Charcoal bags

Charcoal bags

Search This Blog

Burning out Bafang hub motors

The short story

Bike hub motors can overheat if you work them too hard, too slow. This post is about how motors heat, and how to avoid overheating. It is focussed on standard 250w Chinese geared hub motors - like the vast majority of ebikes around the world use.

What makes a bike motor get too hot?

  •  As a standard 250w hub motor slows down, it draws more current (up to a limit set by the controller – about 15A which it reaches at about 17km/h).

  • As it slows down further, the motor draws the same electrical power but gets less efficient, so less of the electrical power leaves the motor as horsepower, and more of the electrical power becomes heat in the copper windings (according to some data I’ve seen, at some point under 10km/h half of the power is going to heat).

  • In a 250w Bafang hub motor going less than 15km/h at full throttle, the winding temperature is rising fast (I measure about 0.5°C/sec).

  •  If the motor is laboured like this for more than a couple of minutes, the motor windings will reach high temperatures which risk damaging the wire insulation.

  • You can’t tell how hot the motor windings are from the outside: the temperature of the outside of the hub motor can be lukewarm to touch, when the windings are over 160°C.

  • If the winding insulation is damaged, the electric current can start taking shortcuts between wire loops, making the windings even hotter, losing power, and cooking your motor until the wires smoke.

How do you avoid cooking your motor?

  •  If you are riding your ebike at full throttle, under 15km/h, for more than about 2 minutes, give your bike a break. Take a break about every minute of riding if the hill is long.

  • Walking while motoring your bike works fine: the motor will keep cooling as you go. Sometimes this is the best solution (especially when carrying passengers – let them walk). There is no shame in walking a bike up a steep hill!

  • The motor windings cool very quickly once you stop, especially if the motor hasn’t been hot for long.  

The long story

A run of burnt motors

My biggest problem with e-bikes lately has been cooked motors. I think I now have some sort of grasp of the problem:
  •  why this is happening: riding too slow for too long up too-steep hills,

  • how to prevent it: stopping and givng the motor (and rider) frequent rests on steep hills

  • how to fix it: rewinding burnt motors – I hope!

I now have 5 motors in the storeroom which have failed through overheated windings – actually 6 if I count a non-Bafang motor which seems to have failed through a combination of water damage and burnt windings. All of these motors have failed after use on our mountain, with long steep climbs.

Unusual conditions

We do have exceptionally difficult conditions for riding bikes where we live. The mountain roads include inclines of 15% and more, which mean riding at speeds down to around 10km/h with our 250w Bafang motors and hard pedalling, especially when carrying loads. The biggest challenge is the “PEI Road” (stands for Public Estate Improvement: the road was built as an employment program during the Depression). The PEI Road climbs up to Mount Glorious from Highvale (coming from Samford), and its stand-out feature is a continuous incline of 15% - 17% for about 2km. I tend to maintain between 10 and 12km/h on this section, which takes over 10 minutes of hard pushing at full motor throttle if you don’t stop (these days I stop…).
My son Jasper rides his Dahon (Bafang SWXH 250w rear motor) home from school up the PEI road most school days (about 17km each way), pushing hard and not stopping till he gets to the top of the main incline before taking a rest (to prepare for the next huge hill…). Jasper has burnt out 2 SWXH motors this year, each after 2500km of this treatment.
My Xtracycle’s first rear SWXH hub motor survived a similar distance (about 2700km) of heavy mountain riding. Then it was fully cooked:
Bafang SWXH rear hub motor with blackened windings - and a terrible smell
I’ve never ridden it up the PEI Road, but I use it to carry our youngest child Luka (now nearly 30kg) to and from Mount Nebo, 11km away. This means travelling slowly for long climbs – which I now clearly understand as the way to kill motors. As Luka has got heavier, I’ve given up trying to carry him up the steep hills, and now routinely let him get off and walk. I now realise this is the only way I’ll get this new motor to last a long time.
[Since this post I changed motors on my Xtracycle to a Bafang BPM, which has done over 11,000 km without any problems. My page on BPM motors is here, my page on my Xtracycle is here. Our son graduated to his own ebike - here is my post on kids' ebikes]

Measuring winding temperature

So that I could get some real information on what was happening inside my motors as they heated up, I bought and installed a thermocouple to measure the temperature of the motor windings as I rode.
A thermocouple uses the junction between 2 dissimilar metals to determine temperature. A meter measures the voltage created at the junction, and translates this to a temperature reading. Effective thermocouples can be bought on Ebay for about $6.
The cheap ebay thermocouple - works well
The probe which measures temperature is tiny – just the end of a pair of twisted wires.
To measure the temperature of my Bafang SWXK 250w front motor, I drilled a 2.5mm hole in the phasewire cover plate on the RHS of the motor (after checking for the best spot). The probe wire then entered the motor through one of the phase wire holes, and the probe was inserted amongst the windings in the copper motor coils. This way the probe is measuring the actual temperature of the copper wires, rather than the temperature of the iron core, or the aluminium case. The wire temperature is what most matters: this is the heat which will damage the thin layer of insulation on the winding wire (often titled “magnet wire”) and cause current to cross between windings and motor failure.
Here is the probe wire entering the motor:

Here is the thermocouple mounted on the bike frame:

This way I can watch the temperature of the windings as I ride (occasionally glancing at the road ahead...).
The ideal way to measure this would be to log battery current, windings temperature and speed over time. However I think you’d come to the same conclusions.

What temperatures does the motor reach?

Most ebiking involves continuously changing conditions: speeding up, cycling above motor speed without power, slowing down at a hill and turning on the power. The windings heat under power and cool when unpowered or at high speed.
Here are some observations:
  • In normal undulating conditions the windings rarely exceed 80°C.
  • Intermittent climbs where speeds are over 15km/h most of the time, get windings over 100°C, which are fine.
  • Once the power is on and the speed has bogged down below 15km/h for extended periods, the temperature climbs by around 0.5°C/sec. Temps will reach 160°C and higher if low speed continues uninterrupted.
  • The motor could be super hot - 160°C - in the windings, but be lukewarm to touch outside.
  • Short stops help a lot, with temps falling fast.
  • Walking up a steep hill while powering the bike alongside you is really effective. The motor isn’t working at an efficient speed, but the power is low so only small amounts of heat are generated.

Bigger motors

A bigger motor can help with the overheating problem. I've been using a Bafang BPM 350w motor on my Xtracycle for several years with no problems at all. This motor is sold as a hill-climbing motor, and is clearly more comfortable with climbing. However I haven’t tried measuring real temperatures in it, and this wouldn’t be easy with a back motor as the wires enter the motor through the axle. It would be possible though, especially if I removed the 5 hall sensor wires which I don’t use (it works fine in sensorless controller mode).  

2-speed hub motors

For a while now I've been using a Xiongda 2-speed hub motor on my Dahon. This can change gears internally so it can climb steep hills without problems (so far). See my post at http://bruceteakle.blogspot.com.au/p/blog-page_20.html

Fixing burnt motors

I have enquired about obtaining replacement armatures for Bafangs (they are easy to remove), but have been told that Bafang doesn't sell them. This is a pity, and leaves rewinding as the only option unless a spare armature can be found. 
I've been using Xiongda hub motors recently, and Xiongda will sell replacement armatures (not that I've needed one). 
I've done some careful dismantling of burnt motors, and hope to try rewinding. I plan to report on that in another page.

28 comments:

  1. Thanks for sharing your experience. I'm just starting recently learning all about emotors on bikes. Planning to get out of the town and on the local hills, South Downs UK.
    What have you got to lose by trying a mid drive, which from what I read are much more suited to hills ? Also have you tried regen? You live in an ideal location to use the hill to recharge and save wearing out brakes. I have a cheapo 250w 36v front hub motor which I fitted to a used MTB frame. The controller that came with it is a regular square wave with no regen so I've also bought a sinewave/regen controller to try out next.

    ReplyDelete
  2. Hello David, good to hear from you.
    I haven't tried a mid drive, but this is my perspective on mid drives:
    - can use the gears and climb well
    - can use any gearing system, but
    - greatly increased wear on drive train
    - greater cost
    - greater complexity
    - can reduce rider control of cadence
    Hub motors are so cheap and simple, it's worth using them whenever possible. Situations with unusually long, steep climbs (like mine) are very rare. I am currently using a Xiongda 2-speed hub motor which has sorted the steep climbs very well.
    See my post at http://bruceteakle.blogspot.com.au/p/blog-page_20.html

    I haven't tried regen. Regenerative braking requires a direct drive motor, whereas pretty much all low-powered (legal in Australia) motors are internally geared and have freewheels, which make regen impossible (unless you ride down hills backwards). Direct drive motors are much bigger and heavier. I have considered regen carefully and I don't believe it is rarely of use. Remember, regenerative braking is braking - how much do you use your brakes on your bike? For most of us, hardly at all, so there is hardly any energy to be generated. Again, my situation with the mountain is very unusual where I have long descents where regen could be used - except that the biggest descent is generally at the beginning of my trip when my battery is full.
    So my advice remains to use a geared hub motor, don't worry about regen.
    Re your cheap front hub motor: it sounds like a geared motor with a freewheel which couldn't do regen. Does it freewheel when spun forwards?

    ReplyDelete
  3. Any ideas as to drilling holes in the motor housing to draw cool air in?

    ReplyDelete
  4. Hello J, holes in a motor case would let heat out, but would also let dirt and water in. There are plenty of people experimenting with ventilating their motors, whose stories can be found in places like endless sphere. My preference is to focus on durability, and I'd rather keep the water and grit out. I think this can generally be done by choosing the right motor for the job and not over-powering it.

    ReplyDelete
  5. Hi Bruce, have you thought about running 2 hub motors? I don't notice the extra weight. Full traction on a steep dirt hill is fun. Kind of like the hand of God is holding my seat and pushing. I don't get 2 x the distance, but close. It made my 35min trip a 30 minute trip.

    ReplyDelete
    Replies
    1. Hi Scott, I've daydreamed about using 2 hub motors but not tried it. Have you described your setup anywhere I can see? I worry about carrying 2 heavy batteries. I've also thought about having 2 motors of different gearing, one for the normal hills and another for the extreme climbs. Our big challenge here is several km of approx 15% gradient to climb, which is a challenge for most motors.

      Delete
  6. mine has a 3 speed (electronic) PAS 8fun 36v 250w front hub, and the cheap 48v 1000w rear hub with a twist accelerator. On the steep hills I tend to use low speed on the fronts controller and slight twist on the back motor to keep the front from labouring. The 1000w motor isn't happy going slow, but there seems to be a balance when equal energy is applied, just enough to help the front wheel. The battery for the rear is 10kg of lithiums, I doubled it from 5kg, cant feel the extra weight difference and certainly don't feel voltage sag under load. A little dirt track I use along side a 200m stair case, probably over 20* and 30* incline I sit all the way, just pedalling to engage the front and help the back a little. In general my bike likes sitting between 30 and 50km/h on the flats. I think geared hub motors, even 3 on a trike would be great for your trip. Regen braking is the option to look at for hills, Im sick of replacing brake pads. Also, These mid drive motors talk about torque, but dont really got much. Weight of the battery packs isn't really an issue if they don't have to work hard, and they live longer.

    ReplyDelete
  7. I know this is an old post, but hoping you're still monitoring the comments. I'm looking at my first ebike commuter with a budget of $2000(USD). I have one major climb on the last 1.5 miles of my 100% all-asphalt street ride home with an elevation gain of 435 ft. translating to a 5% elevation grade. I've made it on a non-ebike in the lowest gear of my 24-speed Trek but it's all I can do to maintain 4 mph. My dilemma is whether to go with a geared rear hub or mid-drive to make the climb and then what size motor (250W, 500W or 750W). This will be a once daily climb and I don't want to be over-stressing the motor resulting in a premature replacement. I've had one mid-drive bike manufacturer tell me a gear rear hub won't make it, and I've had several geared rear hub riders tell me a rear hub will handle it just fine. For the record, I'm 220lbs and may have an additional 5-10lb in cargo.

    ReplyDelete
    Replies
    1. Hello Barry, a 5% gradient is a cinch for an ebike - that's the gradients they are made for. My troubles have come from frequently climbing 15% gradients for several kilometres. A 250w Bafang SWXH or Xiongda single speed should take you up your hill no problem, assuming you're putting some effort in yourself - and you can expect to get stronger with time.
      Mid drives are fast on the flat, great on steep climbs, but expensive, complex, and wear out drive trains quickly. Hub motors are cheap to buy, cheap to run, but have a narrower speed range - fine for the modest and frugal cyclist.
      If you are converting a bike yourself, $1000 should be plenty. If you are converting your 24sp bike you could use a rear hub motor with a spin-on 14-28t 7speed cluster (like I use on 2 ebikes here in the mountains). Cheap, durable, simple. Your 8speed shifter will work if you set the derailleur limiting screws to the 7 speed sprockets - 7 speeds and 8 speeds have the same spacing. Put the controller in a bag and the battery in a pannier and you're on the road - refine with time.
      Good luck! Bruce.

      Delete
    2. Thanks for the quick response. My options are already limited being in Honolulu. The LBS that sell ebikes are either selling high-end bikes $3k+ or unbranded bikes with their own shop name slapped on them. I have more options within my budget online and then I have to contend with shipping charge. Looking at the Voltbike Yukon (750W Bafang) and the E-Glide ST (500W Dapu motor). DIY is definitely an option that I need to look closer at. Thanks again for the help!

      Delete
  8. have you tried cleaning out the grease and putting 30 to 50 ml of auto transmission fluid in the motor then sealing it up well with oil resistant automotive sealant, some others on ES have been doing this and have had excellent results as the fluid gets sprayed around the inside of the motor carrying heat off the stator and to the casing massively improving the cooling.

    ReplyDelete
    Replies
    1. Hello Punjehl, I think this is a good idea. I have tried ATF in 2 of my Xiongda 2-speed motors, so far going well. I didn't do anything to seal the motors (the Xiongdas have a threaded end cap) and I haven't had significant leaks. I'm sure they'd leak if the bike was laid on its side. Using ATF makes good sense to me: the pathway for heat from the armature to the outside is very difficult without a fluid to carry it, as evidenced by my experience that a motor with 160C windings can feel hardly warm to the touch. It also makes sense that the manufacturer doesn't use a fluid lubricant like ATF because then they'd have all sorts of leaking problems.

      Delete
  9. just get an 750w internal gear motor, problem solved

    ReplyDelete
  10. My bike has a BFSW motor, a 350 W motor power reduced to EU legal 250 W. If I climb hills, I reduce the assist mode from 5 or 6 to 3 or 4. The motor's power output is not really afected by this, but the controller reduces the current a lot. So you loose less range, and the motor is not getting too hot.

    But nevertheless, on very steep, long hills mid drives have definitely their advantages.

    ReplyDelete
  11. Bruce,
    As a recently retired aerospace test engineer I want to compliment you on your test methodology and conclusions. Your work is very helpful. Thank you.
    Recently, I appeared to have burnt out my 750w Bafang geared hub motor. The coils measure shorted. I will be getting my new motor today. After I get things going again I will perform an autopsy on the motor to confirm overheating. But based on your writings I will bet that is the problem.
    I have had my bike since 2017 and put on only about 900 miles. I also ride a manual mtb.
    I weigh 215 lbs, 6’ and I am 64. I am classified as a Clydesdale in racing lingo.
    I have tripped the thermal shutdowns probably 3 times on this motor. I should have taken action then.
    One response to your blog stated that a 750 motor will solve the problem. I can witness that may not be the case.
    In climbing some of the hills in my area I may need to climb about 2,000 – 3,000 feet which will take me quite some time. The grade is about 8% and I am traveling about 10 mph (16.7 kmh). That is above your recommended speed to maintain temperature.
    I think you are still correct about slower speeds produce more heat.
    I may be using the throttle at up to 750 w much of the time. I am peddling as well as I can but I have to haul my lard ass up that mountain.
    Going to a bigger motor does not necessarily solve the problem. It seems to me that with the 750 w motor the thermal mass is increased but the surface area is not proportionately increased. So how does heat transfer to the ambient air?
    Now that I am looking into it, I think the thermal management of these motors is deficient. I am currently looking for some heat sinks to mount to the hub to help out.
    Another thing I have learned is that Radrover advertises the Bafang motor as a 750w motor, but it turns out that is the peak rating not the continuous rating of 500w. One solution would be for me is to use PAS at 3 which will limit the power to about 500w and do not or intermittently use the throttle.
    I may also lower the crank gear so I can maybe add more peddle power with a higher cadence and lower torque (like Chris Froome).
    I have read that adding more grease to the motor will help heat transfer (like your bloggers have added ATF). This would void the motor warranty but that may be the best deal. I think the lower viscosity ATF may be better than the grease because of the added mixing but I worry about material compatibility.
    Solutions I will be investigating:
    1. Loose weight – I have been trying that for years
    2. PAS – set to 3, no throttle. May have to walk once in a while but for me that would delay the fun ride down the mountain too much and the day would be hotter.
    3. Heat sinks – should help
    4. It would be nice for motor controllers to be smarter. The bike should be protected from expected failures. The thermal cutoff switch delayed but did not prevent the failure. A smart controller could monitor speed and current draw and automatically limit current based on a predictive heat profile.

    Anyway, thank you again for your fine work. I plan to make a video for other Radrover owners as Rad Power was not as helpful as I hoped in trouble shooting my failure and I did not find much Rad Rover specific videos on motor failures.

    Thank you again for your good work.

    Mike

    ReplyDelete
    Replies
    1. Hello Mike, thanks for your thoughtful comment.
      What sort of Bafang do you have? If it’s a Bafang BPM or CST, it may have a code rating (I’ve described this in my post on BPMs:https://bruceteakle.blogspot.com/p/blog-page_8.html) that describes the number of turns in its windings and sets its speed and torque characteristics. If you have a fast BPM with fewer turns it will take you up to 30km/h (code 10 in 26” wheel) or faster (code 8) on the flat, but will be more prone to overheating at low speeds. If long, steep climbs are part of your e-cycling you may need a motor with more turns, that goes slower on the flat but has good torque and higher efficiency during the climbs so stays cooler.
      You’re right: the old-style Bafang hub motors are poor at losing heat. The coils where the heat is generated are inside the magnet drum, so heat can only escape by conducting along the axle, by radiating from coils to drum to hub case, or conducting into a fluid lubricant + coolant that carries heat to the case where it can radiate away (I’ve been using automatic transmission fluid in some of my motors over the last few years, which seems to work well, but I think you’re right to be concerned about materials compatibility). It’s really hard to improve heat loss from these motors - I doubt heat sinks on the motor case would help.
      Despite this, I think they are a good design for moderate work: , cheap, strong, easy to wind or rewind, easy to repair, durable if not overheated.
      There are newer Bafang motors I’ve read about which can deliver great power, with a very different design. These have the windings on the outside of a rotating magnet drum, so heat can more easily escape (Bafang BBS mid-drive motors are like this too - we have several of them in pieces here currently).
      You’re right about motor power too: I’m only aware of BPMs being labelled 350w or 500w output power by Bafang, and I suspect the power difference is only about being run at 36 or 48 volts. My 36v BPM uses about 650w max electric input with a 350w (motor output) controller.
      In summary: if you’re using a hub motor, you need the right motor speed/winding to handle the slopes you’re climbing, which may involve losing some speed on the flat.
      Good luck with getting your bike back on the road.

      Delete
    2. Since posting I made a video to share my experience with a burnt out motor.
      https://www.youtube.com/watch?v=J60KgOeoT-s&t=23s

      Delete
    3. Thanks for posting your youtube link Mike, it's interesting. I see your motor isn't BPM or CST, it looks smaller. Let us know how your heat sink goes. Bruce.

      Delete
  12. This comment has been removed by the author.

    ReplyDelete
  13. Thanks for your blog.
    I have a RadRhino which has a Bafang geared motor rated at 750W, and run into the same problems. Some of our roads and trails rise at an average >10% (which means sometimes up to 15% grade) over 8 km to 1000m. I had the motor after travelling between Pas 3 -5 eventually cut out at 4.5 km. I inquired and was told that actually the motor does not have a thermal protection, it is the controller to protect the mosfets from the higher amp draw (730W at PAS 5) The problem installing a controller that allows draws to 1000W will leave the motor vulnerable to overheating. I have looked into a higher geared (lower speed mor torque) motor but at 750W the seem all rated at 80Nm. I guess adding a second lower teeth chain crank might help, however due to the config of the Rad it might have to be engaged by hand before the ride...

    ReplyDelete
    Replies
    1. Hello Peterm, it seems unfortunate that this keeps happening. I suspect it is a case of too high speed windings for the long steep climbs. I don't think it would help to put lower gears on the pedal drive (unless you changed the motor characteristics) as the motor only knows what speed it's going and what current it's carrying: going slower will make the overheating worse. I see that Greenbikekit.com (my usual supplier) is sold out of BPM motors, but at bmsbattery.com you can get BPM2 motors in 3 different speeds, which might work in your situation: https://bmsbattery.com/motor/351-15707-bafang-bpm2-48v500w-rear-driving-e-bike-hub-motor-ebike-kit.html#/214-rpm-328

      Delete
  14. Thanks for the helpful post! I need your opinion please. I have a 350w bafang motor on my ebike and the motor started to smell like something was burning after going uphill for about 3-4 minutes on THROTTLE ONLY. The bike seems to be working fine, but I am certain something in the motor got burned/damaged. Do you think the motor will be ok if I never do this again? This is the first time it happened. The bicycle is only 2 weeks old and has 200 miles on it. Battery is a 36V 13Ah. Thanks

    ReplyDelete
    Replies
    1. Hello Prague, not being able to smell it myself I'm guessing, but I'd be surprised if your motor windings are guilty in this case. The motor is inside the metal hub casing that's pretty smell-proof, and I'm guessing your motor is a Bafang BPM (like this: http://bruceteakle.blogspot.com/p/blog-page_8.html) that are very robust. I'm guessing the problem is outside the motor.
      Did you make the ebike yourself? If not you'll need to get to know it. I've often had roasting of connectors through high resistance, perhaps corrosion or poorly joined connectors. Especially check the path of power from the battery, to the battery mount (down tube batteries can cook where the power connects to the battery mount), to the controller, and the 3 phase wires from the controller to the motor. Look for plastic - wire insulation, plugs - that has been discoloured or reshaped by the heat.
      On such a new bike, perhaps a plug hasn't been pushed together properly.
      Going uphill briefly has probably caused enough current to flow that the bad connection has heated, making the smell that tells you to fix it. If careful inspection doesn't show you the problem, you could ride up the hill again to generate the heat again and sniff your way along the path of power, from battery (could be inside the battery) to controller to the motor plugs.

      Delete
  15. I own the Rover with a 750W Motor, and this motor (or the controller) switches off automatically when the amp draw is too high to protect the motor. It can get really warm, but after cooling down for a few minutes the motor is operational again. I guess one method to check he motor is to measure the resistance of the coil against specs to determine if anything is damag3ed

    ReplyDelete
  16. I have a 36v 250w CSC front hub motor which over heated going up a long hill,the windings are nearly black .it will run but under load it makes a loud screeching ,I previously resoldered a joint on the hall effect board

    ReplyDelete
  17. Bruce Teakle'S Pages: Burning Out Bafang Hub Motors >>>>> Download Now

    >>>>> Download Full

    Bruce Teakle'S Pages: Burning Out Bafang Hub Motors >>>>> Download LINK

    >>>>> Download Now

    Bruce Teakle'S Pages: Burning Out Bafang Hub Motors >>>>> Download Full

    >>>>> Download LINK fW

    ReplyDelete