If we increase the L in our coils, can we reduce the rotor speed? I wanted to be safe with my design and that means lower RPM...

>> Anyway, back to Thane, i bet he has realized that fact by now and his countless experimentations hours but so far failed to come up with any reallife working device. (oil black ops?)

There was a situation where Thane reported OU by calculating input power as the formula P=V*I*cos(theta) where theta is the phase angle between V and I. This was with a transformer-like device, whose primary is driven into saturation so it generates odd harmonics. It wasn't a sine wave, in other words...

Thane wasn't aware that the cosine term in the formula he is using implies a sinus function, cos(theta) only works if V and I are both sine waves! If V and/or I are not pure sine waves then they have a harmonic spectrum, or a Fourier series with more than 1 term -- a list of sinus waves at many frequencies, whose sum yields the complex wave shape at hand (a nice Java app demonstrating this is at http://www.falstad.com/fourier/).

The Fourier spectrum is a series of sinus amplitudes *and* phases, so the power formula becomes a series of terms itself:

P(n) = V(n)*I(n)*cos(theta(n))

where n represents each harmonic in the Fourier expansion, so P(n) is the power at each harmonic. Then, summing all the harmonic power gives the true result. Understandably a little complex!

It's much easier to measure V and I using a digital scope that can yield both (a) the instantaneous math product V*I and (b) the mean value of that math product over some integer number of repeating cycles. As of 2011 there are USB-connected sampling scopes at reasonable prices available which can accomplish this. There are still pitfalls to watch out for (such as probe ground loops between V and I measurements, resistor parasitic inductance if I is measured as I=V/R, timing skew between V and I acquisitions, and scope amplitude and time quantization errors)... there are skills to learn before simply trusting what the scope says.

Some inventors just have power measurement errors despite their beliefs and Thane may have been in that category at one time.

-breakthrough

@ Nulpoints , and also in reply to Pm from Mikestocks 2006 , 
  Bifilar in my case is as follows . Take 2 wires , A and B . Wind them "2in hand2 side by side . do not twist the wires together , and try to avoid crossed turns . All 3 electromagnets are wound single layer . Connect the start of wire a to the power supply . Connect the END of wire A to the START of wire B. Connect the connect the END of wire B to the power supply . So current flows from start of A to the end of A which is then connected to the start of B . The source makes no mention of the shape factor of the coil , but shows a similar arrangement to above . The nails I used were three-quarter inch upholstery tacks [tin tacks ]. Smaller nails would give a more accurate measurement , my tests were quick and dirty . If I can find the source I will edit it to this post shortly . You need to try it yourself .
Source:  WWW.tesla-coil-builder.com/bifilar_electromagnet.htp      I dont do links so you will have to type it .

Hi itsu,

Yes you are right, the coils should be 15.5mH, not uH. As for reducing the voltage. That can be achieved by simply shorting the output capacitor periodically to control the max output voltage. Also using the coil-shorting technique, we can precisely regulat the output voltage to impedance match to whatever sized load we want, while still making the output look like a dead short to the input side. My only question at this point is how much power can be extracted from the system.

@Super God,

Yes, increasing L will decrease the resonant frequency, but since you will have to wrap more turns of wire, that will also increase the resistance, which will lower the Q of the coil. The simple way is to use a larger capacitor to reduce the resonant frequency. (The trouble is that adding a larger capacitor will also reduce the Q of the circuit too if it is too big). It's a bit of a balancing act. Here is a website that I have used to get a rough idea of what to use:

http://www.calctool.org/CALC/eng/electronics/RLC_circuit

- Jason O

Hi itsu,
thanks for your videos.

Quote from: itsu on July 01, 2011, 05:06:22 AM
# try to compensate the "running away" under load/short by adding a "normal" not resonating coil pair which introduce drag, but  along the way also generates extra energy (4 pairs in parallel resonance, 3 in normal?).

If it is your goal to avoid that your systems "accelerates too much" , i.e. accelerates beyond the resonance speed/RPMs, you also have the option to regulate/reduce the input voltage.

Just to mention it.
Thanks for your efforts, all.