Pierre's 170W in 1600W out Looped Very impressive Build continued & moderated

[stargate22]

" in Pierre video when he connect power supply to the arduino, begin first one initial sequence (i suppose that is for the core magnetization) and after that the working sequence well knowed. "
I can confirm that. It's easy to see in slow mo.
Has that something to do with all Pierres diodes going to plus?

It is strange that Pierres digital voltmeter COMES ON and lits  BEFORE the main 5 Hz Arduino sequence starts
3:rd film 10.0 min.

I believe he is talking about this....I thought the same ?

[r2fpl]

72 diodes connected Only to + and somewhere ?   (72 relays, 36 coils)

And middle connections without labels  35 empty 36 = one single point between two coils and two relays with TWO diodes in parallel ??
How are the diodes connected on the Arduino compared with Pierres connections?

35                    36
coil       relay    2xdiodes
orange|blue|black,green

[Andy71]

Hello
Try it.
Which type of diode does Pierre use?
Are sb 580 schottky diode sufficient?
80V, 5A

[forest]

Andy71


Thanks.Very valuable circuit.I wonder if we can make smaller scale model this way. How to eliminate trigger coil and turn both MPS transistors from Arduino at the same moment  ? Can you post such modified version too ?

[jerdee]

The easiest way to vary the current through all of the coils is with PWM.

We are generating a travelling wave via the rotating poles lets say at 10Hz, then via PWM we superimpose 60Hz on top of the 10Hz wave?



L192

  I believe you are on the right track! Look at Peirre's scope shot, you clearly see the variations of pulse width.  Pierre has 24 lines of code, while PMGR(LUC's code) has 20 lines of code.  BTW, the bug discussed earlier in the code, does not make a huge enough difference in the current variation.   You have three ways to approach variation.  1.  Time, 2. Amplitude, 3. Frequency.  Just like sound has three principle elements.


If we divide the code into 4 quarters, you can think of each quarter as a typical quarter wave!  Set each quarter of the code with alternating delay values.  The total freq can remain the same, in Luc's code, 20 total lines = 1 cycle or one FULL WAVE.  We already know that varying the freq. of the loop with delays varies the current draw.    Why not purposely set alternating groups of delays to simulate variation of current!!  Does this make sense?


Other thoughts,
1. Do the variation of delay values on the half wave.  Not sure this will work, but worth a try. 
2. Vary the offsets (instead of 5,5,5,5  make it 1,5,5,5,4)
3. Alternating on the the full-wave!

I don't believe variations of delay between the HIGH's and LOW's will make a difference! You don't establish WAVE, you just establish overlap control of the windings.  So you can't think of delay variation based on the switch, you have to think of it as a variation on the wave, in quarters, half, fullwave. 

Cheers,

JerDee