainly wasn’t trying to be critical, but didn’t see where the Rb resistor variation test was going, particularly since I havn’t seen your output waveform shape over one period.  I don’t happen to have a resistor value of 4.6 megohms in the house, so can’t give you that info.  My LED was very dim with less than 1 Megohm.
The input battery voltages for my 1.5 hour charging period was nearly constant, 2.68 volts start, and 2.66 ending voltage.  The charging battery  voltage was 0.2V when discharged the night before the test.  It was up to 1.140 when starting the charging test, an increase similar to your experience.  That is a concern with the use of batteries. However, it was certainly depleted.  The discharge values were given before. 
Alberto

OK -- the photo and answers are very helpful.  One sees differences between your circuit and mine right off:

1.  You show FOUR resistors; I have only ONE = the base resistor.  Each resistor represents LOSS in the system -- what are your other three resistor for?

2.  "I don’t happen to have a resistor value of 4.6 megohms in the house, so can’t give you that info.  My LED was very dim with less than 1 Megohm."  So again, with this empirical result we see a major difference...  I would ask if you can get the base resistance up to about 4 or 5 mega-ohms, so we can further check, but this result helps.

3.  A major difference may well be in the bifilar-wound inductor -- mine is toroidal, yours is rectangular. 

4.  Next, I agree we should compare waveforms.  Suggest we look at the voltage across the LED for different values of Rbase.
I'm going to post this then go and photograph the waveforms for 4.6 K-ohms (left screen-shot, which is what I ran last night), and 0.9 M-ohms (right screen-shot photo, ran earlier this week).  You will see, as one changes Rbase, the waveform changes and it is QUITE POSSIBLE that the Pout/Pin ratio will change also (that is one of the main variables I'm testing).

I also made two short vids from my runs on this circuit last evening, which I hope to upload to Youtube...  Fact is, I have never uploaded a vid to youtube personally, so any pointers on how to do that would be appreciated.

Edit-- added photo-shots of the waveforms, only change in circuit was Rbase of 4.6 K-ohms goes to 0.9 Mega-ohms.  Vertical scale = 1 volt for both waveforms.

[albertouno]

Joule Seeker:
Yes there are differences.  I see one of my postings didn’t get listed properly on the web site.  I had described the rectangular coil in more detail.  Note, the picture of the coil got cut off when uploading.  You can only see the lower half.  I tried the blue “Goldmine” purchased  toroidal coil with 12 turns, but for some reason it drew 20 ma of DC current on the input, so went to my rectangular version , which drew only 4 ma of DC current.  The toroid measured .03 millihenries (both sides),  while the rectangular one (with 16 turns) measured 0.67 millihenries (with my 2 digit meter, on  both sides).  It must be a better ferrite material?
One of the 4 resistors was a 1000 ohm load (removed for battery tests), another in series was 10 ohms for measuring output current, the third was 1.2 ohms  for measuring input current.  The base was the 4th.  I had suggested building a hard wired circuit with short soldered wires.  I just did this, and noticed a difference with the blue toroid.  Resistance of the breadboard may also be affecting these results. Somethings  to explore later.
I have to start packing for my Hawaiian trip, so won’t be able to do more for several weeks.