Motor Conversion Magnet Mounting Question

[Spdlmt150]

I recently reworked the armature on the Baldour motor conversion (repositioned the mags a hair to reduce cogging a little more) and was thinking of trying something a bit larger. I looked through Zubb's posts once again, and I noticed that he turned the armature down, made a sleeve to position the mags, and then potted the whole thing.

When I did the Baldour, I milled flats to position the mags.

My thought, or question as it is, is this.... When doing a motor conversion is there any benefit to milling flats for the magnets? Without milling, only the center of the magnet is actually in contact with the armature. Is that slight gap out toward the edges causing any noticeable loss of field strength?

I know I have my thoughts on this. I also know there are many here that have more than speculation to provide on this subject (Flux?) If anyone could model it to see if there really is any gain, it would be an interesting thing to see.

If anyone knows if the time invested in milling flats for the mags is beneficial, or a complete waste of time, cares to speak up.... Here's the place to do so. After seeing the Baldour perform, I am now working on a 1.5HP 3 phase, and need to decide if I will spend the time to put it in in a mill to put flats on it, or just make a sleeve to position the mags over a round armature.

[Bobbyb]

Hi I did a motor conversion my self an I did mill flat's. I'm sure it makes a difference bun I' have no experience how much. What I do know that it took twice or maybe even three time's as long to turn the rotor down that it took to mill the flats.

So if u have the right equipment like I did, I'd do it anyway. What's 4 ours more work on het years that you windmill will probably run?

[Flux]

What magnets are you using?

The effect depends on the air gap that you introduce. Perfect magnets are curved to match the rotor core and the stator core. If you use lots of little magnets placed in rows as Zubbly usually did then the air gap you are introducing is small and not likely to have much effect. The cage will hold the magnets adequately.

If you use rectangular blocks as magnets then you are forced to have a significant air gap next to the stator which you can't avoid, in which case it is as well to mill flats and not have the one that you can do something about. Also for the block magnets I would not be happy mounting them on a curved surface unless you have a banding over them or some other fairly foolproof means of keeping them in position.

All the air gaps require you to use a longer magnet length, so how serious it is depends on the magnet shape. I saw a comment somewhere that 1/4" thick curved magnets performed at least as well as the 2 x 1 x 1/2" blocks so it seems that the gaps are significant. If your magnets are greater than 1/2" thick then from the magnetic point of view it may not be worth milling flats but from the mechanical point of view it would seem a wise move.

All this of course assumes that the flat is being machined on an oversized rotor core. If by milling a flat you actually increase the air gap to the stator then it would be a backward move.

Not sure if Peter has run any FEMM plots on this.

Flux

[Spdlmt150]

The rotor will be turned down to keep a minimal gap with the mags used. Similar to how I did the baldour. As for the magnets, doing things Zubb style, I think smaller magnets would be best to avoid a large gap on the stator side of the assembly. The 1/2"  X 1/2" X 1/4" mags used on the baldour seemed to work well. I may use them again on the next project. I looked into curved mags for it, and would be paying a small fortune to have them custom made.

My only other thought was to use large bar magnets with shoes. Fewer magnets to mount. Shoes would eliminate almost all of the gap. I don't know if the field would be any better than with many smaller mags. I would think there would be some loss of field strength in the shoes. I've wanted to make another shaft for the baldour trying this to have something in hand to compare. Someday I'll take the time to do it.

I will most likely end up milling the next rotor. The one big benefit I see to it is that magnet placement/spacing can be held more accurately.

[Spdlmt150]

My equipment is a full 3d modelling software package, and access to several cnc machines. Time is the big thing for me.

[Flux]

I think I would avoid the rectangular block magnets with pole shoes. Unless the shoes are well laminated you leave yourself open to significant eddy losses within the shoes from slot ripple unless the air gap is wide, in which case you might as well keep the gap as close as you can without the shoes. Producing thin laminated shoes is not easy and fixing them is another issue so I would avoid that option.

I agree that milling flats is far better mechanically and would seem justified even if magnetically it made little difference.

Flux

[SparWeb]

I made a conversion much like Zubbly's, with flat faces and block magnets on each.  The block magnets that I chose were quite large (bigger than necessary, it turns out) and I made a brand new rotor because the old one would have just fallen apart if I turned it down that much.

Here's how I did it:

http://www.sparweb.ca/Conv/CONVERSION_V1.pdf

It has been in the air since summer, putting out plenty of power for the barn's lights.

[Spdlmt150]

I had an idea today which I am investigating. Shrink an aluminum sleeve onto the turned down rotor. Space things to put the center of the face of 1/2" round by 1/2" thick magnets right at the inside diameter of the stator teeth. After the magnets are epoxied in place, the whole assembly gets ground down for stator clearance. End result is multiple magnets creating each pole, flats milled on the rotor for maximum field, and still maintaining a minimum airgap.

The big issue with this is trying to grind a combination of soft & hard materials simultaneously. I should have an idea tomorrow if it would be possible. I know aluminum can be a real mess with a grinding wheel. I also know of a continuous dressing system that is designed to knock the alum off the wheel while in use.

This will be tested on a good size 3 phase motor - at least 2HP. I'm hoping to find something around 5HP.

Estate issues should be dealt with tonight, and hopefully I'll be back to tinkering soon.

[Flux]

Not sure I follow this. If you are proposing to grind the neo magnets then I suggest you don't. With plenty of water they may grind but even if that is successful then they will start to disintegrate with no protection and I doubt that anything you try will stop it.

At the very worst grinding a combination of aluminium and neo without sufficient water would not be a good combination. I hope I got the wrong impression of what you are trying to do.

With multiple little magnets the air gap is not going to be a big issue compared with   the rectangular blocks.

Flux

[Spdlmt150]

After talking to a few people on this one.... Goal is to remake the rotor to the original diameter with an aluminum sleeve to locate the mags. Pockets milled, with a small flat on the steel rotor. The sleeve will be turned down to finished size before installing the magnets. Once the magnets are epoxied in (sticking out above the sleeve slightly), the whole assembly will go on a cnc grinder, which will take the magnets down .001" per pass until they match the diameter of the sleeve. After grinding (which I have been assured can be done with no damage to the magnets) the rotor will be sealed with resin. The finished assembly will be multiple magnets per pole, with the proper radius to match the stator (as the original assembly did) to maintain the same airgap that the motor had when functioning as a motor.

[dinges]

I wouldn't bother grinding the magnets down but if you do, report back on how it went so the rest of us can learn too.

You will likely gain much more flux by building a complete new, all-steel, rotor - as opposed to re-using the old alu/steel rotor.

I've always built new steel rotors myself, I've never re-used the old (turned down) rotor. I will likely never go through the trouble of grinding the magnets/rotor down though.

If you download FEMM and play around with it (it's easy software, shouldn't take you more than half an hour to master it) you could find out for yourself. Make a .dxf drawing in CAD (which is more work than the actual simulation) and export to FEMM. Run simulation. Instant answers.

Regards,

Peter.

[ghurd]

I expect it would be easier to mill a "D" rotor spot for each magnet.

No reason to sink them if they will have an AL cage.

If it's going to have a coating and a gap, might as well leave the plating.

Anyway, with the resources at your disposal, no way it could turn out bad.

G-

[Flux]

I agree with Peter on this. If you are proposing to keep the original cage rotor then don't. The composite steel/ aluminium combination will lose you more flux than the little air gaps you are worrying about. Those cage rotors have so much aluminium in the bars that the remaining steel will be saturated.

Start with a new core. Fine to use the cage and spot mill the core to let the magnets full contact.

I still don't advise grinding the magnets, you will save little air gap and you will probably never cover the bare neo properly with any potting without introducing more air gap than you are trying avoid.

Having done all this I would suggest you use long enough magnets to achieve the flux you need with an average air gap to the stator of at least 1mm. There is a very  good reason why induction motors have incredibly small air gaps but this is not the place to discuss it. There is good reason for synchronous generators to not have such tiny air gaps, you will gain little in terms of flux, you will make yourself more open to mechanical rubbing which will kill the magnets. You have more problems with heat from the stator affecting the already dodgey neo which is running close to its temperature limits. You will let yourself open to excess slot ripple loss in the neo magnets and particularly the aluminium cage and you will exaggerate any residual cogging.

I can see the point of avoiding large unnecessary air gaps from sloppy construction but there comes a time when absolute perfection becomes counter productive.

I am not even convinced that the aluminium cage has any benefit except for holding the magnets safely and if it was kept below the height of the magnets it would still do this perfectly well without having to worry about the possibility of eddies being induced in it.

Flux

[Spdlmt150]

As always, thanks for your insight Flux. Grinding is out (not enough gained)

Solid core is not a problem. The thickness of the core should be 1.25 X the thickness of the magnets... I think? My thought on that is to make a steel sleeve, over an aluminum core, pressed onto the shaft. Just an attempt at cutting down the weight of the assembly. A larger motor with a solid steel core would be insanely heavy.

Now... this may lead to another thread if it goes somewhere... I was looking at inverters today, and realize the benefits of jumping voltage up to 24 or 48V on the batteries. The one big drawback to going to 48V is that the inverters are big money. My thought this afternoon is this - Why have I not seen anything using a 120VDC battery bank with a circuit to convert the 120VDC to 120VAC? Is there such a circuit? Is there any reason this isn't done? My thinking is that it would be quite nice to get away from the high-amperage current draw on the DC side of the power system. More or less run the same amperage on both sides of the system. Am I way off in thinking this would be much better than running low voltage and high amperage on the DC side of the power system?

[dinges]

Flux, I've noticed you saying this several times before. This time I can't resist asking - what do you exactly mean:

"use long enough magnets to achieve the flux you need with an average air gap to the stator of at least 1mm."

Do you mean 'long' as in what I would call 'thick', i.e. magnets being 'long' in a radial direction of the rotor ? Your remark would make sense to me if interpreted like that.

Or do you mean long as I would normally call it, i.e. long in longitudinal direction ? ('long' in the direction of the shaft)

Also, I've never imagined that the aluminium sleeve had any other function than providing a non-magnetic case for the magnets, so as not to short flux yet provide a way of safely mounting the magnets and positioning them. I find another (maybe the most important) benefit is that an aluminium sleeve (that's higher than the top of the magnets) protects the delicate nickel coating of the magnets during insertion and removal of the rotor from the stator. Though if that's the only purpose then the method the original poster used in his previous generator would be even better: casting the rotor in epoxy. There will be no eddy currents induced in the epoxy...

To the thread starter: I think the extra weight of using an all-solid steel rotor is hardly an issue. Even in my latest project (a 10 hp, weighing 65 kg originally; the rotor itself weighing ~15 kg, from memory) the few extra kg of an all-steel rotor would be hardly noticeable. It would be even less of an issue in smaller conversions, whereas the added complexity (machining accurate workpieces for the interference fit) requires quite a bit of extra effort.

In an earlier 3 hp conversion I made the rotor/shaft combination out of a solid piece of steel on request of my machinist friend. It was much simpler/quicker for him to turn down a solid piece of steel and grind the bearing locations than to try to re-use the old shaft. Pictures of the work can be found here:

http://www.anotherpower.com/gallery/3HP-induction-conversion

It's the plan to use the same method (but without aluminium sleeve, and using rectangular magnets) on the 10 hp conversion. The rotor/shaft will again be made out of one piece. Costs a bit more in material but saves much more on machining.

Regards,

Peter.

[Flux]

Peter

Sorry about the confusion. By magnet length I mean the length in the direction which causes the mmf. It does happen to be the physical thickness with neos whereas with the older magnet steels it would have been the physical length.

Very difficult not to be ambiguous sometimes when some people are thinking of physical properties and I am thinking of the magnetic ones. I just assumed the physical length was fixed by the stator dimensions.

Regarding the aluminium cage, I have to agree with you that using it to protect the magnet plating during assembly would be a higher priority to me than the losses it might incur from induced ripple. I wouldn't use a cored machine for a low wind area and if I used one it would be for robustness rather than absolute efficiency.

For someone so concerned with absolute efficiency as to consider grinding the magnets then having a cage so close to the stator would be doubtful but there again everyone sees their own issue in isolation from the whole picture.

If you want to cover all options then the cage could be kept above magnet diameter and milled back in the region of the magnets, leaving a guide pad mid way between the magnets to steer it during assembly.  I don't think there will be significant surface loss anyway but there is a real issue with solid steel pole shoes and small gaps .

It is conceivably possible that the cage could prove beneficial just as damper bars help a single phase alternator, but with 3 phase the main field will be constant, the only significant flux ripple will come from the stator slots. The danger region for eddy loss would seem to be right close to the magnet tips where fringing to the stator teeth would be greatest.

Flux

[Spdlmt150]

Grinding is out. Not enough gain for the time. I have to agree with the idea of leaving the cage above the surface of the magnets to protect them. As for absolute efficiency.... Why not get the most bang for the buck? If I can put a little more work in to the project to boost the output then I will. I would also like to avoid spending time on things that won't benefit the project.

I have found a claim of 91% efficiency on a ferroresonant transformer. 60A capacity. Can be dead shorted without damage. Could convert 120vdc to 120vac directly without a lot of components. Cheaper than any 48V inverter options. I'm wondering what the problem with using this would be? Why is it not common? Anyone have any experience with a setup with high voltage dc through a ferroresonant transformer?