Hub motor MXUS V3 3K Turbo and hot temperature (Problem solved!)

VelluK

1 W
Joined
Feb 25, 2018
Messages
64
Location
Helsinki, Finland
My MXUS V3 hub motor get really hot when driving by bike. I made 30 min test drive and controller cut the power to the motor couple of times. Motor temperature limit was set to 130 C degrees (266 F).

It's interesting to know is this normal temperature for the Mxus V3 motor or is it too much for the motor? What would be the good high temperature limit for this motor?

And is there any magic tricks what I can do for the motor to get the temperature down? I have already put 5 ml ferrofluid to the motor. Motor is 5Tx12 / 7,2 kV model and I have 21" moto wheel at back.
 
What kinds of voltage and amperage are you putting through? What controller are you using? Any other info I.e. weight of rider plus bike, hill grade, time to heat to 130 C at x kW?

Bad phase combos will in some cases appear to run fine but heat thw crap out of the motor. The Mxus halls and phases differ in color from other motors.
 
molybdenum said:
What kinds of voltage and amperage are you putting through? What controller are you using? Any other info I.e. weight of rider plus bike, hill grade, time to heat to 130 C at x kW?

I'm using 18s Lipos so voltage is about 66 V. I don't know the current exactly because I don't have any current/wattmeter installed but I guess current is about 50 A max. Controller is Kelly KLS7230S. Rider + bike weight is about 130 kg.

I have to check if I find some other way to connect the motor wiring and then test how the motor works.
 
VelluK said:
molybdenum said:
What kinds of voltage and amperage are you putting through? What controller are you using? Any other info I.e. weight of rider plus bike, hill grade, time to heat to 130 C at x kW?

I'm using 18s Lipos so voltage is about 66 V. I don't know the current exactly because I don't have any current/wattmeter installed but I guess current is about 50 A max. Controller is Kelly KLS7230S. Rider + bike weight is about 130 kg.

I have to check if I find some other way to connect the motor wiring and then test how the motor works.

I see from your other posts that you are indeed using the Kelly KLS7230S and that you have already checked the hall phase combos. I remember when the MXUS motors first became available (I bought one on the initial group buy). They talked of hall phase combos for the original MXUS V1 with Kelly controllers at https://endless-sphere.com/forums/viewtopic.php?f=31&t=66472&p=999176&hilit=mxus+kelly+phase#p999176. Do these match yours?

I don't think they changed the wiring colors for halls/phases of MXUS motors on later iterations, or at least I haven't heard of any changes...

I don't think the motor should heat to 130C at the power levels you are using, so something is amiss.
 
I run something similar and its not hard to get the motor temp to 90 deg (30min trip time, input 300wh, starting temp 24C, pretty flat and
plenty throttle on/off). No ferro or other mods on mine tho. What I mainly notice is that heat does not dissipate very well at all, it just keeps building up gradually. Staying at lower power limit, say 1000w will help a lot. My small mid drive heats up fast but also sheds the heat at like 1C degree per second or more at higher temps. My mxus is more like 1 degree per minute..

I would be interested in what the highest temperature that is considered safe, since the casing is not more than 30C at any time and the magnets are in direct contact with the casing. As long as I keep the magnets way under "too hot to touch casing" and what ever temp the isolation of the winding don't suffer I should be good, right? These magnets don't suffer permanent damage until what, 80-90C?
 
Here's the simulation for your motor with 50A at 66V no statorade. As you can see, this motor is not predicted to overheat at these power levels. Of note is that the motor stabilizes at 1000W at final speed in this sim (approaching unloaded speed) but runs higher watts when accelerating. I ran a crappy 1000W clyte HS3540 at these power and temps never got past 110C.

Mxux3k5Tferro.jpg

As for permanent demagnetization due to heat, this effect varies greatly on magnet quality. There is some discussion here: https://endless-sphere.com/forums/viewtopic.php?f=3&t=91625&p=1336402&hilit=magnet+heat+damage#p1336402

You should also check out the dedicated MXUS 3k threads: https://endless-sphere.com/forums/viewtopic.php?f=2&t=92101
 
Calculation shows that consumption should be about 24 Wh/km when my bike's real consumption is about 45 Wh/km. Battery capacity is 13,2 Ah and I can drive about 20 km when I should be able to drive like 40 km.

So there should be something wrong on my motor. But what? Do you have any ideas what to do next? I have to open the motor to see if there is something weird inside the motor.
 
Here is the quote from the original Mxus thread. I'm assuming the hall-phase wiring hasn't changed with the latest Kelly controllers or Mxus motors.

Do these match your wiring?

How many amps (or watts) does your motor pull when run unloaded?

Motors themselves are exceedingly simple and I know of no other fault that would cause these problems. It walks and quacks like a hall-phase problem. I doubt you would see a problem by opening the motor.

You can also try posting in the Mxus thread, as this would get the attention of numerous experts on this motor.

\\/ampa said:
Known Settings and Parameters (please help extend this. PM me if you have more settings and parameter to be added)

CAv3

Odometer

23 Poles

Temperature sensor CAv3

LinearMode
Units 0,67V
Scale 177,2 Deg/V


Recommended motor shut down temperature 120°C - 130°C



Halls and Phases combo's for various controllers:


Kelly Controllers

Phase wire
B = Yellow
C = Green
A = Blue

Hall sensors
A = Yellow
C = Green
B = Blue

For Infineon (both Lyen and Keywin) controllers

Halls:
Blue-Blue
Yellow-Green
Green-Yellow

Phases:
Blue-Blue
Yellow-Green
Green-Yellow
 
Here is the wiring:

Hall sensors
Motor blue - controller green
Motor green - controller blue
Motor yellow - controller yellow

Phase wire
Motor blue - controller green
Motor green - controller yellow
Motor yellow - controller blue

I don't know the current because I don't have any wattmeter but maybe I can use the multimeter the measure it.
 
I had a look at your hall-phase wiring and it is identical to what @\/ampa posted in the mxus thread; thus, your wiring is most likely correct.

Is your temp sensor displaying the right temperature? i.e when cold, it should display somewhere near ambient. Does the motor feel hot to the touch when it displays 130+ C?

More info here https://endless-sphere.com/forums/viewtopic.php?f=2&t=76133&p=1380620&hilit=Mxus+Temp+Sensor+Cycle+Analyst#p1380620

What are you using to display Mxus temp sensor data? Are you using cycle analyst?
 
molybdenum said:
I had a look at your hall-phase wiring and it is identical to what @\/ampa posted in the mxus thread; thus, your wiring is most likely correct.

Is your temp sensor displaying the right temperature? i.e when cold, it should display somewhere near ambient. Does the motor feel hot to the touch when it displays 130+ C?

More info here https://endless-sphere.com/forums/viewtopic.php?f=2&t=76133&p=1380620&hilit=Mxus+Temp+Sensor+Cycle+Analyst#p1380620. This thread indicates that the Mxus V3 uses the 10K thermistor for temp sensor. Earlier Mxus 3k motors used the KTY83/110 temp sensor.

What are you using to display Mxus temp sensor data? Are you using cycle analyst?
 
I had to wire my mxus 3k 4T with the yellow and green swapped on halls and phase, At 10s 40amp batt 80amp phase I get no detectable heat with the hand even when climbing at low speed stalling it out more or less if it weren't for the torque being able to pull out of it even on the steepest of hills, I run a 16 inch moped rim.
 
molybdenum said:
Is your temp sensor displaying the right temperature? i.e when cold, it should display somewhere near ambient. Does the motor feel hot to the touch when it displays 130+ C?

What are you using to display Mxus temp sensor data? Are you using cycle analyst?

Temperature sensor gives right value, at garage it's less than 20 C degrees when I have not ridden at all.

I read the temperature from Android app which is connected to Kelly controller via bluetooth.
 
VelluK said:
molybdenum said:
Is your temp sensor displaying the right temperature? i.e when cold, it should display somewhere near ambient. Does the motor feel hot to the touch when it displays 130+ C?

What are you using to display Mxus temp sensor data? Are you using cycle analyst?


Temperature sensor gives right value, at garage it's less than 20 C degrees when I have not ridden at all.

I read the temperature from Android app which is connected to Kelly controller via bluetooth.

This seems reasonable; thought I'd ask..


Ianhill said:
I had to wire my mxus 3k 4T with the yellow and green swapped on halls and phase, At 10s 40amp batt 80amp phase I get no detectable heat with the hand even when climbing at low speed stalling it out more or less if it weren't for the torque being able to pull out of it even on the steepest of hills, I run a 16 inch moped rim.

Although Ianhill runs the best wheel size/motor wind possible for heat and efficiency, the fact he needed to switch the halls and phases relative to earlier MXUS motors is very interesting. You should try as @Ianhill has done. In your case, it would be switching halls and phases to:

Hall sensors
Motor blue - controller green
Motor green - controller yellow*
Motor yellow - controller blue*

Phase wire
Motor blue - controller green
Motor green - controller blue*
Motor yellow - controller yellow*

Be careful when testing, especially when testing things unloaded; it is easy to blow the controller with the wrong combo as it can essentially pull unlimited amps in some scenarios. the proper way to do this is to limit amps to the controller by using an amp limiting power supply and not the battery. I've done this with a battery, but test at very low throttle and only do this briefly to establish which hall/phase combination gives you lowest unloaded power consumption and runs smoothly. I never go higher - or to WOT if the motor doesn't turn with a nudge of the throttle.

Also note that if your android device measures watts in real time, this will work to establish which hall/phase combo is most efficient on the road at a given throttle input.

See also:

https://endless-sphere.com/w/index.php/Determining_the_Wiring_for_a_Brushless_Motor

and https://endless-sphere.com/forums/viewtopic.php?f=16&t=3484
 
The real problem is the size of the wheel you're running the hubmotor in. Hubmotors ALWAYS perform better and run cooler in a smaller diameter wheel, and your MXUS is no exception. ie I run a 3kw rated hubbie in a 10" scooter wheel with an OD of about 16", the smallest profile 10" scooter tire I could find, and I'm pushing 150A at 115V nominal into that motor (over 17kw peak input) for my 25km daily commute over hilly terrain pushing over 350lbs of total load with performance that embarrasses all but the 1000cc motos, and my hubbie has never seen 130°C.

If I was running even just a 20" OD wheel, much less a 27" OD wheel, I couldn't run nearly the power I do. Sure my moto is more efficient and can handle the high rpm better, but that's only a small part of it, since I've also run hubbies similar in design to your MXUS in different size wheels. I have little doubt that my cargo bike with one of those pushing a heavier load could out run your ebike while maintaining cooler temps simply because I run a 20" OD tire on that ebike.

FWIW, before waking up to the advantage of hubbies in a smaller wheel, run a lower phase current limit on your too-steeply-geared hubbie and you should see lower temps, with little change in performance.
 
molybdenum said:
Although Ianhill runs the best wheel size/motor wind possible for heat and efficiency, the fact he needed to switch the halls and phases relative to earlier MXUS motors is very interesting. You should try as @Ianhill has done. In your case, it would be switching halls and phases to:

Hall sensors
Motor blue - controller green
Motor green - controller yellow*
Motor yellow - controller blue*

Phase wire
Motor blue - controller green
Motor green - controller blue*
Motor yellow - controller yellow*

WOW, wiring fixed the problem :D Now my bike is like a rocket, front wheel cannot stay on the road, top speed is more than before and motor temperature stay under 90 C degrees. Thanks a lot molybdenum, I'll give you beer when you next time visit in Helsinki :D

I tested the unloaded full throttle and I found the real time motor speed and phase current values from the Android app. Before the wiring change the motor rotating was really weird. When the throttle were about 80 % of maximum the motor rotated about 2400 rpm and the phase current was some amps, max 5 A. When the throttle were 100 % the motor rotation was only 1900 rpm and there were really weird sound on the motor like some negative force were braking the motor rotating. Also the phase current increased a lot, it increased maybe 10 times to about 50-60 A.

When I changed the wiring like Ianhill had documented the motor started to rotate smoothly. Phase current stayed low at unloaded full throttle and the real test drive was pleasure :)

I found a new problem, when I had some throttle and I set the throttle to zero and again gave some throttle when the bike were just freewheeling the motor made really strong braking and after that started to accelerate again. I didn't found any setting which would fix the problem. I think it's good to start a new thread because my original problem is fixed now :)
 
I'm glad you were able to get it working! From previous experience, I know the wire colors vary between manufacturers both for controller and motor wires. I wasn't expecting the V3 mxus to differ from previous versions - well, lesson learned and fortunate that Ianhill had come across this thread to add his experience with the V3.

Cheers, and happy riding!
 
Glad to be of assistance, it's a cracking motor I've really put the hurt on it only at 10s mind, Goes up some serious inclines and I can get upto 30 miles on the flat or 10 miles serious climbing with a nice ride back down on the stock 10s 6p pack that's just amazing.

It's great fun blasting past the convoy of drop handles as they climb, I bait them then on the flat into a cadence they cannot maintain after 2 or 3 miles I normally get a comment oi mate wtf is that.
_IMG_000000_000000.jpg
_IMG_000000_000000.jpg
 
Back
Top