Shorting coil gives back more power

[gyulasun]

Hi Mags,

I have to tell (I know you are aware of this  ) that generally it is not good to parallel rectifier diodes UNLESS you can select some out of several for nearly equal forward voltage drop at the load current. I agree with what you wrote on the paralleled LEDs.
If someone chooses a rectifier diode type that meets the current and voltage requirements for a task, then there is no problem if the diodes to be paralleled are unselected for a (nearly) uniform forward voltage drop, so paralleling them bring the benefit of less voltage drop. I agree the UF4007 type mentioned here sounds a better choice (a single UF4007 is rated 1000V at 1A and ultra fast) but studying its data sheet it has 1.7V drop at 1A current! ( http://www.fairchildsemi.com/ds/UF/UF4007.pdf )
So paralleling them is a must but maybe searching for another type is also good idea in this case, depending on the actual AC voltage-current amplitudes involved.
I have access to an old transistor curve tracer at work and tested some diodes placed in parallel that is why I wrote those mV values above to Romero. OF course the best way of reducing forward voltage drop is to use low barrier Shottky types, problem with them is they are not high voltage rated (normally under 80 - 100V). PC power supplies have them on their output side (5V, 12V rectifiers placed on heat sinks).
On the heat issue: yes, using paralleled diodes improves the dissipation burden of the individual diodes, remember also that when junction temp is increasing, forward voltage drop is reducing.

Re on MOSFETs used as diodes: see Figure 193 in PDF file Page 193 of this link for the simplest but practical full wave MOSFET rectifier:
http://www.ieeta.pt/~alex/docs/ApplicationNotes/Rectifier%20Applications%20Handbook.pdf

The drawback is you have to make a center tap on the coil and also two small extra coils for controlling (switching) the gates of the MOSFETs (dots show proper phasing of the voltages for the correct control). Basically the body diodes are shunted by the ON resistance of the MOSFET in the correct time so that the voltage drop can be estimated by Ohms Law. IF you use say a R_DSon=50 milliOhm FET, then at 2A load current the voltage drop is V_d=2A*0.05 Ohm=0.1V instead of an Si diode's 0.7-0.8V or higher voltage drop.
Here is another good (4 page) paper on this:
http://elth.ucv.ro/fisiere/anale/2007/194.pdf

A full wave bridge can also be made by 4 MOSFETs, however their control circuits becomes a bit involved. I showed an article on such here:
http://www.overunity.com/index.php?topic=6116.msg141191#msg141191  and a direct link to that Electronic Design article is here:
http://electronicdesign.com/print/power/-greener-rectifier-loses-the-diodes-adds-power-mos.aspx 

Gyula

I had seen a setup that mosfets were configured to act like diodes. It was very simple, ill see if I can find.

Maybe they could be controlled with timing.  A solid state commutator.

Thats interesting that diodes in parallel can lower the voltage drop of diodes. Never heard that till now. Im surprised that parallel works at all. If we had an led with a limit resistor, it would light when V is applied. But 2 leds paralleled with 1 resistor in series, only one would light. All due to the fact that each led is slightly off tolerance from the next. The one with the lower V drop will be the one that turns on and the other not.  Maybe it is due to the resistor and rectifier diodes are different.

How many volts are we looking at here going through the diodes? If it is high, the dif between .7 and 1 volt drops should make very little difference. But if they do work in parallel, I would have to say that amperage capabilities would help keep the diodes cool, and maybe it is heat that causes the loss. I have had some hot rectifiers before, and maybe I should go back and try increasing the amperage with parallel diodes to see if I was missing something.

Mags

[Magluvin]

Hey Guyla

Thanks for the response.  ;]

Here is another example.   I have another, looking for it.

http://mackys.livejournal.com/928372.html

Mags

[Magluvin]

This is from the link on that page.

Mags

[Magluvin]

One more. It uses a fet to replace the germanium diode in a crystal set. =]

Mags

[yssuraxu_697]

Would add a note about Tesla bifilar vs regular coils. While for drive coil one certainly wants low inductance and fast response and Tesla bifilars are in business, it is a bit mixed bag with shorting coil.

If you look from one side - when shorting there must be rapid current buildup - low inductance helps with that.
If you look from other side - when unshorting there must be large backspike - high inductance helps with that.

So it is very complex problem with multiple variables. I guess some smart dude must do a dynamic über-coil that morphes between low and high inductances inside on shorting cycle

For example in the military they get kicks out of charging up big inductance and then get EMP by shorting individual turns step by step.
http://en.wikipedia.org/wiki/File:Flux_compression_generator_2.png
Hmmm......