In my opinion: No. For one, the Supply as positioned in the wiki image is not needed for Luc's circuit.

Your 'adviser' knows by the way that the battery is disconnected after an initial cap charge?

--
NextGen67

I think he does, and I bet he would say: after the cap is disconnected the voltage rise is due to leakage from the mosfet.

Sorry for the back and forth, but I know very very little about electronics and he is an EE so I can't really question his judgement.

It's ok, I think your 'adviser' would be right with your bet. Ha maybe you ask again to confirm and see if your bet comes true. Without the supply, the cap is not parallel with it anymore by the way [since it is not there], and the leakage from the mosfet is also not parallel with the Cap, at least not directly, but via the Gate, through the Drain.

Hmm, which I just now realize poses something.... The leakage goes from the Gate through the Drain, and then through the 6.9 ohm coil resistance, and then back into the Cap.... The cap can -with correct tuning- gain charge.

Huh, I think it should be possible to calculate how much current is leaking when we can monitor the voltage raise over the cap during a few minute period [or better said, should be able to calculate how much current goes through the circuit]? The Caps are 3900uF each at how many volt? Ok... It's way to late here. I'll sleep first and tomorrow see what I can do. (It's actually early morning so to speak already haha).

--
NextGen67

I should have asked this before, but could someone state the meaning of the letters G, D, and S in gotoluc circuit diagram?

@Harry
The letters refer to the wiring terminals (pinouts) on the MOSFET.
G-Gate
D-Drain
S-Source

http://en.wikipedia.org/wiki/Mosfet

Hi everyone,

wow... you all have been working at this today. Thank you for all of your participation.

@MarkSCoffman, that is some massive calculations blow me away. How about you just ask me to do a specific test that would take less than 10 minutes and I can make a video of it. Just let me know.

I was able to get the CNY17-3 OPTO today and I tried it. What's left of the pulse signal at the mofet gate is very ugly    @20KHz and has next to no pulse width (flat top). Much worse @30KHz. There is no way a signal like this will switch the mosfet correctly. The other problem is the 10K resistor across the gate and source to turn the mosfet off after the pulse. That alone consumes about 600 micro amps. at 20KHz let alone what the gate will be using. A higher value resistor makes the signal even worse. These cheap OPTO's is not the way to go. Mostly because a resistor is needed across the gate and source to make it work:P   We need a better solution. I know you are all looking into it, so I'll wait till the verdict is out to buy more stuff.

I also picked up a CMOS 555. From what I can measure, it seems consumes 300 micro amps, so I'll play around with it and see what it can do.

Thanks for all your help and support as without all of you this would not be possible for me alone to do. If this circuit ends up being real! the credit will go to all who participate in developing it.

Luc

@All

Here's a data point for transistor devices that better matches
the gotoluc coil system, I think. There seems to be a fairly
linear  relationship between Rds and Ciss for all mosfets.

Tmos Device = VN0300L n-channel
Rds ohms max = 1.2
Id amps = 1.0
Vgs(th) min = 0.8
Vgs (th) max = 2.5
Vdss Volts =  60
Ciss pf Max = 100
Crss pf Max  = 25
ton ns max = 30
toff ns max =30
case style TO-92

note: Rds = 1.2ohms is worse than .15ohms of IRF640
        Ciss = 100pf is much better than 1560pf of IRF640

        The question is how important is lower Rds to operation
        of the overunity energy coil? Input gate AC power
        is much less...Just a data point.

:S:MarkSCoffman

Ha, I have to re-read your previous post a few times to digest what you wrote, but thanks for the info. Your calculations might need to be interpreted a bit lossy I guess because there was not really time to get good readings regarding Cap. Depending on what you saw on the video, with or without battery makes a lot of difference.

I think the most important part is to have the lowest Ciss, thus I would prefer a Ciss of 100pF with a Rds of 1.2 Ohm above a Ciss of 1560pF with a Rds of .15 ohmm.

I guess must take some time to find the most optimal type. There must be something available like Rds<0.5 Ohm / Ciss<100pF

The thing with the opto switch is that it uses a tremendous amount of energy (compared to a non opto fet). And actually, there is nothing bad with Gate leakage, as long as we can determine how much is leaking.

Suppose we run the fet from the Cap, then there is no worries about gate leakage, since the energy would come back in the same cap again.

The VN0300L has a higher Rds and Ciss tough, according to my data sheet (still pretty low).

Ohh for ALL who would know....

I am not good with non DC resistance thing's.. Is there a way to determine the resistance of his coil at say 35Khz [or rather at his optimum frequency] ?

--
NextGen67

Gotoluc’s Real Self Runner... Gentlemen: “Let us Compute!”
 

The data used for the following calculations is taken from:
Self Running Coil Test #7 video – (Search for gotoluc on youtube.com)
Since we can’t experiment physically, we will do some theoretical
calculations.

<...>

The voltage on the bulk capacitors: = 16.64volts
Total size of the bulk capacitors =  2 x 3900uf =  7800uf
Approximated change rate in voltage on the bulk capacitors when circuit
is tuned taken from video #7 = 1/100 or   .01volts per each (1) second.

<...>

:S:MarkSCoffman

MarkSCoffman,

On the moment Luc took off the battery, the circuit became in - less then optimal- tuning, because the overall LC has changed by taking out the batteries.

So the voltage reading going up you see on the meter is not as optimal as it could have been [then when the circuit was somehow re-tuned on the moment he took of the battery].

Look at time frame 08:49 of that video.. You see a reading of 0.000061 [most optimal tuned at this point]

Then he takes off the battery and we see the charge go up. However, take a look at time frame 09:08 [you might need to replay a few times there], and you find the reading has now become 0.000031 [about half what it was before], and not in optimal tune anymore [I assume].

Due to the battery missing impedance [when they were taken out], the circuit changed in performance. [In a fact, we could say that when the battery was attached, they were re-charging also.. at least the circuit tried to do such].

--
NextGen67

Hi Luc,

First, I am sorry again I suggested the CNY17 opto for this task, as I wrote in the meantime but it must have been late. I owe you a beer or two.. 

While I agree with the idea of using a MOSFET or a swithing device with much better parameters than the IRF640 (and here is another suggestion to a MOSFET type, use them both in parallel to reduce R_DS: http://www.vishay.com/docs/71434/si1026x.pdf )

Question still arises: how your circuit works?  if there is resonance at the output side of the FET with the toroidal coil, then a FET with much less output capacitance could be tuned to resonance? surely it could but perhaps at a much higher frequency that might demand more input power (in switching circuit the higher the speed the higher the power demand gets, though this is nearly a linear dependence.) 

I have been pondering on the following: let's say 4 pick up coils (optimized) could be used around the present toroidal coil and their individial outputs are summed up either in series or parallel or both.  From this pickup power the CMOS 555 could be run.
Now by choosing a more preferred MOSFET type instead of the IRF640, the 1-2mW summed output from the pickup coils could already be enough juice for driving the gate source input, that would have less than 100pF input capacitance.  This is what I think.

rgds, Gyula

PS Luc, what is the type of the CMOS 555, please?

Your suggestion to use the si1026x Mosfet [ http://www.vishay.com/docs/71434/si1026x.pdf ] and use them in parallel mode to reduce the R_DS seems like a perfect solution to me !

I would say forget about the optical one, and go for this instead. If easy available and not to costly, it is definitely worth to pursue

Gyula you think the eventual 4 pick-up coils would deliver enough energy to power the 555 circuit AND drive the si1026x ?

In such a case this would be a very good idea, as driving anything from the Cap tank is not preferred.

Luc, lets go for this if you can

*If* the 555 is able to work and drive the moset, while using energy from the pick-up coil(s), next would be to see what the 'new' optimal tuned frequency would be.

Once resonance is back and it seem to work, the 555 could be replaced with an specific designed low-current usage design, to minimize energy use from  the pick-up coil(s).

Yes, this seems the way to go :-)

I am by the way not worried about the [eventually] higher resonance frequency.

--
NextGen67

Well, maybe finding the optimum pick up coils is the first now, I think what Luc temporarily uses is too long, the further end of the coil from the toroid can only receive much less juice, unfortunately. I think a size of his toroid core he uses now could serve as the size for  multiturn pick up coils, including the thickness as the max lenght for the coil. And ALSO I repeat here: it would be definitely worth trying to use some small piece of ferrit core inside these pick up coils, to increse their useful self inductance, hence received induced power.
<...>
The main input power to the IRF640 is needed for defeating its big input capacitance, and if any mW from the presently needed 15-16mW signal generator input could be reduced, it would be a bonus for sure.

Yes the higher operating frequency will not be the limiting factor, I mentioned it as a tendency.

rgds,  Gyula

I've send you a PM by the way Gyula, Subject: Normal Amp Draw?

A bonus indeed.. it would make more then 10mW difference !

Agreed, about the pick-up coil shape... And in his video's I saw some nice yellow colored toroid core laying around... It will be a few minute job to wind 2 halve coils on it [meaning 2 coils of each about 180 degrees... maybe make it 175 degrees].

Luc, could you try that? I would do only a single layer of not to thin awg to start with and see what it can pickup. In that way you have created 2 pick-up coils following the resonance coil quite well, and it would be easy to adjust.

If 4 pick-up coils in series would work, they would need to provide some 2,5 volt, or the 555 need more?   The gate can operate with 2.5 volt I believe.

EDIT: I remember that Luc said any ferrite material going near his resonance coil destroy the effect... In such a case probably cannot use the toroid as core material... Ha, you might want to find something that has a similar shape... Optionally, you could drill out a circle of Plexiglass and drill a hole in the center and use that as core.. [it is even a bit more 'flat', which might be better].

--
NextGen67