Hub noise/knocking?

[averhamdave]

Purely out of interest could someone explain in simple terms why the Bafang hub motor on my old Ezee knocks and grumbles badly below about 10mph?

Some background: The bike is an Ezee Torq Mk1. I have a ECrazyman sensorless controller fitted (I did this as I had problems with the hall sensor wires). The bike runs well and has done hundreds of miles in this format.

After running for a period - I think more pronouncably having run it at highish speeds, say 15/20mph plus - when I slow below 10mph the motor starts knocking and shuddering, worsening as it slows. Sometimes I even apply some throttle input whilst simultaeneously applying the front brake to minimise the effect. It can be quite severe and embarrassing!

I appreciate the issue is due to the non use of hall sensors and I'm presuming is due to the build up of some magnetism?

I would be interested in an explanation of what is happening and any suggestions how I may minimise it - if at all possible?

 

[neptune]

This reply is a bit speculative , and I am inviting being shot down in flames . Your hub motor was obviously designed to run with hall sensors . I am assuming that you know what hall sensors are for . You have changed to a controller that does not use hall sensors , but relies on back emf to sense rotor position , and thus determine when to fire each of the 3 sets of coils in the stator . I have read somewhere that this is notorious for being a bit hit and miss at low RPM . If I am right , you can only cure the problem by repairing the hall sensor wiring , and if necessary , replacing the hall sensors , and then refitting the original controller .. That is my two pennorth . Anyone know better ? Another possibility is that due to a fault in the new controller , or a broken phase wire , only two of the three sets of stator coils are being energised , That would cause lumpy running at low revs , but would be smoother at high revs .

 

[flecc]

I agree with neptune on using a non-Hall controller on this bike, the fact is that the Torq 1 motor is very overgeared since it's a 20" wheel motor in a 28" wheel to give it the 22 mph high assist speed. That means it's revs are much lower at low speeds than most hub motors, so accentuating any phase firing roughness then.

As for what is happening, I think it's just the poor timing of phase current pulses at low revs due to inaccurate reading of the rotor postions then. The mis-timed pulses just tend to snatch the rotor back and forth, creating the noise and vibration.
.

 

[Deleted member 4366]

If it makes the noise as you accellerate, it's probably because of the timing of the controller. If it does it when you decellerate, I reckon that your gears need re-greasing because when your throttle's shut, the controller doesn't send out any pulses.

 

[averhamdave]

If it makes the noise as you accellerate, it's probably because of the timing of the controller. If it does it when you decellerate, I reckon that your gears need re-greasing because when your throttle's shut, the controller doesn't send out any pulses.

No, its not that sort of noise. Bearings have always rumbled, its an old bike.

It definitely sounds like a timing isue but as D8veh identifies there is no power there when the throttle is shut. Its much much worse when decellerating. From about 10mph down to standstill it really does shudder - it even does it when immediately coming to a stop.

I'm happy to live with it and its still my favourite bike - and I have a really nice original one as well. Hall sensor wires are long shot so that's not an option. I'll live with it but am curious to know what's happening.

The lost phase wire doesn't sound right really, it works too well otherwise. The bike is powerful and fast.

 

[averhamdave]

Cpould there be a regen action occuring I wonder? Is regenerated energy being stored in the controller and released regardless of throttle position?

 

[flecc]

As its with the throttle closed, I wonder if the gear teeth are the problem since stripped or damaged gear or rack teeth on these Torq motors is quite common when old?

 

[Deleted member 4366]

Sounds like your clutch is jammed too. Does your wheel spin freely when you lift it off the ground and spin it by hand?

 

[averhamdave]

Sounds like your clutch is jammed too. Does your wheel spin freely when you lift it off the ground and spin it by hand?

Mmm. I'll check that tomorrow. I'll compare both bikes as they are identical other than the noisy one being sensorless.

 

[Deleted member 4366]

The clutch should disengage when you decellerate, so the motor shouldn't be able to make any noise. When it's free-wheeling, However, the internal gear-wheels still rotate at quite a slow speed (I think it was about 1/3 speed of the wheel). That's pretty slow, so they wouldn't make much noise. When the clutch is engaged they spin very fast and would probably rattle if not under load.

If you find that your clutch is jammed, try giving the tyre agood kick in the forward direction. If you can;t kick, use a heavy hammer like a lump hammer, but make sure that you knock the tyre tangentially - not inwards.

Please let us know how you get on.

 

[shemozzle999]

Are you still using the original cable on the motor?
I assume you had a problem with the hall sensors if you no longer use them.
The wheel didn't spin out of the drop-outs at any stage twisting the motor cable did it?
As the hall cables are thinner they tend to shear first but the thicker phase cables can get damaged as well, while not completely breaking but shearing a few strands of wire therefore limiting the current carrying capability, unbalancing the power phases to the motor.
I know you said you have plenty of power and speed but could you check your 3 power cables to the motor.
Disconnect 1 at a time and see if it makes a difference. If on disconnect you notice no dramatic change then that cable is faultly.

 

[averhamdave]

Are you still using the original cable on the motor?
I assume you had a problem with the hall sensors if you no longer use them.
The wheel didn't spin out of the drop-outs at any stage twisting the motor cable did it?
As the hall cables are thinner they tend to shear first but the thicker phase cables can get damaged as well, while not completely breaking but shearing a few strands of wire therefore limiting the current carrying capability, unbalancing the power phases to the motor.
I know you said you have plenty of power and speed but could you check your 3 power cables to the motor.
Disconnect 1 at a time and see if it makes a difference. If on disconnect you notice no dramatic change then that cable is faultly.

Hi. No its not the cable. I completely rewired the bike when I ditched the old controller and the Hall Sensors (and fitted a rack mounted duct tape battery). With a degree of pedalling it will cruise at 22 to 25 mph on the flat. Its definitely not lacking power. I have two of these (well my wife has one as well really) and there is no noticeable difference between them other than my hub motor is noisy - always has been - and I have this juddering issue.

I'm now fairly sure that D8veh has hit the nail on the head with the clutch problem. We are away in Spain at moment and don't have adequate workshop facilities but when back I'll fully strip and rebuild the hub.

 

[Deleted member 4366]

Hi. No its not the cable. I completely rewired the bike when I ditched the old controller and the Hall Sensors (and fitted a rack mounted duct tape battery). With a degree of pedalling it will cruise at 22 to 25 mph on the flat. Its definitely not lacking power. I have two of these (well my wife has one as well really) and there is no noticeable difference between them other than my hub motor is noisy - always has been - and I have this juddering issue.

I'm now fairly sure that D8veh has hit the nail on the head with the clutch problem. We are away in Spain at moment and don't have adequate workshop facilities but when back I'll fully strip and rebuild the hub.

You don't have to strip it down: just give it a good kick like I said.

 

[Scottyf]

You don't have to strip it down: just give it a good kick like I said.

Sounds like my kind of engineering.
Now stop, Hammer Time!

 

[Deleted member 4366]

Sounds like my kind of engineering.
Now stop, Hammer Time!

The best tool is the one that works

BTW Dave, It's easiest to turn the bike up-side down first.

 

[shemozzle999]

OK, but there is no harm in checking.
It's a quick check, no tools required.
Just disconnect each motor wire in turn.

 

[averhamdave]

Well after fair bit of kicking and work with a lump hammer on the front wheel, it's no better. In fact I'm sure at one stage it got worse!

However having drawn many strange looks from nearby campers I decided to stop, live with it and when back home strip the hub down.

I am now quite sure the problem is the non disengagment of the clutch. Thanks to everyone who posted.

I'll let you know what I find. PS: Anyone got a link to a hub drawing.

Thanks

 

[flecc]

I'll let you know what I find. PS: Anyone got a link to a hub drawing.

Not a drawing, but the link below from my site illustrates the opened eZee Torq 1 motor and gives stripping details:

eZee Torq bafang motor

 

[Deleted member 4366]

You don't need a drawing. I think you have the Bafang with the side-plate retained by screws. They're very easy to disassemble in about 5 mins.
take the wheel off
Undo the 6 screws on the wire side
Tap the whole motor out by banging on the axle end and the cover will pop off with the motor internals attached

This will give you access to the clutch rollers that will just need a bit of help down the ramps. You might find some rust inside. It should look something like this:
http://98.131.176.65/endless-sphere/GEDC0799.JPG

 

[flecc]

I think you have the Bafang with the side-plate retained by screws.

Yes it is, the only SB unit that was used on the eZee Torq 1, my site specialising on that model only as shown on my above link.