Bifilar pancake coil overunity experiment

[ayeaye]

Many thanks Partzman, i will analyze your results. Of course you didn't do the right thing, and your results are worthless. No conclusions about overunity or no overunity can be made either. I said that my experiment was about overunity in the coil, you again measured overunity in the circuit, where of course there is none. No need to test whether it's raining when one can see that it's raining. Your traces are also small and not detail enough, very difficult to measure them, not to talk accurately, so i don't know whether i can get anything useful at all from your results, but i may try.

The other thing, mosfet, it's a horrible thing and leaks as hell through its capacitances. This experiment requires that the circuit is open when the pulse is off, and no current goes through. Transistor does that almost completely, mosfet leaks as hell. The green there i think is the current, yes? It goes a lot back and forth when the pulse is off, so even when one measures any overunity there, the result is likely false, because the power came from mosfet leaking.

So far none anyhow useful results from any replications, not for and not against overunity.

I consider that there are no replications so far.

[TinselKoala]

I consider that there are no replications so far.

Yep--- especially not from YOU. You took it apart ! 

Let's see you put your circuit together again, and reproduce those scopetraces you showed in the early posts, and let's have a dialog where we examine some scope settings and probe points and component variations.

You are also wrong about your interpretation of my scope traces: As I noted, the CH1 trace has a large DC offset as anyone looking at the schematic can predict, so I displayed it AC-coupled to remove the DC offset. There is also a _lot_ of ringing which can't be resolved at the timebase setting you specified. This ringing largely is due to circuit parasitic inductances, not the coax bifilar coil.
You are also wrong about the Rigol's math capabilities. All measurements, horizontal and vertical, that the scope has available can be performed on the Math trace. The scope can do integration and differentiation of channel data, display that as a trace and then do all measurements on that math trace. It can also display a reasonable FFT, but without all the bells and whistles of a spectrum analyzer of course. The scope can export trace data in several formats via USB stick, and it can communicate over LAN or via USB with third party programs that can control the scope, display the screen and download and record and analyze data.

You are wrong about spreadsheet data analysis. A six thousand datapoint spreadsheet with multiple graphs and lots of math analysis including integration is child's play for anyone with a little experience with statistics.

And I don't know what kind of transistor you have and _neither do you_, because every datasheet we can find for C945 under several aliases (plain C945, just 945, 2sc945 and even BC945) all give the same ECB pinout -- but yours doesn't -- assuming you have actually connected it properly.


And furthermore, if you did have an actual working NPN transistor in there connected as your circuit shows, driven with a 5v pulse train to its base .... what would anyone expect to see on the CH2 (collector) signal? 

[TinselKoala]

Now, just what is a replication, according to you? I used the components you specified, connected according to a schematic you posted and then verified my redraw. No, I did not use a breadboard and a bunch of stray wiring, I soldered the circuit together in my usual style. I used the voltages and duty cycle you specified. I probed the points in an isolated manner as you specified. I used a genuine C945 transistor, verified working, connected according to its data sheet pinout. The only obvious difference is the wire I used for the coax bifilar coil. And of course I used a function generator to provide the base drive, but you said that would be OK, I thought. So how is what I did not a replication? Is it not a replication because I didn't get the scope traces that you got?

Maybe the problem is the bifilar coax coil. Can you describe it in detail? How did you make the series connection between one file and the other (between inner conductor and outer braid)? What is its DC resistance? What is its total inductance?



ETA: I am doing further tests using a different probing technique. Maybe I did it wrong the first time, let's see what I get now.

[ayeaye]

Let's see you put your circuit together again, and reproduce those scopetraces you showed in the early posts, and let's have a dialog where we examine some scope settings and probe points and component variations.

Well what sense does it make. They cannot even understand that  P = V * I . Everywhere is only waste of lies, all you get when you try to do something important. I have seen enough that already, i don't want more, i better be friend with the drunks on the street, i don't need to put up a lot of effort to be treated the same as they do. Or are you supermen, above all other people you know, doesn't matter.

You are also wrong about the Rigol's math capabilities.

I only read the manual of some simple Rigol, was it 1024, i cannot remember all these numbers, sure more expensive ones do more.

I try another thing, i saved oscilloscope screen image as c source code with gimp. This is string, modify a bit and Python eats it, i already tried. Then leave only pixels in the right color range, that is traces, then save as an xpm file that whatever graphics editor can then open.

[TinselKoala]

Well, I think you would agree that numbers are important --- whether you remember them, or got them from the correct manual or datasheet or not.

Also, it does you no good to insult and stomp your feet when someone is trying to work with you and to understand what's going on in your circuit.