Author Topic: Power ratio over one (Read 92978 times)

gyulasun · « **Reply #195 on:** December 29, 2008, 11:45:58 PM »

Hi Wattsup,

Maybe it is only a drawing mistake but I do not like your IRF840s Source electrodes receive a positive polarity from the DC supply and their Drain electrodes get a negative one.

The correct polarity should be just the opposite, Sources get negative and Drains get positive polarities, normal for an N-channel MOSFET. (Think also of the built-in body diodes across the Drain - Source electrodes, they get a forward bias from the DC supply now in your drawing.)

Will think of a simple phase shift circuit and a pulse oscillator.

rgds, Gyula

gyulasun · « **Reply #196 on:** December 30, 2008, 03:24:08 PM »

Hi Wattsup,

I am going to show you a schematics on a simple pulse generator and phase shifter circuit but there are two more things I still have mention to you.

The first problem is you wish to use max 3V DC to feed your switches + its driving circuits if I understand your last letter well. The 3V is just the threshold voltage for the IRF840 MOSFETs to just switch them on, if you look at the data sheet, http://media.digikey.com/PDF/Data%20Sheets/ST%20Microelectronics%20PDFS/IRF840.pdf it turns out the 0.75 Ohm drain-source ON resistance comes at least with 10V gate-source switching voltage. At and around 3V it is very uncertain and its drain source ON resistance can be several Ohms or higher at and around the 3V treshold.

The second problem is connected mainly to the MOSFET type because if you consider your coils DC resistance which are very low with their few turns, well under the milliOhm range and the MOSFETsay 0.8 Ohm ON resistance, out of these two resistances in series the (I*I*R) loss is much bigger on the MOSFET, it 'eats' most of the input power and very little gets to the coils.
The remedy for this is to choose a MOSFET with much better ON resistance, though in this case there is no much choice because the best MOSFETs in this respect have 2-3 milliOhms...
Nevertheless, if you assign your test for an approximate one, what could be had as output, then go ahead of course, keeping in mind such considerations.

The Schematic partly comes from nul-points http://www.overunity.com/index.php?topic=4419.msg97648#msg97648 for the pulse oscillator and partly comes from this link http://www.tedpavlic.com/teaching/osu/ece209/lab1_intro/lab1_intro_phase_shifter.pdf , worked together by me. The data sheets for the CD4093BC and the LT1077CN8 are here:

http://www.national.com/ds/CD/CD4093BC.pdf
http://www.linear.com/pc/downloadDocument.do?navId=H0,C1,C1154,C1009,C1099,P1275,D2472

Hints for the circuit: the value of C1 is 10nF rather than 47nF, to cover higher than 13kHz.
In the phase shifter circuit the R potmeter can also be 47kOhm, be careful never go down with it to very low values (put a 4.7KOhm resistor in series with it) so that the waveform should not suffer at the output.
The value of the C capacitor here can also be a 10nF for a start. The simplicity of this circuit lets you know to figure out the amount of the phase shift only by calculation, in practice by tweaking the phase shift potmeter like you tweaked the phase shift by your software.

I suggest you run this circuit from a 9V dry battery and the separate DC supply for the MOSFETs can be set to 2-3V only, because the consumption from 9V will be under the milliamp range on the avarege, the main input power neccessity will come from the 2-3V supply to the coils.

rgds, Gyula

wattsup · « **Reply #197 on:** December 30, 2008, 04:03:44 PM »

@Stefan

Last night I tried only on the outer coil and a parallel capacitor that was fed via my FG through a small bulb trying to find the resonance frequency that would turn off the bulb (as taught by @armagdn03) and nothing up to above 2mhz. I tried all frequency scales and could not identify the resonance frequency. I will try again this time with my dc pulse generator that I usually find resonances at much lower frequencies with DC pulsing. I now have a bridge rectifier on the outer coil output.

@gyulasun

Thanks for your indications on the IRF840. I did in fact connect it as shown in the diagram with source to positive and two drains going one to each center coil and the results were as indicated. I have done this many times with the IRF840.

Now following your lead I connected the source to the power supply negative and put the feed positive on the center coil common and this also worked and I would say much much better indeed.

The voltage rises were up to 330 volts when I stopped and the rise would happen at more varied frequency mixes and phase settings. I used source voltages of 1.5 to 12 volts and the range gave me different rises speeds also.

My intention was mainly to prove the non-moving @handyguy1 concept of not rotating but simply using the pulses to shift the same polarity on each side of the inner section of the outer coil and it works exactly as I had envisioned as a non-moving replication. This will maybe inspire me to push my VAR project using an alternator stator as the outer coil and making a center non-moving coil/core. But not right now, too busy.

@handyguy1 was using two long and very strong neo magnets that developed a much higher magnetic flux and I am confident that my center coil could be made with much higher flux potential. I used 12 standard off the shelf (25 cents each) choke coils for this build and the open spacing between the inner coil ends and the outer coils would be considered very wide in motor coupling standards. So I am sure with a tighter spacing, better inner coil wire choice and a better core configuration for the inner coils will be extremely efficient. Just the fact that each inner coil is using both its north and south polarity means reduced waste of driving energy.

Also @handguy1s generator coil was a good 3 lbs. of winding and mine is only 1/2 lb with the 1500 turns.

It is really incredible how much fun you can have with two IRF840s, a computer audio card and a frequency generator program that can give both left and right frequency as well as the left and right phasing. Even though I can only go up to 20khz, when using pulsed dc, I have found that you do not require to go up into the 100khz levels as you would need with AC square and sine waves.

I will take the next day or so to find the best levels offering fastest and strongest voltage rises and once I know this, will put the final result and a final diagram and hopefully we can then look to make a circuit to do it as a battery driven stand alone with loop back.

Hopefully, by then I will also have the @groundloop circuit board via @najman100 (see @gotolucs Resonance effects thread) and once it is built, I can try it on this also with reverse polarity pulsing although I think this design will not work in such a way and simply cause field cancellations.

But in general, it is good to envision something, build it and see it react as planned. This confirms to me that thinking about your build very carefully in advance of building anything is important and things should work first in your mind, then in practice. Now I can't wait to have a circuit to run this thingy. lol

More fun to come.

Added:

@gyulasun

Just saw your last post right now. I will read it and respond or maybe some of my above information will answers some of your questions. Thanks.

gyulasun · « **Reply #198 on:** December 30, 2008, 05:31:36 PM »

Quote from: wattsup on December 30, 2008, 04:03:44 PM

@Stefan

Last night I tried only on the outer coil and a parallel capacitor that was fed via my FG through a small bulb trying to find the resonance frequency that would turn off the bulb (as taught by @armagdn03) and nothing up to above 2mhz. I tried all frequency scales and could not identify the resonance frequency. I will try again this time with my dc pulse generator that I usually find resonances at much lower frequencies with DC pulsing. I now have a bridge rectifier on the outer coil output.

Hi Wattsup,

Would like to tell you that the best method for the above problem is to use a GDO (Grid Dip Oscillator). Maybe you can borrow a GDO from amateur radio operators near to you or from a university lab.
Here is what it is, how to use:
http://home.att.net/~w4cwg/ngdip.html http://en.wikipedia.org/wiki/Grid_dip_oscillator
http://users.tpg.com.au/users/ldbutler/NegResDipMeter.htm

Tha latter link includes a negative resistance oscillator too, with which you could make your coil to oscillate very near to its natural resonant frequency (no any capacitor connected in parallel with the coil). The oscillator needs only a p and n channel junction FET (jFet) and a variable voltage supply to find its negative resistance range, nothing else and if you bring an oscilloscope probe very near to this oscillating coil to make a loose inductive coupling, the picked up resonant oscillations could be evaluated on the scope.
One notice here: your outer coil must have a low resonant selectivity because of the high DC copper resistance the many turns involve, (214 Ohm) this is the main reason why it is a bit difficult to find its resonant frequency with the method you tried, even if you use parallel capacitors.

Quote

@gyulasun

Thanks for your indications on the IRF840. I did in fact connect it as shown in the diagram with source to positive and two drains going one to each center coil and the results were as indicated. I have done this many times with the IRF840.

Well, in those cases you operated the IRF840s in reverse mode and up to 0.8-0.9 V drain-source DC supply voltages the body diode remained more or less closed in spite of the forward bias received, but at higher supply voltages the diode started to shunt (i.e. short circuit) the drain-source path.

Thanks for the further info on your findings with this circuit.

rgds, Gyula

News:

User Menu

Author Topic: Power ratio over one (Read 92978 times)

gyulasun

Re: Power ratio over one

gyulasun

Re: Power ratio over one

wattsup

Re: Power ratio over one

gyulasun

Re: Power ratio over one