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

[listener192]

Thank you Pierre for the explanation.


This is my take on his four points:

1. He mentions the CHANGE of a magnetic field (which is achieved by rotating the magnetic poles at a frequency f by the Arduino) and he talks about the change in the STRENGTH of the magnetic field itself (which is achieved by the change in current through the coils). The largest current change occurs for the coils that are shorted for a short time during the Arduino overlap; before the overlap, the current through this coil is e.g. positive, during the overlap the current is constant, then after the overlap it reverses direction and becomes negative). So it changes from a positive current to a negative current (large dI/dt), so on top of the rotating magnetic field, there is an additional change in magnetic field due to the change in current direction. As a side note: generated voltage at output coil (and for BEMF) V=-dflux/dt where flux=L*I (inductance*current), so V=-(I*dL/dt+L*dI/dt), first term is caused by the spinning magnetic field (also called parametric variation), second term is caused by change of current (caused by switching of current direction).

2. He talks about how fast the magnetic field changes which is proportional to the frequency at which the field spins around (frequency f). So the faster you spin the field the larger the generated voltage. However he mentions the TIME CONSTANT (load speed) of a coil. The highest frequency a coil can respond to is f=R/(2*pi*L). He says that if you spin the field too fast, you will get a loss trying to charge coil and thus also in the return of the coil. In order to get the highest possible frequency, there is a trade-off between R and L. R is proportional to the number of windings, L is proportional to N^2 (square of N), so frequency is proportional to 1/N. So if you want a large frequency, you can't make too many windings per coil.

3. He talks about the MOVEMENT of the magnetic field here which is done by the Arduino and he says that is we need to study that first. This is why I suggested Luc go back to the small magnet and play with the Arduino code and speed it up and down to see where the TIME CONSTANT (load speed) of his coils are. For Luc's coils, inductance is about L=2mH and R=0.5ohms, so highest frequency possible is about 40Hz. This might explain why he has no output at 50Hz as the coils can't respond fast enough. Typical safety factor is a factor of 2 below the f_max, so keep the spin speed at 20Hz or below.

4. Lastly he talks about the RETURN OF THE COIL (output coil) which will be reflected back to each of the six coil sets. He says that if you collect them at the right spot (coil) and at the right point in time, it will provide useful energy instead of destroying your mosfet. He says this is the most important point. In my understanding, this will also explain why his device is not dependent on load. Actually, the more load he puts on, the higher the collected current will be.

PmgR

To look at what we have at the moment running...
The distributed (lapped) windings that a pole consists of, create an increasing and then decreasing MMF in the rotor as the pole approaches and departs, then continuing with an opposite polarity pole in the same manner. It does this without having to increase the current through the pole coil group.

As we know that does not induce much EMF in the rotor winding, there must be something else of which I have still not seen anybody clearly explain. Pierre previously referred to his system having a variation in current.

1. Did he mean a variation in current  that would result in a further variation of MMF already occuring due to the distributed windings? If so how so?

2. Or did mean something more drastic such as a complete reversal of current for the whole system, at some point i.e. swapping the relay feed rail polarity?

Regards

L192

[d3x0r]

I believe this is achieved during the overlap time of the Arduino when a coil is shorted, see my post above. The current in that particular coil will change direction, e.g. from +1 to -1 in a very short period of time (determined by L/R time of coil).
PmgR

ya could be more or less windings inbetween tap  points too ...


like here if the coils were in series too I suppose you could activate sets of them... and for some, just reverse the polarity on the tap which would generate 'instant' opposite parlaity


http://overunity.com/17653/pierres-170w-in-1600w-out-looped-very-impressive-build-continued-moderated/msg519932/#msg519932

[Jeg]

Sine modulation of the DC rail also boosts output but once again I have only tested asynchronously, so this was also cyclic.

Also I have placed a 20V AC transformer secondary in series with the 30 coils (30 slot stator), this also cyclically boosts output .

I would like to thank all the people here especially Pierre who gave us a motivation to continue questing in this field.
Jerdee i am impressed! 

L192
By placing the secondary in series with the coils isn't actually the same as modulating the dc rail? In fact this is what i had in mind to do for varying the strength of the magnetic filed. But with a diode in between transformer and coils, so to feed the whole arrangement with a varying dc power signal.
What did you chose for driving your coils? L298 boards as Luc?

Guys i have my fresh painted stator on bench and plenty of aug22 wire. I am thinking to go like Pierre's second generator so to be able to cover this method as the most of you have built his first prototype. I have a small objection in the number of turns (Pierre used 95 turns) as it will not permit frequencies higher than few tenths of Hz. Perhaps D3's middle tapping idea is what it needs for experimentation. Will see..! 

[listener192]

I would like to thank all the people here especially Pierre who gave us a motivation to continue questing in this field.
Jerdee i am impressed! 

L192
By placing the secondary in series with the coils isn't actually the same as modulating the dc rail? In fact this is what i had in mind to do for varying the strength of the magnetic filed. But with a diode in between transformer and coils, so to feed the whole arrangement with a varying dc power signal.
What did you chose for driving your coils? L298 boards as Luc?

Guys i have my fresh painted stator on bench and plenty of aug22 wire. I am thinking to go like Pierre's second generator so to be able to cover this method as the most of you have built his first prototype. I have a small objection in the number of turns (Pierre used 95 turns) as it will not permit frequencies higher than few tenths of Hz. Perhaps D3's middle tapping idea is what it needs for experimentation. Will see..!

Hi Jeg,

I have given up on the L298N boards, as soon as you increase current much over 4A I had multiple failures.

I have a BTS7960 twin bridge board on order for evaluation, but I have decided to now go with the relay boards until I can get an understanding of exactly how the recovery diodes are working, connected as Pierre has them. It looks like the current path is from ground via the remaining series coils in the group, though the coil that's been turned off then up through the diode on that side.
I would be reluctant to remove the lower diodes on a half bridge.

I went with 1000V 10A fast diodes for recovery.

Regards

L192

[onielsen]

I made the attached gif animation of how I believe the fields are switched.

Regards
Ole