Hilden-Brand Magnet Motor

[kadora]

Hi all
Just an idea.
What if Jack valve was built in a bit
different way. I think , instead an iron jacket
to use thick coil wound with the iron wire only
and then to switch this iron winding .
Would it be more efficient?

even a stupid idea can shift you to better one.

[MeggerMan]

Hi, Gyuda, Kadora, JackH,

Iron wire as a coil is a very poor conductor and will be of no use here.
The ideal material would be room temperature super-conductor wire which does not exist (at the moment).

Thanks for pointing out the flaw in my idea, I will make sure all future simulations are done with/without a solid core of the same material.

OK, I have just completed a lot of simulations on ferrite and silicon steel and the results are:

Ferrite will only work with a very very low power magnet, even less than cermic 5.
So this is a non-starter.
Ferrite will actually produce less change of flux with a ceramic 5 than with a solid ferrite core at 20mA coil current.

The only way to go is with silicon steel and the most powerful NdFeb Magnet you can get.
With the torroid setup the difference at 20mA is 150% of the setup with a solid core of silicon steel at the same current.
M19 Steel:

Cur. I(mA)	Core material	Flux min(mT)	Flux max(mT)
0	NB37	43	150
20	NB37	152	586
20	M19	67	294

As you can see the max. difference is:
(On state - Off state) - On state (solid steel core) = 586 - 150 - 294 = 142 mT

This means that I need to scrap the idea of using ferrite, and even using silicon steel, the design will need to be very tight to make use of all the available additional flux.

Regards

Rob

[gyulasun]

Hi Rob,

Thanks for sharing yours simulations but I fail to notice any answer to my question before and may I repeat it with your kind permission:
"Have you checked that with your latest simulation how much the flux is in the outer leg when you remove the permanent magnet and let the the 20mA current flow into the electromagnet? 
Would be curious to know.  Thanks."

So I mean to leave the volume of the permanent magnet empty with air and then see the flux in the outer leg solely by the electromagnet . Because if you take out the magnet, its lack will change very little the resultant permeability in that leg area where it was placed before.

I agree with voting for materials that give the biggest flux change possible but would not you have serious eddy current losses by choosing silicon steel as the core? Or you choose a very low switching frequency to minimize it?  Jack wrote about iron particles mixed with insulating material as the best which is just the CURE against eddy losses (just like to use ferrit in this respect). Oh I remember now you have plenty of laminations that is ok.

Thanks
Gyula

[MT]

This is maybe slightly of the current discussion but I though it might be interesting to post. On Sunday did simulation on the valve to see how it works in FEMM. So I made a 3 leg circuit with magnet in the middle leg. On the picture first circuit is with shorted magnet. In the left leg in position X=-107.9,Y=42 we get |B|=0.04676.
Second circuit is with coil switched on with 0.4A current. In this case there are no flux lines around magnet but left leg on the same spot has flux about 0.3741T!
The last third circuit is without magnet with coil switched on to the same current and on the same place in left leg |B|=0.1959T.

So in circuit without magnet by switching coil on and off flux difference = 0.1959T
In the circuit with magnet we get 0.3741-0.04676=0.3273T flux difference or 67% more flux comparing to circuit without magnet.
Although I expected 2x more flux it is still nice isn't it? 

with kind regards,
MT

PS: Thanks kingrs for your FEMM files I used them as starting point.

[MeggerMan]

Hi All,

Gyula, I have included your air core results in the attached spreadsheet.

Nice to see that I am not the only one trying the Femm simulations out (MT well done!).
MT, I think your units are wrong, I think you mean T not mT, out by a factor of 1000.
Check the polarity of your current, try -ve current to see what difference that makes.

I have been playing with lua script tonight (attached) to speed up the results and they are as per attached spreadsheet:





So to sum it up I have managed to get a 186mT excess flux density over a steel core, which is 127% excess.
The circuit properties for the on state:

Total current = 0.0095 Amps
Voltage Drop = 0.397284 Volts
Flux Linkage = 7.51958 Webers
Flux/Current = 791.534 Henries
Voltage/Current = 41.8193 Ohms
Power = 0.00377419 Watts

Thats 3.8mW of power, quite low I know for a device that is 10" x 15" x 2".

I have run hundreds of simulations to arrive at the shape shown above for the best efficiency.
Try and beat it if you can, everything you need is attached as a file.

Regards

Rob