How quickly can hall sensors break?

Get all your technical information about electric bikes here.
Post Reply
Yermommy   1 W

1 W
Posts: 55
Joined: Mar 17 2018 1:18pm

How quickly can hall sensors break?

Post by Yermommy » Sep 19 2018 6:54pm

I replaced the hall sensors in my mxus 3k on August 5th and yesterday (September 18th) they broke. I run about 6k-7k watts through it. Does this seem normal?

dustNbone   10 kW

10 kW
Posts: 622
Joined: Mar 13 2016 9:59pm
Location: Vancouver, BC

Re: How quickly can hall sensors break?

Post by dustNbone » Sep 20 2018 12:15am

They are probably just getting too hot. I don't know if there are some other ones that can tolerate high heat, but that might be something to look into. All semiconductor components have temperature limits though.

Yermommy   1 W

1 W
Posts: 55
Joined: Mar 17 2018 1:18pm

Re: How quickly can hall sensors break?

Post by Yermommy » Sep 20 2018 7:41am

dustNbone wrote:
Sep 20 2018 12:15am
They are probably just getting too hot. I don't know if there are some other ones that can tolerate high heat, but that might be something to look into. All semiconductor components have temperature limits though.
The quoted temperature range is -20c to 125c and I never let my motor go above 90c. I guess maybe my set wasn’t made with strict tolerances.

User avatar
TommyCat   10 W

10 W
Posts: 93
Joined: Jan 22 2018 8:05pm
Location: North-west Illinois, USA

Re: How quickly can hall sensors break?

Post by TommyCat » Sep 20 2018 9:02am

Seems odd that all three would expire at the same time... Is your temp sensor reading what the sensors are actually receiving. Location and thermal conduction? Or perhaps riding it hard, then abrupt shut down causing thermal overshoot.
See my completed Magic Pie V5 rear hub E-Bike build Here!

dustNbone   10 kW

10 kW
Posts: 622
Joined: Mar 13 2016 9:59pm
Location: Vancouver, BC

Re: How quickly can hall sensors break?

Post by dustNbone » Sep 20 2018 4:16pm

Also maybe check the voltage going to the sensors. If they're getting overvolted (should be 5V going in) because the regulator on the controller is borked that would explain the short lifespan.

Yermommy   1 W

1 W
Posts: 55
Joined: Mar 17 2018 1:18pm

Re: How quickly can hall sensors break?

Post by Yermommy » Sep 21 2018 8:13am

dustNbone wrote:
Sep 20 2018 4:16pm
Also maybe check the voltage going to the sensors. If they're getting overvolted (should be 5V going in) because the regulator on the controller is borked that would explain the short lifespan.
That's possible but I can't be sure unless I take the controller apart.

The only thing that is strange is that the controller and throttle work for about 5-10 seconds of riding, then error out with a hall fault code. This has lead me to believe that the sensors are to blame.

User avatar
inedible   100 W

100 W
Posts: 139
Joined: Mar 18 2014 9:22pm
Location: Calgary, AB, Canada

Re: How quickly can hall sensors break?

Post by inedible » Sep 24 2018 10:34am

Have you considered that the problem might be mechanical? I've had something similar happen where the pins were snapping off the sensors due to vibrations and centrifugal force. Solved it by potting the pins using RTV silicone elastomer to dampen vibrations.

rumme   100 kW

100 kW
Posts: 1073
Joined: Jan 28 2011 7:47pm

Re: How quickly can hall sensors break?

Post by rumme » Sep 24 2018 12:35pm

I doubt I will ever go back to a ebike motor/controller setup that relies on hall sensors. My latest ebike build, uses a LYEN sensorless 18 FET controller, and it works flawlessly . Good Riddens to hall sensors and those stupid thin little hall wires.

dustNbone   10 kW

10 kW
Posts: 622
Joined: Mar 13 2016 9:59pm
Location: Vancouver, BC

Re: How quickly can hall sensors break?

Post by dustNbone » Sep 25 2018 12:44am

That's one of the things I like about the Grinfineon controller I use with my BBS02. It will work sensorless, but running sensored it's much smoother and quieter (I think it only runs sinewave with sensors). If there's a problem with the halls or their wiring, I can just unplug them and carry on.

User avatar
dogman dan   100 GW

100 GW
Posts: 34366
Joined: May 17 2008 12:53pm
Location: Las Cruces New Mexico USA

Re: How quickly can hall sensors break?

Post by dogman dan » Sep 26 2018 8:02am

When you run hard, then suddenly stop, the motor loses all wind cooling, and the inside temp will spike more than you might think possible. When I fried halls in smaller motors pretty regular, they typically died while the motor was stopped, despite my attempt to run a bit of cool down time at low speed and power. Basically, once outside temps got to about 40c, the motor would not cool enough, and I'd fry the halls in a motor. Why would cooking off your halls not cook off all three? Assuming they did cook off.

Even very small, like 3 mm holes in the hub cover can help a lot with quicker cooling of the motor when you stop, even if pretty much totally ineffective while the motor runs. It also lets you smell it when you stop, when you are cooking off the windings. poor mans temp monitoring, stop when you smell fairly toasty windings, before you smell burnt.

Yermommy   1 W

1 W
Posts: 55
Joined: Mar 17 2018 1:18pm

Re: How quickly can hall sensors break?

Post by Yermommy » Sep 27 2018 4:36pm

dogman dan wrote:
Sep 26 2018 8:02am
When you run hard, then suddenly stop, the motor loses all wind cooling, and the inside temp will spike more than you might think possible. When I fried halls in smaller motors pretty regular, they typically died while the motor was stopped, despite my attempt to run a bit of cool down time at low speed and power. Basically, once outside temps got to about 40c, the motor would not cool enough, and I'd fry the halls in a motor. Why would cooking off your halls not cook off all three? Assuming they did cook off.

Even very small, like 3 mm holes in the hub cover can help a lot with quicker cooling of the motor when you stop, even if pretty much totally ineffective while the motor runs. It also lets you smell it when you stop, when you are cooking off the windings. poor mans temp monitoring, stop when you smell fairly toasty windings, before you smell burnt.
My hall sensors actually died in the middle of a ride; I had throttle and power then all of a sudden I had nothing...

I replaced all 3 sensors with new, higher quality sensors by Allegro Microsystems even though I verified that only 1 sensor was broken. I believe the broken one was making intermittent contact which was allowing the controller to pick up its readout for a few seconds, then lose it. I did not pot the sensors but I am now worried that they will break again in another month or two. If I get stranded 20 miles away from home, I will seriously re-consider ever using a sensored controller again.

Thanks for everybody's help and input.

Bg-electric-cycle   10 mW

10 mW
Posts: 23
Joined: Jun 06 2018 7:08am
Location: Yambol

Re: How quickly can hall sensors break?

Post by Bg-electric-cycle » Sep 28 2018 2:32pm

I had a cable short while out on a ride 40km from home. Thankful was visiting a buddy and in the car the bike went.

Fried one hall sensor I also replaced all three.

Now run with a sensorless controller as a back-up. Although it’s got cruise control and regen therefore on occasions utilise a sensorless controller.

Does require some excertion as it’s not idea to try and apply power from a stop position.

Really need to get a meter into the Halls to attain what voltage is going through them as they should last the life of a motor. Subject to normal wear and tear.

User avatar
dogman dan   100 GW

100 GW
Posts: 34366
Joined: May 17 2008 12:53pm
Location: Las Cruces New Mexico USA

Re: How quickly can hall sensors break?

Post by dogman dan » Oct 03 2018 8:12am

You cooked em off, and will again. But since its fall, maybe not till next spring. 8)

Post Reply