Muller Dynamo

[mariuscivic]

Hi konhead
I have tested IRFS640.  They are rated 0.18ohm but only 200V. The gen coil was ringing more but the output was much lower: 60V. My ''shorting are around 0.4 miliseconds and they are too wide. When searching for best output (under shorting) i have found that releasing the short at the peak  is best but maybe couse the shorting pulse is too wide.Right now the width is dictated by the rpm. The faster spins, the shorter are the pulses. How do we make this pulses shorter?
I'm using ferite couse is boosting the output. I'll try the no core coils

Hi Khwartz

For now let's play with a driving coil and gen coil. I did tryed to short the driving coil when is not pushing the rotor but nothing intresting so far. Just more current drawn.
The caps are filling like this:
- 56uF goes to 300V in 20 sec.
- 22uF goes to 200V in 5 sec

Another directional coil in progress with teflon insulation between the layers (once again ran out of wire)

[mariuscivic]

Just finished testing parallel mosfets.I'm usingIRF840 wich have 0.8ohm(400V / 8A). I put 10 of them and the result was not good.The voltage droped from 150 to 60V.Just like the IRFS640 wich have 0.18 ohm.(200V / 9.8A). Maybe i must change the coil...thicker wire....i dont know.

[konehead]

Hi Khwartz

you almost got the formula right - but at the very end of it you have f  2  - (cant make a small 2 on this keyboard like you wrote)...but anyways it is how many discharges within one second that you times it all by at the end of the formula - you dont square the freqpencuy or anything like that jsut times it all by discharges per second..
You cant measure current without resistance...if you have no resistance across a cap, then the cap fills up without current really - only resistance really is the cap itself then -
good thing is, to make anything resonate you need a cap, just like you need a violin body to resonate the vibrations of the strings...without a violin body, there would be no resonance, without a cap you cant resonate anything electically....so having a cap in system is good, filling it without any resistance is good, and when you hit aload with cap, disconnect the cap from its "source" when the cap hits load pretty simple eh - this way the "source" never sees the load.
I cannot  understand a single sentence that Tom Beardan writes sorry.....

[konehead]

Hi Mariu
Cant say for sure why those paralell mosfets dont work right -  one thing is the turn-off might not be like it should, since you have that crazy 100ohm resistor on one bank of mosfets and the 10K on other (I assume you are doing it like your drive coil circuit)...and you arent using a driver chip too.
I will bet you are going to eventually need to use a driver chip for paralell mosfets and to improve your system and make it reliable.......
500V rated mosfets are proabably the minimum voltage you should use.
Maybe those paralell mosfets you used are  have some internal diode that restricts them (dont know)
If you voltage drops like that I assume its a diode problem but dont know...other thing is the timing is way off, and you arent at the peak when you short...
mabye there is timing issues with so many mosfets in paralell - they are all doing their "own thing"
That pull-down resistor across the gate is very important in most any mosfet so that the internal cap quickly drains, and the mosfet turns off properly, and very quickly....when you have a lot in paralell, then what I do is each mosfet in the bunch has a 10K resistor from its gate to its source, plus there is another single 10K resistor that streches across the whole cluster too from the paralell-gates to the paralell-sources - so for sure paralell mosfets should turn off correctly this way..
The way I do pulse-width adjustment with hall effects and mosfets is to use a 4421 driver chip on one mosfet, (or cluster of paralell mosfets) and this will be normally OFF, swithces ON and this has its own halleffect too.
Then I have another mosfet (or cluster of paralell mosfets) with a 4422 driver chip, instead of the 4421....now this makes this mosfet (or cluster of paralell mosfets) become normally ON, and switches OFF, and this has its own hall effect too...(same type of hall effect as other)
Then the two mosfets (or cluster of paralell mosfets)  hook together in series...
now put the hall effects very close together, and have them drift apart or get closer....to make it easier you can have a halleffect trigger by another "in phase" trigger magnet so you dont have to have them right together....
theres no way I could do the 100% no lenz coil shorting without the super narrow pulse width...on the scope, it is just a narrow slit no real "pulse" to it...at 60hz it is .25 millisecond or less....
I made a couple directional drive coils today and instead of doing a wrap of telfon over the whole layer, instead I just wrapped the "pull back" portion of the wire that streches from the front back to the back with that teflon plumbers tape - so no "external" wires this way - its all inside, between each layer and the teflon tape prevents chaffing...with two layers of teflon tape per layer, it made the coil fairly fat looking - I wanted the layers to be  tighter, and closer to the ferrite cores....
Played with magnets behind the cores today in my Romeor variant machine - that is really amazing and makes no sense at all in how they improve the perfoormance and rpms....having the magnet behind the GENERATOR coils is super important, when pulsing with the drive coil...(why???) I just had single dirve coil on bottom plate as experiment - and ALL the generator coils on top plate need magnets behind them - but why is this? makes no sense, but it triples.quadruples the rpm with less draw ....
on bottom plate I have core-backing-magnets on only two of the 5 coils in the bottom plate...also this makes no sense whatsoever - why 5 magnets on top plate and only 2 on bottom?? seems logically there should be 5 on each coil on bottom plate too,  but if you put magnet on the other 3,  it slows badly....very strange...I've gone through the possiblities over an over it always comes out this way on this particular  machine - very strange but it really works great...I think this is the greatest "discovery" Romero made - the backing magnets and nowadays I am thinking this is why his machine looped itself too - and why he cant really explain "how" it did since the backing magnets are going to be very particluar and peculiar to each machine in how they make the rotor race up in speed.....
anyways I bet you could double your speed with same draw if you could get some backing magnets to work on those long ferrite cores - mabye ring-shaped neo magnets...this is what magnacoaster has - ring shaped neos on end of core...
I have  stack ring shaped magnets for my backing-magnets behind the cores, since I have a SS mounting-bolt jutting out the back that they slip around....sometimes it likes 3 stacked, and with some coils,  4 stacked for longer stronger magnet...that is the other variable with the backing magnets- how strong they are, and I control this with how many are stacked....

[mariuscivic]

Hi Konehead

That's pretty amazing what you just described about your Romero's variant. Any chance doing some video about it? My rotors are all NSNS and adding magnets behind the core of driving coil increase the rpm but with huge amount of current drain. Not efficient. I think that every setup has it's own rulles
Played again with mosfets in parallel and just dont work better; at least in my case.
Now, here is something that i'll let you really scream at me  : only the 100ohm resistor; no 10K resistors, the same voltage in output.
The rpm was around 5500 and the pulse width was just a little more than 0.1 miliseconds (only half wave shorting;ran out of triggering magnets).Lenz is still there doing his job