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

[konehead]

Hi Luc
Just to enter my two cents, here is how I think wiring should be for the rotating field in the stator if 6 pole NSNSNS:
I don't think I need drawing it is pretty simple to describe....

Wind BIFILAR wires throughout the pole slots.
One of the two bifilar wires will be power-wind, the other induced "returned-power" wind...
One of the two wires in the bifilar sends in power one way, the other sends in power the opposite direction to switch polarity of the poles very simple.

Wind this bifilar wire around for example pole #1 then stretch it over to pole #5 (in your case with 30 poles)  leave those poles (2 3 4) in between  the two energized ones (1 an 5) blank for now - later they will be filled up...
Wind both pf these two poles same direction....now you have established a N or a S pole stretching 5 poles long of length, and its all depending on N or S in which way you send in power to the respective bifilar wind.  I think so far easy to understand eh?

So now do same thing at 120 degrees so you establish 3 poles....I would think best to connect all three pole sections in series too

Do all this it again but opposite winding direction, for the remaining 3 poles of the eventual NSNSNS 6 pole rotating field.
Connect ALL the poles for example between 1 and 5 poles to ALL the other N un-wound poles....
Do same thing on all the S poles - that is to connect all those un-wound poles together  (I would think in series)
So now you have some very powerful "pickup" winds within each pole whenever energized...

For example on that 1 to 5 pole "structure" of winds, you would have two (1 and 5) being energized, and the un-energized bifilar half becomes an induced pickup wind plus you will have all the power created between those two poles 1 and 5 (2 3 4) becoming also induced pickup winds....so tie the poles together like this, now you have lots of power created to fill the caps whenever you energize the poles.

Also stretch a long length of bifilar for all three N poles, and long length for all three S poles so not so many switches needed to send in power, and to switch polarities....so for three N poles at 120 degrees, you would wind a total of six coils,, and a total of 9 pickup coils (plus the induced half of the bifilar)
For the rotating "crawling motion" of the field, you would have to envision the three poles like this: one on, then two on, then one on (like Pierre describes) except now it is not one pole "segment" at a time miving in rotation, (30 times)  but the 6 poles being rotated per total rotation....so now faster rotation needing not such fast switching....I would think having the three N poles not pulsed at same time as the S poles would give better power with less cancelling...

[konehead]

in attachment a drawing I did awhile back.
This has H switch to switch polarities of poles, instead think of bifilars winds and sending power in one direction of one half of bifilar, and the other half the other direction to switch polarity.
Also this is to energize all 6 poles at once, but maybe better to do only 3 N and then 3 S....
For "induced pickup winds" you would connect: 2, 12, 22 and 3, 13, 23  and 4, 14, 24 for red N poles
then of course 7, 17, 27 and 8, 18, 28 and 9, 19 and 29 for blue
S poles...

[pmgr]

Hi everyone
Is that a mistake on the drawing Pierre just put up message 1050?

There are two red arrows circled in yellow, and also to the right of that on bottom of circle he has the small black current flow arrows clashing too.
Seems like one of those red arrows should be blue, pointing in other direction, and the current flow small black arrows corrected.....

I think the "parallel" question is that yes he shows 6 coils in series per pole position, but those six coils will be connected in parallel to the other poles of either N or S poles. (the red arrows indicate current input and direction it comes from and these arrows are not showing the NSNSNS poles to be in series)

It is open question (to me at least) if he connects ALL coils at once, or only energizes the N poles, then the S poles independent..... I would think much better to only energize only the S blue or only the N red poles otherwise you will lose a lot right where the poles meet through canceling.

But I don't know whats going on, since I thought this drawing was for an untested idea he had for a new version...

Yes, that's a mistake. Just follow the arrows and Pierre drew some in the wrong direction.

Anyway, I agree with you that Pierre has six coils in series per pole and then six poles in parallel. If you check his first video and his fifth video, this is also what you will find. In these videos, none of his coils are isolated.

I note that this kind of configuration, although the current direction for all of the 6 poles is correct, this is indeed no good for driving the coils. Simulations show this as well. So it is a big mystery how he got a system like that looped...

PmgR

[TinselKoala]

Yes, that's a mistake. Just follow the arrows and Pierre drew some in the wrong direction.

Anyway, I agree with you that Pierre has six coils in series per pole and then six poles in parallel. If you check his first video and his fifth video, this is also what you will find. In these videos, none of his coils are isolated.

I note that this kind of configuration, although the current direction for all of the 6 poles is correct, this is indeed no good for driving the coils. Simulations show this as well. So it is a big mystery how he got a system like that looped...

PmgR

Not to me. And I suspect... not to you, really, either.

[r2fpl]

If someone is missing pins, propose to use MCP23017
It is controlled by i2C. Everyone has 16 output.

There is a library for this so the service is easy.

mcp1.digitalWrite (0 ... 16, HIGH);

mcp1..8 * 16 = max 128 output