170 watts in - 1600 watts out - looped - Very impressive build and video

[Magluvin]

A repost from Pierre at the moderated topic 

Here is the latest from Pierre


"The only concern I have are the mosfet's you use, I'm not sure if they will withstand the the input amperage? just a single coil takes a lot of amperage, check before how much one coil draws compared to the 2 amps the mosfet's can handle.

Well I have to argue that very strongly. Todays motor controls of just about every kind is done via transistors(mosfets mostly) from the tiniest toy drones to the controllers to drive electric bicycle hub motors to electric cars. Im working with some mosfets that are 1.4mohm on resistance and can handle gobs of current compared to the relays. Can run .2ohm loads at 4v, 20 amps continuous and the transistor barely gets warm, no heat sink. But i do use a sink to be rid of as much heat developed for max performance. 3034 mosfet.

Ive seen some very small surface mount mosfets that are just amazing(to me) as to what they can do, and no heat sink needed. If any of you do the vaping instead of smoking cigs, if you have a chance to take apart your hand held vape device, you will see what Im seeing. these are used as a buck/boost converter to drive the vape coils and have digital power adjustments.

Not to come down on Pierre but I think his thinking on this needs to be corrected.

Mags

[Slider2732]

Mags, ikr, as they say. It really is a bit of a different situation if you are used to swapping FET's on RC stuff.
I can vouch for the A49T SMD MOSFETs, blew a factory fitted FET on an Eachine EO10 micro quad and the ridiculously powerful but cheap replacement is still running a year later. Altered another to run with 8mm motors with no issues and those things can draw 2A.
5.8A capable, likely with a heatsink plate of copper stuck on top with superglue or something. Or superglue a penny on top of it   
Probably older design regular sized FET's are what he is referring to, but i've ordered the N-channel A49T and P channel A19T's.


https://www.ebay.com/itm/20PCS-AO3404-A49T-5-8A-30V-SOT23-N-Channel-MOSFET-SMD-transistor-NEW-/232077638730
$1.24 free shipping for 20 !!!

https://www.ebay.com/itm/20PCS-AO3401-A19T-4-2A-30V-SOT23-P-Channel-MOSFET-SMD-Transistor-BH-/183001699892
$0.77, $0.16 shipping, again for 20 of them

[T-1000]

 Hello all,

I just spotted this thread and I was following Pierre's videos. Just I missed long discussion, wow!

For Pierre - great job! I always dreamed about creating rotating magnetic field without moving magnet or coil and you just did it!..
For Luc and other builders - great job on taking interest and trying to assemble setup. Now need to finish your rigs for actual testing if you still have spare time.

Here are my thoughts for breaking down whole thing into building blocks:

1) To make induction we have to alternate magnetic field. Without moving magnets or coils we only can achieve that when adding or removing power from additional electromagnets. Then have neutral point and polarity reversed when going around circle with switch on/off groups. This is also reflected in given numbers from Pierre. Which are 36 electromagnets with 6 poles. Which breaks down to 6 electromagnets per pole (5 active and 1 switched off to give space between poles as per Pierre's explanation video).
2) By activating 1st then 2,3,4,5th electromagnets we increase and decrease magnetic field strength so we can alternate resulting magnetic field in the core inside of stator. Which makes induction as the magnetic field strength changes.
3) When shifting starting coil number we simulating moving magnet poles
4) As the electromagnet is not treated like transformer primary we spend all current on plain wire resistance when each coil is activated. Which means when we put load on or off on the center core coil it does not lower electromagnet inductance any more as we relay on coil wire resistance already.
5) By collecting BEMF after each coil is switching off or getting induction from neighbour coil we recover partial energy back to the bank of super-capacitors.
6) The generated power is looped back to the input and the excessive power is taken to the load.

Now to my questions and things required to test so we can understand exactly what is going on:
1) On the BEMF/inductance recovery.  Luc and others - can you try to power on coils in the sequence I mentioned and see what amount of power is getting to recover? This defines overall COP of the system before we start talking about induction part. This is very important so we will know how much power in percentage is required to make system self running.
2) How much power you can induce after smoothing over shotky HV diode bridge and high voltage cap?
3) If COP of the system before looping back is less than 70% - do we miss any crucial steps when making magnetic field increase/decrease in 6 steps resolution? Like too long duty cycle on coils or something like that...
4) Would Pierre elaborate on my post and give some comments on his view of the whole thing?

So, any volunteers for making tests?

Cheers!
T-1000

[T-1000]

When we are talking about rotating .magnetic field it is huge difference when creating it in stepping motor style with DC and the rotor is actually is not moving. It cannot be compared to polyphase motor as the effects on driving coils are different from the sine wave AC.
With what was done here would require 6 phase AC which is not doable so easy.

[Magluvin]

Mags, ikr, as they say. It really is a bit of a different situation if you are used to swapping FET's on RC stuff.
I can vouch for the A49T SMD MOSFETs, blew a factory fitted FET on an Eachine EO10 micro quad and the ridiculously powerful but cheap replacement is still running a year later. Altered another to run with 8mm motors with no issues and those things can draw 2A.
5.8A capable, likely with a heatsink plate of copper stuck on top with superglue or something. Or superglue a penny on top of it   
Probably older design regular sized FET's are what he is referring to, but i've ordered the N-channel A49T and P channel A19T's.


https://www.ebay.com/itm/20PCS-AO3404-A49T-5-8A-30V-SOT23-N-Channel-MOSFET-SMD-transistor-NEW-/232077638730
$1.24 free shipping for 20 !!!

https://www.ebay.com/itm/20PCS-AO3401-A19T-4-2A-30V-SOT23-P-Channel-MOSFET-SMD-Transistor-BH-/183001699892
$0.77, $0.16 shipping, again for 20 of them

Hey Slider.  Good to see you around here.

My examples were not Hv type but that can also be done.

here is the 3034 data sheet below

Applications

DC Motor Drive

High Efficiency Synchronous Rectification in SMPS

Uninterruptible Power Supply

High Speed Power Switching

Hard Switched and High Frequency Circuits

Mags