Dismantling and repairing the motor :o

[bob2.0]

Today I studied the rim/magnets more closely.

I think everything that is not quite right is this one magnet.

1ca37bdc-754b-4317-8959-667ea45e8945.jpeg (104.22 KiB) 494 mal betrachtet

On the far side (away from the cable side of the motor, it sits flush with the other magnets.

On the other it sits around 0.3mm proud.

The triangle shape is therefore understood. At the tip of the triangle it is precisely zero clearance with the stator.

At the base of the triangle, the middle is closest to the centre, because the magnets are flat and are each a short chord to the circle. So the middle and as far as the tilt on the magnet interfere with the stator. So, it is a triangle.

I realised that there is no reason for any wear on the stator. I measured more accurately and all I could find was wear on the one side of it. It is about 0.1mm of wear, on the cable side edge compared to the other edge.

I cannot say if something has made this magnet tilt, or if it was like this since it was made. I think it is since it was made. I don't know how they make these, but maybe it was the last magnet in and didn't quite fit. Or maybe the first one and it popped out.

I don't know.

All I know is that the only solution I can see is to sand this magnet back by at least 0.15mm. I know this because the wear triangle shows it is about half way between the tilted up edge and the correct placement.

I cannot see any corrosion on the rim or the magnet to show bubbling underneath. I think it was just too much glue, didn't quite fit. Or maybe it was one-off magnet not quite right.

Trust my luck. I always seem to end up with this sort of thing. Nothing at all I could have done to stop it, and yet there is no-one else to take responsibility.

I assume the plating is already worn off and I am looking at sintered magnet? I do not know the best solution to remove this thickness.

Maybe I do not bother and I let the parts grind themselves into place? I can open it again next year and take a look?

What would you do?

[MEroller]

Grind away another 0.1 to 0.2 mm on the triangle side, WHILE having someone else VERY closely have a small-tipped vaccum cleaner nozzle going at full tilt (best is an industrial one), in order to whisk away as many magnetic particles as possible, because you won't be able to get them off magnet otherwise. Maybe covering the not to be ground off magnet area with masking tape, which you should be able to peel off in one piece after vacuum as much of the magnetic powder off as possible. And before peeling it off, put on another covering layer of masking tape to as good as possible take prisoner the magnetic powder.

And after you are done apply a thin coat of clear coat or the oil you prefer to the enlarged triangle, for some corrosion inhibition.

[bob2.0]

Here is the rotten detail.

The magnet is completely demagnetised in the area of the triangle. It is clear it has got very hot with friction and demagnetised.

The other end is OK and so are the magnets either side.

So, it is only half a magnet anyway, now.

I used sand paper and it produces only a fine dust that can be wiped.

But, is there any point to go further?

I cannot tell if there is corrosion beneath this piece, which has pushed it out?

In any case, I think, like a dentist, I might as well take it out? It will get worse and corrode more, stick out more.

Could this be a reason my controller gave a motor error? With only half a magnet, the currents are wrong?

What if I take it out?

If it snaps in the middle, this solves the problem for today. Then I have removed the extra thickness!

I am looking now for a replacement magnet. I measured it as 45mm x 16mm x 2mm thick.

The closest I can find is 40mm x 15mm x 2mm.

I am sad. I seem to always suffer these sorts of complex problems.

(I have a small violin, I can play sad tunes on it.)

[jeff-jordan]

That's pretty weird!

I'm wondering what's the red stuff between magnet and rim?
It's only visible at THIS magnet.

Might it be that it's a manufacturing fault? Even some doctors leave tissues or scissors inside a patient....

I think going for a new 40x15x2mm neodym magnet would be a solution... not perfect, but it won't get worse.

...
Could this be a reason my controller gave a motor error? With only half a magnet, the currents are wrong?
...

It could cause problems when the damaged magnet tries to trigger the hall sensors... so the controller might get a hall sensor error, for a short period of time.

[bob2.0]

It might well be rust. I gave it a scrape with a clip-puller tool (the pointy one).

If the back of the magnet gets scratched and then it ends up at the bottom and condensate forms, well, you can imagine.

I don't think 2mm shorter either end will make any difference, and the 15mm is sure to fit (I measured 16 but couldn't get the verniers in squarely, so better too small than too big.

The magnets I can get. What worries me is the glue to use. The cheapest magnet-specific glue that Henkel do (trusted brand only here, if a magnet comes loose and the wheel locks on a bend .... not nice) looks like AA326 (£30) plus the activator B7649 (£20).

I mean, the glue alone costs 10% of a new motor. Plus magnets, plus sealant, plus etc etc etc....

[bob2.0]

I ordered the 40mm x 15mm x 2mm magnets because it says they can take a month to get here. Maybe just get my order in before another Chinese holiday?



Will I ever ride my bike again? At least the battery was at a middle charge, so should be OK for 3 months storage.

[bob2.0]

It could cause problems when the damaged magnet tries to trigger the hall sensors... so the controller might get a hall sensor error, for a short period of time.

I had this thought too.

When the motor is going faster and field weakening, does it expect a smaller Hall effect signal? Because it worked OK over 60km/h.

Or maybe there is a very small field in the magnet and when it goes very fast it is just enough to trigger the sensor.

I think these things make sense. Maybe that last trip when I was going fast, the magnet finally demagnetised after a long time, hot and with field weakening that would demangetise it.

If I fit these shorter magnets later on, I will have to decide if it needs to be over to one side to pass the active Hall sensors.

So many complications! AArrrgh!

[bob2.0]

I have been trying to decide how this magnet might be removed.

I have looked at getting an oscillating blade tool, to cut behind it, but it will only go in so far before the magnet distance to the rim is too small for it.

I have thought of using a fine drill to weaken it, then hammer down.

Or a very sharp blade.

All of these will be thrown around by the magnetic field of the other magnets, and even if I do a great job of it, at the far end there is probably only 0.05mm of glue bonding. So, in any case, the magnet will disintegrate.

I am thinking I might as well just hammer it off with the coarse tools I have.

I will tape it over with the strongest adhesive tape, and hope that this is enough to capture the main broken parts.

I just don't really know. There are no good outcomes.

One other thought is to drill two holes into the magnet down to the glue, and then thread screws in, and lift it off by tightening the screws. Again, I think it will just break up.

Maybe more holes?

I also have to deal with the tool trying to snatch away from the magnets either side.

Why can't they make SR hub motors!!!!!!

[bob2.0]

Today I looked again in more detail at the job to replace the magnets. I had started sanding down the bad magnet before realising it was lost magnetism. So, I abandoned doing that.

I realised that the magnet one along was also sticking out. But at the other end, so I did not see it.

This also has the triangle shape. It has also lost its magnetism at the end, too.

So, that is two to replace.

It will be like going to the dentist.

I used my favourite small screwdriver, it is very hard metal. I was sad to abuse it. But it did the job very well.

I just tapped it in behind the magnet. Of course, it broke into pieces, but it came of very easily.

See how the plating was left behind, so it rusted underneath the plating itself.

[bob2.0]

I have ordered the 40mm x 15mm x 2mm from China, but it seems they are not even yet posted.

I have found 15mm x 15mm x 2mm square magnets here in the UK. I think perhaps I should buy those and fit 3 in each slot. It will be either easier, or more difficult. One or the other! At least I am sure it is then the same area as the other magnets.

I am not sure if it is better to wait for the single pieces (though shorter).