Increased drag after DD motor was close to burn

hias9

1 kW
Joined
Jul 11, 2018
Messages
422
I was testing the temperature limits of a 3kw DD hub motor (+ Statorade+Hubsinks) and went up to 150 degrees Celsius. After that, drag was increased.
Consumption during no load speed (without field weakening) went up significantly from 300W to 1300W while no-load rpms stayed about the same. This happens both in sensorless and hall sensor based mode.
The motor still works fine, but has increased drag, gets hot more easily (less efficient) and vibrations at some speeds have increased.
I measured phase-to-phase resistance and it was about equal between the three combinations, so no short.
When I had the motor open the last time, the windings were already black at some spots, but it was still working fine.

Unfortunately I won't have time in the next weeks to open the motor to see what happened.
Do you have any ideas what could have happened?
 
You demagnetized it. Now you have a totally different winding that is faster but way less efficient.
100c is around the point when neodymium magnets undergo negative changes.

Sorry to break it to ya.

What kind of wattage were you running through the thing?
 
The magnets are rated 120 degrees celsius, but repeatedly up to almost 150 degrees (at the windings) resulted in no change in Kv at all.
Now Kv is slightly increased. No-load rpms are now only about 1% higher than before, but now it consumes 1300W instead of 300W at these rpms, so yes Kv increased slightly.
But how does this explain the increased drag? Shouldn't drag be less with weaker magnets?
I have stronger 150 degrees celsius rated magnets here, but no time to change them at the moment.
I run peaks a bit above 20kw.
 
Maybe your bearings got destroyed also by the grease liquifying and getting out.
 
I use quality brand bearings instead of the stock Chinese ones and the peak temperatures are not in the area of the bearings, so I would be surprised if it would be bearings.

Is increased drag normal for a motor with weakened magnets? I thought drag would be less with weaker magnets.
 
hias9 said:
The motor still works fine, but has increased drag

What does increased drag mean? Does it have to do with coasting or pedaling without power (or turning the wheel by hand without power)?
 
With the wheel up in the air, after accelerating to no-load speed it usually took maybe about 20 seconds until the wheel came to a stop.
Now it only takes about 5 seconds.
 
The exact nature of the change could be magnet damage, bearing seal damage, or a little of both.

You will have to disassemble the motor to find out what the issue is.

For those builders reading, I feel that 140F/60C is near ideal for a core heat limit, to ensure a long life.

For years, I have recommended 200F and 93C as a hard ceiling. 150C is certain to result in some type of damage.

Demagnetized magnets
Melted Chinesium solder
Damaged generic Hall sensors
Damaged insulation on the windings
Damaged seals on the shaft bearings, leading to rapid failure
 
For stock DD hub motors (without windings, magnets, hall sensors, insulation upgraded), a hard limit of 120 degrees celsius should be fine based on my experience.
Above 130 degrees celsius problems start.
On a geared hub I already had nylon gears destroyed at 90 degrees celsius motor temperature.

If efficiency matters, better keep the motor temperature low.
If you want to get the maximum power out of a motor, you have to go close to the motors operating temperature limit (while efficiency suffers of course).

When magnets got permanently weakened (demagnitization temperature would be over 300 degrees celsius), is it normal to have increased drag? I would have expected less drag.
 
line resistance increases in copper wire, as it gets hotter. This means a hot motor converts a higher percentage of input battery watts into waste heat.

Don Garlits was known for saying that he didn't care about efficiency, he just cared about winning, but...he would only race a 1/4 mile for a few seconds.

I do very much appreciate your posting of your results. All data is useful. Some manufacturers use higher-quality ingredients than others, so its good to know your brand could reach that heat level...
 
With increasing temperature, efficiency gets worse, but the motor case + cooling fins can dissipate more heat when hotter.
If you allow a higher continous operating temperature, you can push more power through it continously (while efficiency suffers however).
If you allow a higher peak temperature, this is nice especially if you live in a hilly area. Then you can ride hard until you reach the top and then give the motor time to dissipate the heat to the air while riding downhill.
 
Testing temperature limits does imply intentional damage, even if not testing "to destruction"
 
Yep, exactly. This motor is meant to test the limits of the components, to see which components need to be replaced if one would want to run it up to a higher temperature.
It worked fine before for 2 years with a maximum limit of 120 degrees celsius at the windings.
 
Not that I really understand what I am asking but ....
All those thin laminations that the coil wires wrap around have to be magnetically separated ... yes?
Is it possible that somehow the higher temperatures and pressure resulted in fusing some of those together ?
 
Very interesting. Yes, they have to be magnetically separated and a short would increase core losses.
But I never heard of a hubmotor where this has happened.
Does anybody know at which temperatures this starts to happen? I would expect this to happen only at much higher temperatures than the 150 degrees celsius the motor has seen.
 
hias9 said:
The magnets are rated 120 degrees celsius, but repeatedly up to almost 150 degrees (at the windings) resulted in no change in Kv at all.
Now Kv is slightly increased. No-load rpms are now only about 1% higher than before, but now it consumes 1300W instead of 300W at these rpms, so yes Kv increased slightly.
But how does this explain the increased drag? Shouldn't drag be less with weaker magnets?
I have stronger 150 degrees celsius rated magnets here, but no time to change them at the moment.

Amount of time in the danger zone spent determines the damage. It's summer and your motor probably spent more time in the danger zone than usual.

As for the rotational problems, check your bearings for signs of melting.

hias9 said:
I run peaks a bit above 20kw.

Whew, that's way too much for a motor that's rated 2.5kw in a 26" wheel and you're probably well beyond saturation ( the point at which you make more heat than power per watt inputted ) at 12-15kw.
 
It's a 24'' wheel and it can take about 5-6kw continous (at higher speeds) with Statorade + cooling fins.
20kw+ peaks are only for a few seconds when accelerating full throttle at highest level and motor heats up quickly of course.
As already mentioned, 2 years no problem with 120 degrees celsius limit.
Saturation should depend on phase-amp(-turns) not watts as far as I know.
Now I increased temperature gradually to see when the components fail. I have some spare motors, but plan to uprade the parts that cannot take 150 degrees (at the windings) for a few minutes.
 
hias9 said:
Saturation should depend on phase-amp(-turns) not watts as far as I know.

Yes, and your phase amps need to be very high in order to make that 20kw.

hias9 said:
Now I increased temperature gradually to see when the components fail. I have some spare motors, but plan to uprade the parts that cannot take 150 degrees (at the windings) for a few minutes.

Open up your motor and let's see it's guts. If bearings are not your problem then windings or wire insulation may be. Some other components in the motor may not be rated to 150c.
 
I will open it when I am back in about 2 weeks. The last time I had opened it (after it was around 135 degrees for a few minutes), the yellow insulation plate got soft, some bamboo sticks got loose (removed them all) and windings (claimed to be rated 180, but manufacturer corrected to 130 degrees later) were already black at some spots. Hall sensors are rated 150 degrees and no problem at all so far. Kv was still completely unchanged after windings were at 145 for a few seconds and at 130-135 for some minutes.

I would be surprised if it were the bearings because I replaced the Chinese ones by quality brand bearings.
 
Right, those other materials that prevent your motor from shorting out are not rated to stand up to he kind of temperatures it's been experiencing for a long time..

Check for a phase short because i'd guess you probably have one if your windings were starting to get extra crispy before.
If you got the motor that hot, you absolutely demagnetized the motor to some degree, and had lower efficiency before, which would have started a death spiral of progressive inefficiency and heat damage to the magnets.

I'm impressed that the motor handled that for so long!
 
Please check the first post. I measured with a milliohmmeter and phase-to-phase resistance was equal over the three combinations. Motor still works fine, but has higher drag + vibrations at some speeds.

Comlpete demagnetization happens way over 300 degrees celsius for neodymium magnets as far as I know. But magnets can get permanently weakened earlier. Also in this case, but I would expect lower not higher drag with weakened magnets.
These magnets are rated 120 degrees and I have stronger magnets here with 150 degree rating.

I only had increased no load comsumption the day before it happened (about 20% increased. now about 300% increased).
Before that, even after a few seconds at 145 (at the windings) and 130-135 for some minutes, absolutely no permanent change in Kv and no load consumption.
As mentioned before, it was used with a 120 degrees limit before I started that test with this motor a few months ago.
As long as you always stay below 120 degrees, the stock motor should be fine, no matter how many amps you push through it.
 
Oh my bad i didn't read all of the post.

If you saw a 20% increase in your motor no load speed then those magnets were already affected pretty badly.
The MAC from 2011 i ran at 4kw for 5 minutes was never right afterwards and had a 15% speed increase, which was nice except for the greatly reduced range and how it would heat up so easily from then forth.

Anyway let's see it's guts when you get time.. the pictures will tell the story :)
 
The day before (the motor also was above 140 degrees that day), a few hours after riding, there was no 20% increase in no-load speed, but in no load consumption (ampere and watts). No-load speed was unchanged.
Even now no load speed is only 1% higher, but consumption while the motor is spinning no load (without field weakening) is 300% increased. Kv has slightly increased by about 5% in this case.
 
Its definitely toasted. I mean the magnets. Sounds a lot like what happened to some brushed motors I cooked, before the final death by shorted windings.
 
I finally had time to open the motor and replaced the stator by a new one.
Seems like the magnets are absolutely fine. Kv was only slightly increased and is back to normal now.
The old stator which I ran up to 150 degrees Celsius had extremely increased eddy losses:
1300W consumption at 950rpm
350W at 430rpm
50W at 175rpm.
(Measurements without field weakening of course)

Now with the new stator consumption at 950rpm is around 250W (now also without Statorade).
The stator has 21 strands in parallel and 3 turns. The windings are kept in place by little cords which seems to have burned on this stator and cut through one strand.
Winding resistance was only slightly increased and was still about equal through the three combinations. There does not seem to be a short.

Does anybody know the reason for the highly inceased eddy losses? How can I check if the laminations are still magnetically separated?

Also there seems to be only little Statorade left at the magnets (I put in 15ml about 2 years ago). Covers are sealed and the ferrofluid moved to other places in the motor like phase wires and bearings outer ring.

Pictures of old and new stator to follow. The one with the dark windings is the old stator ;)
 
Back
Top