Sports Biomechanics Lab > Blog > Hub motor woes
Personal tools

Hub motor woes

by Jason Moore — last modified Jan 11, 2011 05:00 PM

On our first day of testing the bicycle fell over and sheared the wires to the hub motor causing the motor to stop working. Here is a little story about what was wrong and how I fixed it.

null
From Bicycle Dynamics

We did the first day of preparatory treadmill testing on December 28 and things went fairly well. We tested the harness setup, using the bump with the treadmill, the pull cord and the data collection software. Things went pretty well until we heard a crash. I saw the bicycle falling from the corner of my eye and yelled out a lout "Noooo!" (or it may have been "Oh shit!", my memory eludes me at the moment). We put were using this lightweight bicycle stand to support the bicycle. The bicycle with the rigid rider setup weighs in at about 77lbs and is top heavy so it needs some serious support and the dinky stand didn't cut it.

null
From Bicycle Dynamics

Regular bicycles typically weigh from 15 to 30 lbs but with the variability of human weight, this bicycle should still give relevant data. We didn't see any visible damage to the bicycle so we took it back to the treadmill for some more testing, but the motor was not working and just vibrating a lot. I stopped and saw that the wires from the hub motor have been partially sheared.

null
From Bicycle Dynamics

I took the bicycle back to the lab to try to patch things up. I opened up the hub motor and rewired the the damaged wires. Then opened up the hub motor, snipped the wires below the bad spot and rewired everything with identical gage wire and colors. When I reinstalled the motor and tested it with no load, I found out that it was not working properly. The motor would spin up like it was working and the after a 20 secs or so it would peter out and the throttle did nothing until you released it and tried again, which produced the same results.

null
From Bicycle Dynamics

The first thing I did was to open up the controller box to check for burnt electronics. I found nothing out of the ordinary so I put it back together. Then, I opened up the hub motor again to see if I could find any damage to the motor coils or the hall effect sensors inside. I connected the red sensor wire to +5v and the black to ground and checked the sensor output with a permanent magnet. For the blue and green sensors I was able to get the voltage to change from 0V to ~0.14V depending on what pole of the magnetic was brought near the sensor. But for the yellow sensor the magnet seemed to not have an affect, leading me to think that sensor is damaged. But .14V didn't seem correct either. I couldn't figure out what type of hall sensors were in the motor because the numbers and letters on the tiny sensors were pretty impossible to read. The top line seemed to be "F41" but the bottom line was something like "9K06", "9K0G", "9KJG", or "9KCG". Google searches with wildcards didn't turn up anything. We finally hit a google search for "*41 hall effect" and found the Honeywell SS41 sensor that seemed to be the same physical dimensions as the ones in the motor.

null
From Bicycle Dynamics

Other notes on the electronics:

  • The controller seemed to supply +5v to the red, green, blue, and yellow lines (ground [black] as reference).
  • The sensors seemed to stay on or off until the opposite magnetic pole is brought near the sensor face.

The jackpot to figuring this out was an instructables article about a guy's e-bike motor that basically had the exact same thing happen to it as my bike: http://www.instructables.com/id/Electric-Bike-Hub-Motor---How-to-Replace-a-Hall-ef/. He has a Golden Motor (as opposed to my Amped Bikes motor), but the insides look uncannily similar. So I finally new what sensors they were and began reading all of the great documentation that Honeywell supplies including how to use the sensors with brushless motors. 

One main difference in this sensor to other sensors I worked with was the "sinking" output. I still don't understand it completely but a voltage is essentially applied to the output and it will go low when the sensor is on. This youtube video helped reveal the correct way to check to see if the sensors were working:

It turned out that the yellow sensor on my motor was in fact dead and while trying to inspect and repair the motor I managed to break leads on the other two working sensors. So I ordered 5 sensors from digikey (2 as backup) for a total of $15 including shipping. I carefully removed the old sensors and scraped away the glue that held them so that the new sensors easily slid in and then connected the original wires to the three sensors being careful not to overstress the leads. I used a bit of 5 minute epoxy to secure them back into there notches and rewired the hub motor like the original setup. I checked each sensor with a permanent magnet to make sure it work before connecting the three power cables to the hub motor. They worked fine and after I connected the power, the hub motor ran like new again!

null
From Bicycle Dynamics

I also made a space age wooden stand that is supposed to hold the bicycle better...but it isn't that great. Maybe I can improve it.

null
From Bicycle Dynamics

More repair photos can be found here: http://picasaweb.google.com/moorepants/BicycleDynamics#

Document Actions
Add comment

You can add a comment by filling out the form below. Plain text formatting.

(Required)
Please enter your name.
(Required)
(Required)
(Required)
Enter the word