Hi gyulasun - you're right, neos are according to Wikipedia >1 Tesla. I've been using the magnet calculator here: http://www.kjmagnetics.com/calculator.asp - which says an N52 has a field of 6619 Gauss - i.e. about .66T. I'm not sure how to reconcile that difference TBH so I was working on the assumption that the folks who were making the magnets were probably right - and to assume the lower field.

Electric Steel - http://en.wikipedia.org/wiki/Electrical_steel

Assuming the neos were 1.5T - it's still no more than electric steel - and the cost of the big magnets you'd need would be high. Agreed that eddy currents are a problem in steel - and laminations are always desirable - but maybe not necessary if the design is OU enough.

I have been looking at the following arrangements for a 'Magnet Piston' type engine - where the coil is the stator - i.e. fixed, and the magnets / core moves. Input is pulsed DC.

 Case 1) A single plain steel armature - as in a typical solenoid.

 Case 2) 2 opposing magnets on a non-magnetic shaft - both situated within the core of the coil (at the extremes of stroke the opposing magnet would just appear out of the end of the 'cylinder', while the attacting magnet is pulled to the center, of course)
In this case - there's no steel at all - just the magnets. It's a good arrangement.

 Case 3) As 2, but with a fixed central steel core, perhaps 1/5th the total length of the coil - to provide a fixed central magnet for both the moving magnets to act upon. (The stroke length is reduced...). This is an excellent arrangement - as the magnets act on the B-field of the central core, and the H-field of the coil too. My design uses no permanent magnets, so is a 2-stroke.

I've calculated power at 1000RPM for each of the above. Obviously my calcs could be totally wrong, but I'd be happy to go thru them with you guys... One problem is that I don't believe how little power it takes to fully saturate steel:
http://en.wikipedia.org/wiki/Saturation_(magnetic)
...says 100 Amp Turns per inch is more than enough (and diameter doesnt seem to matter!). So for my 8" coil below - I should only need 800 Amp Turns, not 8,000. In which case the coil can be made much thinner... I've been trying to err on the side of caution.

Given 50w input into coil: Length 200mm, InsideDiam 100mm, OutsideDiam 300mm, 3mm copper, 2178 Turns, 3.23 Ohms (3.9A, 12.7V) -> 8500 Amp Turns, 42830 AmpTurns/meter (Total mass of coil is 85Kg!)

 Case 1: Solenoid) Output power = 1,300 Watts
 Case 2: 2 Magnets = not calculated, as i've recoded and dropped this calc... Tended to be about 2-3 x case 1 with 1.5T magnets
 Case 3: 10,800 Watts (14.6Hp)

I've been looking at using aluminium for the coil - to cut weight. The 'attacting' magnet / armature would be a steel cylinder. The 'repelling' armature would be a smaller steel cylinder wrapped with a coil. Connected via a non-magnetic shaft through a hole in the central core, and to an external crank.

So, if I could get case 3 running at 5,000 rpm it should give 70Hp approx, which is enough to run a motorbike - off a standard 12v battery...

I've started a new thread about my 'Case 3' magnet piston engine design. It's here:
http://www.overunity.com/13673/tims-magnet-piston-engine-design/

Hi gyulasun - you're right, neos are according to Wikipedia >1 Tesla. I've been using the magnet calculator here: http://www.kjmagnetics.com/calculator.asp - which says an N52 has a field of 6619 Gauss - i.e. about .66T. I'm not sure how to reconcile that difference TBH so I was working on the assumption that the folks who were making the magnets were probably right - and to assume the lower field.

Hi Tim,

Yes, I accept that info on Wikipedia may be questionable in some cases, albeit what I quoted was saturation magnetization which goes together with residual flux density, Br  while your example sounds like surface field strength data. If yes then surface field is magnet shape and configuration dependent. For instance in case of magnet type DX08 N52, in free space, the field in the surface center is 5237 G and  at any one of the 4 outer 90° corner edges it is 9417 G.  And I agree, even with 1.5T values it would be in pair with electric steel.  (If inserting this magnet into a closed magnetic circuit, the field can be higher inside the circuit versus that of the free space surface data but you surely know this)

Electric Steel - http://en.wikipedia.org/wiki/Electrical_steel

I think Luc has used laminated core scavenged from microwave owen transformers and the Wiki description sounds like transformer cores are made from electrical steel, so no problem here, just the term electric steel what may have caused some confusion, not mentioning extensively for transformer cores.

I've calculated power at 1000RPM for each of the above. Obviously my calcs could be totally wrong, but I'd be happy to go thru them with you guys... One problem is that I don't believe how little power it takes to fully saturate steel:
http://en.wikipedia.org/wiki/Saturation_(magnetic)
...says 100 Amp Turns per inch is more than enough (and diameter doesnt seem to matter!). So for my 8" coil below - I should only need 800 Amp Turns, not 8,000. In which case the coil can be made much thinner... I've been trying to err on the side of caution.

Your Case 3 setup looks interesting indeed and later I will try to go through the calculations in your other thread.  Regarding the curves in the Figure http://en.wikipedia.org/wiki/File:Magnetization_curves.svg which suggest a low power is needed for saturating steel material: I think such data always refers to toroidal shape cores (i.e. in closed magnetic circuits) made from the different materials to compare or characterize them. In case of an open core like a piston setup I think higher power is needed for bringing the steel into saturation.

rgds, Gyula

Thank you Tim, I'll have a look

Luc

Hi telecom! Nice to see I am not alone to advise for outpower measurement by weight lifting: kg/s against the gravity; it is for me most undisputable way to demonstrate any O.U. What do you think?
Cheer.

Not sure to very inderstand the principle and the calculations for this kind of device; just be about giving true ideas onhow to demonstrate O.U. so that nobody could contest it.
Good luck for the rest of the quest

Hi Luc,
was very excited looking at your videos - great work!
Just want to point out that in order to generate 1 w of energy you should be able to lift weight of 10 kg by 1 centimeter at 1 second,
or 1 kg by 1 cm 10 times per second, or 250 grams at 1 cm 40 times per second!
 But generally speaking, in order to achieve this you need a) - much bigger magnets, b) - much smaller travel of the coil.
According to my calculations the magnet ( the source of the energy) should have the pull of about 400 kg, and the travel about
.5 cm. With your setup this doesn't appear to be practical since magnets will be fighting each other (IMHO).

Hi telcom,

thanks for your interest and test information.

I have a new (stronger magnet) model that I have tested. I noticed that as I drop the current the pull grams per watts is not linear.

Example:

My new model pulls 520g for 1.2 watts input.  However, if I drop the current it can pull 200g with .3 watts input.

Luc

Hi Luc,
this is a big step in the right direction.
If it takes 1 second, than the mechanical equivalent of .3 w X sec (.3 joyles) equals
3 kG x cm, and the efficiency is still below 10 %.
Perhaps you may add few more magnets, or increase number of strokes per second or both?