Partnered Output Coils - Free Energy

[TinselKoala]

Next, some scopeshots.

[TinselKoala]

More scopeshots:

(last one is with insufficient Gate drive for full mosfet action)

[EMJunkie]

@TK,

The gentlemanly thing to do would be compliment you.

Nice work! Its a nice replication!

1: If I can ask, Open Circuit can you test for spark on Output Terminals.
2: Your waves are a bit different, have you inverted them?
3: Have you tried this with 2X Coil configurations on the Partnered Output Coils? If so, was the result better or worse?
4: Does your Input drop when you load the device?

Regards

  Chris

[TinselKoala]

@EMJ: No, the channels aren't inverted BUT I discovered that I had the Primary connection to the rest of the circuit flipped, compared to Itsu's connection. Here are the traces with the Primary connection reversed from my previous posts. The spike now goes positive and hits an even higher value than before, and the trace looks very much like Itsu's except for my delay before the high-amplitude ringing, and Itsu's DC offset. I may have a DC bias issue with my scope on that channel, I'll do another measurement on the other channel for comparison. The difference in this delay may be a result of our mosfet choices. I have some other ones I can try.  I can also set up my DSO to repeat some of these measurements, which I may do later this evening.

I could probably charge a capacitor with the spike through a diode. I doubt if it will generate a spark on its own since it is so very narrow, but I'll try it.

I don't know what you mean by "2x coil configurations". I have tried it with the Partners in "aiding" mode and the input current goes way up (from about 380 mA to over 1.3 A, raw reading from inline ammeter). So that would be "worse" I think.

I have not tried "no-load" tests, that is, opencircuit in place of the 10R, yet.


These are interesting results, but as others have pointed out, the V and I measurements aren't valid for output power computations.

I hope you try my suggested test, comparing the current readings you get by simultaneously using your current clamp and the Vdrop across a 1R current-viewing resistor in series with the 10R load. You may see the result of the bandwidth issue that MarkE pointed out: The current probe may not be able to respond to sharp spikes and rapid changes in the current, whereas the Vdrop will show these much better. Please also show (or state) the channel probe attenuation settings for these tests, I'm still not sure how to interpret the voltage/current ratio of your CC-65 probe as displayed on your traces.

Here's the video showing a comparison between the detail of a _high quality (or at least very expensive!)_ current probe and Vdrop across a 1R CVR in the JT circuit, and the effect of moving a probe reference lead 3 inches along a bit of wire:
http://www.youtube.com/watch?v=KWDfrzBIxoQ

[TinselKoala]

Holy moly! I just tried an IRF830 mosfet and that makes the positive spike go to a bit over 3.3kV !!

I'm having trouble believing this, I'll have to come up with some alternative measurement for that voltage. Anyone have any suggestions?
Maybe trapping it on a capacitor will work.


ETA: ZAP! I'm believing it more now, I just got a nice little shock off the open end of a NE-2 connected with one wire at TP A.

ETA2: Mosfet gets very hot, use a good heatsink if you try this at home.