Mostly Permanent Magnet Motor with minimal Input Power

[gyula]

Hi David,

thanks for the positive comment

I'm not sure if the magnet poles are all the same on each bar but it looks to me (I could be wrong) that Butch LaFonte's demo is the same effect as the Flyn Parallel Path. The main difference is he's using a PM instead of EM coils to redirect the PM flux.

If we use coils across the bars instead of the PM would this not create a path (bridge) for the PM flux to cross when the coil is not activated?

Interesting that you came to the same Flyn motionless idea I did 4 days ago.

Here is the post at OU: Solid State Orbo System

Thanks for sharing

Luc

Hi Folks,

My take on the LaFonte setup in the video is that it is very much a paralell path variant, except the moving magnet has permanent poles while the Flynn setup has electromagnets in the middle and their poles can be changed at will.

Yes all three permanent magnets are magnetized length-wise, the poles face to the bars sides. The upper and lower magnets should have their like poles on the same bar side and the moving magnet has the opposite orientation, this is what I think, otherwise the effect he shows cannot happen like that.

Yes, this setup could also be used for the Paul Noel setups, combining them is ok but how would you move the middle magnet, Luc? maybe with linear motor you started this thread...

On your question to replace the pm with a coil across the bars, if it were an air core coil then it would conduct flux when you switch current into it, but if it had a iron core too then it would conduct flux always in itself, except when you switch the current on so that the poles work against that flux. So it would be better to place the coils onto the bar as Flynn shows, not across.

rgds, Gyula

[gyula]

Gyula,

thanks, i put it back into the circuit, connected the power on the "load" side, bulb lit, i shorted with a wire from the gate to the source......bulb still lit....

= bad mosfet ?

Not yet... check the drain source polarity, drain to the bat+, source to bat- via the lamp? otherwise the body diode can conduct.

USe a ohmmeter in the diode test range to check drain source path: positive tip to drain negative tip to source: should be open circuit. when you reverse, the body diode should manifest. gate always tied to source here.

Gyula

[rave154]

gyula, ive just done the same test, using the mosfet "out" of the circuit, using a brand new IRF740.same thing.....bulb is lit all the time.

[gyula]

gyula, ive just done the same test, using the mosfet "out" of the circuit, using a brand new IRF740.same thing.....bulb is lit all the time.

With your voltmeter check the battery polarity wires that connect directly to the drain and source if they are interchanged?

what can you measure with the ohmmeter across the drain source in diode test range and also in higher Ohm ranges? (gate is tied to source)

do you have a third MOSFET, any type, n-channel? test it too.

Gyula

[rave154]

ok, its MAJOR BLUSH TIME HERE !!!

i had the positive and negative leads of the battery to the load....the wrong way around.......oh gawddddd ( wheres the embaressed icon????? )

at least let me explain.... the battery powering the pwm side...has red on the right..... black on the right...however.... the battery powering the load......has the red and black reversed.........yikes...

im off to crawl into a hole till you all forget about this embaressing series of posts

[gyula]

You sorted out in the end...

Just relax, such things may happen to anybody here and there.

Gyula

[rave154]

right, GOTO, first, sorry to take up space with my embaressing etc etc...

now, as i understand it, the flynn, while it produces 4X the force with the dual fluxes ( minus loses = about 3.7 times )... i dont know which avenue would be the best to go down... the kinetic way to try and utilise that 4X advantage but of course now your talking rotors & bearings etc with all the loses associated with that.......or.......to go down the solid state avenue.... you wouldnt get 4X the force... but its a lot easier to switch a flux via flynn / laFonte / etal... than it is to spin a large set of neos around in a circle.

i guess, i would have to vote....... solid state.

Thanks again gyula

[gyula]

David, using the solid state way will bring Lenz law into the picture we do not really need... at least this is what I think. When the flux from the load current enters the cores, normal transformer effects can manifest due to the common magnetic pathes...
Maybe using Paul Noel's suggestions on embedding several such PP setups into each other could minimize Lenz effect. Needs much and careful testing, I am not aware of any reports on that, unfortunately.

Gyula

[gyula]

....
The Paul Noel full PM suggestion and Backdraft are very interesting experiments also. The main problem is saturation as you say. I have no Gauss meter so it maybe difficult to replicate.
...

Luc,

Although I answered you here http://www.energeticforum.com/81371-post256.html , an addition to the saturation: the simplest remedy is to leave airgap in the magnetic path, either between two magnets in the stack or in the core. And the good thing is that if you design for the airgap in advance then you can change the gap size much easily with different thickness pieces of plactic or wood 'laminates', this way the flux strengths can easily be adjusted, trimmed to come back towards the linear part of the cores BH curve. And if you find you had started with too big airgap and your magnet stack is short, you can use transformer lamination pieces cut to size for filling into the gap.

rgds, Gyula

[rave154]

gyula & goto,

i just reviewed the p.noel stuff,

im attracted ( no pun intended ) to the use of a FPP device in conjunction with another FPPM device.

that idea is striking me as very efficient, its almost a little bit like the magnetic equivelant of the transistor darlington-pair, where a little transistor controls a bigger one which in turns switches or controls an ever bigger current.

[gyula]

Yes, I agree, a magnetic darlington transistor setup... and the good thing is inherently isolating input from output when two or more pairs are used.
I cannot state it will have no Lenz but surely it must have less than in a single stage normal parallel path solid state setup. I hope my take on this is not mistaken...
Gyula

[gotoluc]

Hi Luc,
great experiment as always.
It is odd that you can not get something useful out of the flyback spike. I made an experiment similar to this, but I did not have cores like you have, I just used iron bars and nevertheless I managed to charge up a capacitor with flyback spike using that setup. Can you do a quick test for me? First of all I would like to know how much energy it takes to fully redirect the flux, not in joules, just put a variable voltage source to the coils and then rise the voltage until one of the end pieces falls off and there is no flux there anymore. See what voltage and current is needed for that. In my iron bar setup it took about 5v and 4A. Then use the same voltage and do an experiment like this:

See if the capacitor will charge up. In my setup I used 10000uF 60v cap and it charged great, about 4-5 spikes was needed to get the cap to 12v. Maybe this is because material used, the iron is not the best thing to use in these setups as it has residual magnetism and lower permeability, so it takes more power to saturate the core.
If you have some free time for this, then please do that experiment for me
Thank you,
Jetijs

Okay Jetijs,

here are the results of the test you asked for.

First here are the coils (in series) specs. 36 Ohms DC resistance, 360 mH Inductance. The coil design is for use with 120vac 60Hz

I used a variable DC power supply and had to go up to 30vdc @800ma to get the magnet flux to a point where most metal object won't stick to it. However it would need to go even higher if the weight of the metal object was very light, like a single steel lamination of a transformer would require more power then the above. The heavier the weight the less power is need for the metal object to let go as gravity is doing work also. I guess my large transformer letting go was not the best example to use to see how much energy is needed to release as much of the work was done by gravity. It's a more useful test to show PM magnetic power.

I did the next test with 38vdc battery bank and a 6000uf cap attached for flyback and used a wall plug transformer core and did a quick pulse by hand to release the small transformer and only got about 1 volt on the flyback cap for one pulse. When the transformer gets magnetically connected it induces a voltage of about 0.5vdc on the flyback cap also.

Hope this helps.

Luc

[Jetijs]

Thanks Luc
Those are great news. The power requirement for complete redirection of flux is higher than I expected, I thought that my crude setup should be way less efficient than yours. This is interesting. But I like the fact that the flyback can be captured that is encouraging Here is a picture of my setup so that you can compare:


Thanks for doing that experiment for me
Jetijs

[gyula]

Hi Jetijs,

I think the reason Luc experienced a relatively low amplitude flyback pulse is that transformer laminations do not like DC current bias, they can easily saturate and REMAIN in saturation for a longer time you would expect it to recover hence the collapsing field sees a much lower inductance value to induce into. (Maybe you both saw Nali2001's videos on core tests like that.)

If you do not mind doing another test on you setup in your picture: insert a wood or plastic piece between the magnets and the inner side of the bar to reduce magnet strength and this way the switching input power should also be less, hence the core saturation will also be less, this way the flyback pulse should also increase, dispite the lower input current... I think. Use at least 3-4 mm air gap at the magnets for a start.

rgds, Gyula

[Jetijs]

Thanks Gyula,
I saw the core test videos and I will do the test you proposed tomorrow.
Thanks,
Jetijs