Potting windings on a hub motor

Has anybody tried to pot the windings of a 3kw DD hub motor with thermally conductive epoxy/silicone?
Was there a sigificant improvement in combination with Statorade+Hubsinks compared to Statorade+Hubsinks only?

Obviously on a hub motor it is not possible to build a thermally conductive bridge to the case because the case is rotating so I guess the improvement of potted windings on a hub motor would only be minimal.

Justin did it in his heat thread, one of the best threads on here. There was a small improvement but i don’t think it was worth the effort and cost.

Long but worth the reading as long as justin’s writing. After a while it derails into non-facts again but the first 20 pages or so are really good.
https://endless-sphere.com/forums/viewtopic.php?f=2&t=48753

hias9 said:

Has anybody tried to pot the windings of a 3kw DD hub motor with thermally conductive epoxy/silicone?
Was there a sigificant improvement in combination with Statorade+Hubsinks compared to Statorade+Hubsinks only?

Obviously on a hub motor it is not possible to build a thermally conductive bridge to the case because the case is rotating so I guess the improvement of potted windings on a hub motor would only be minimal.

Click to expand...

Potting can help adding thermal mass, so the motor doesn't overheat as quickly. Though it will take more time to cool down, and add some extra weight. What you can do is some light potting (epoxy the coils and oil proof cable exit) to allow for oil bath cooling which will be much more effective, but also quite difficult to implement. Beware that mineral oil will eat through the silicon; only synthetic stuff will work.

I am now planing to pot at least the end windings because I usually run the motor up to 135 degrees Celsius and on only a few wires the varnish is already black. So heating at the windings might not be equal everythere.
Which potting epoxy should I use and are there any videos/pictures/instructions about potting a stator (or at least the end windings)?

Potting the windings is pretty useful for an underwater motor, but not for ebikes.

Even with thermally-conductive potting material, it would slow the heat-transfer to the ferro fluid.

Here are two pictures of a 3kw DD hub motor which I run up to 135 degrees Celsius.





Would thermal epoxy not help to spread the heat at the windings more equal and conduct heat better than the paper and wooden sticks? (I have now removed the wooden sticks because some came loose at high temperatures.)

The basic problem won’t be solved by potting the end turns. Lowering the current or duty cycle would :wink:

When windings start to blacken it’s time to rewind, anything else is a gamble that a short takes out the controller eventually.

The copper winding wire is rated 180 degrees Celsius continous according to the Chinese manufacturer and the temperature sensor on the windings (which I have calibrated with the controller, so there should not be more than 5 degres error) never saw more than 140 degrees Celsius.
Did you already rewind auch a stator? I guess this is a lot of work. Motor still works fine and ASI controller should have a protection for that.

It’s a lot of work to rewind, unfortunately. No controller has protection for a burnt/short circuiting winding as far as i know. It’s not exactly a planned/foreseen operating point..

You could get some winding varnish to save the windings before short circuiting, if they’re blackened at 140deg i’d lower the temp limit in the controller to avoid breaking the enamel down further.

Lowering the temp would be a solution of course, but with this motor I was also testing how far I could go and I was worried more about the magnets which are rated 120 Celsius. Kv did not increase at all. Hall sensor are rated 150 degrees (no problems) and winding wire has a 180 degrees rating (which should be continous).
The motor is still working fine, but I am disappointed of the quality of the Chinese winding wire.
Does anybody in Europe offer a reasonable price for rewinding with quality wire?

Brocotts.co.uk
Wires.co.uk

But the damned brexit almost stops ordering from the uk, i waited 2 months for my last order.

:wink:

hias9 said:

Lowering the temp would be a solution of course, but with this motor I was also testing how far I could go and I was worried more about the magnets which are rated 120 Celsius. Kv did not increase at all. Hall sensor are rated 150 degrees (no problems) and winding wire has a 180 degrees rating (which should be continous).
The motor is still working fine, but I am disappointed of the quality of the Chinese winding wire.
Does anybody in Europe offer a reasonable price for rewinding with quality wire?

Click to expand...

Increased temperature will lower the lifetime on all components. 180 is not continuous, it’s a max, hall sensors 150 is also a max so run at 140 for any longer time period will make them age and fail prematurely.

Look at any datasheet with mean time between failure ”MTBF” vs temperature and it’s like this.

Getting some wire is not the problem, I meant if somebody offers a rewinding service for a reasonable price in Europe.
As far as I know, the temperature rating of the winding wire (like W180 for example) is continous, not peak.
Maybe not for the Chinese wire.

If the hall sensors fail, no problem. I only need them for a smooth start, after that the contoller switches to sensorless FOC.
If they fail, I will look for sensors that work up to a higher temperature.
Regarding the magnets, I already have 150 degrees rated magnets here, but have seen no change in Kv at all for the stock 120 degrees magnets so far.

I think motors shouldn’t be run that hot, it’s wasting energy that could be better used. Might work, might not.

If you find a spec sheet saying the wire is 180deg continuous then please link it..

180deg rated winding wire is nothing special, there are lots of them. Just google it. I guess they used a 130 degrees rated winding wire which is the cheapest and told me incorrect information.
The winding wires which use PAI as insulation material have an even higher rated temperature (210-240 Celsius), but its surface is harder, so more difficult to wind.

Obviously efficiency suffers a lot when the windings are hot, but if you live in a hilly area you could for example finish the climb without power being reduced and cooling fins will shed the heat after reaching the top.

Just saying.. see that “max winding temperature” per class?

Yes, but according to spec sheets and wikipedia for winding wire the rating is max continous temperature.
Peak temperature would be higher.

The insulation class means the more expensive insulation can take more heat before it breaks down.

Also, even when a given Chinese company produces a useful product, they often post fuzzy specs. You might consider taking the posted specs with some healthy skepticism.

Xxxxxxxxxxxxxx

If your motor is cool all the time, then its probably larger, heavier, and more expensive than necessary.

The opposite is also true. If it is 200F all the time, then it is smaller, lighter, and not appropriate for the job you are giving it.

Hot copper wire has more resistance than "warm" copper wire, so an inappropriate system passes beyond a distinct tipping point.

Another side issue might be the thickness of the stator laminations. The common 0.50mm thickness has more eddy-current waste-heat than the less common thinner 0.35mm.

Also the kV can have an effect. All other things being equal, a higher kV (*low turn-count) fills the available stator-coil volume with fewer turns (shorter runs) of thicker wire.

So if we compare two similar hubmotors, the "slower" winding will provide "x" speed at full throttle, and the faster winding will provide the same top speed at a partial-throttle...however...the fast winding uses thicker wire and shorter runs of wire. This means a fast winding has less inherent resistance.

More resistance creates more waste-heat at the same power inputs between the two morors.

hias9 said:

Yes, but according to spec sheets and wikipedia for winding wire the rating is max continous temperature.
Peak temperature would be higher.

Click to expand...

You’re right, i had to check, 20000 hours at the rated temp should be enough. Then it’s pretty clear the chinese wire isn’t up to the said spec, or temp sensing inaccurate at higher temperatures.

I have calibrated the temp sensor with the controller. And yes, controller shows less than real on very high temperatures, but that’s already in the calculation. There should not be more than 6-7 degrees inaccuracy. Highest it ever was at the winding was about 145 degrees, but only for a very short time and that’s several months ago. First time I had problems was on a day where it was around 130 for a few minutes which made the yellow insulation plate become soft and some bamboo wedges came loose.

They probably use the cheapest wire available (rated 130 Celsius) and gave me incorrect information.
edit: They have now sent me a picture of the used winding wire. It is Chinese winding wire rated 130 degrees.

Mystery solved then! Might help others if you link the motor you’ve got. Could be advised to buy something else for the hotrodders.

It is the usual Chinese 3000W 45H with 23 pole pair design like mxus 3k for example, but with more copper.
Mxus also used 130 degrees wire for the DD motors by the way.