New "Lingbei MM28.250 torque sensor mid drive"

rick.schoenmaker@gmail.com said:
Chito said:
Hi, I have a mm28.500, I have noticed that it only has 85 rpm of pedaling assistance. this is normal?

Quite normal for electric assistance as Max torque is at 1rpm? Maybe this something cut by the torquesensor?
I measured it directly with the throttle, and without the chain in place so that it didn't force
 
mm28 really uses a torque sensor, use a spring between the axle and the transmission to the pinion.It has two magnets and measures the offset between the two to record the displacement of the spring. The transmission may have a small displacement with respect to the axis. One magnet is attached to the shaft and the other magnet is fixed to the transmission.
 
https://www.instagram.com/p/B3zKYrABY-o/
https://www.instagram.com/p/B3zKWpRBNcK/
https://www.instagram.com/p/B3zKjQEBqXb/
https://www.instagram.com/p/B3zKlnRBd7V/
https://www.instagram.com/p/B3zKnfpBvvs/
https://www.instagram.com/p/B3zKq2QhAqm/
 
maitilupas said:
The downside is that is still annoyed with its PAS functions programing, specially with a 500w battery, ive ordered a cable to see if i can change the amps on the pas levels. If i can do this im happy with it, as for the torque sensor maybe in The future ill add something different if possible..

Can you tell us more about this cable and the programming process ? The existence of this cable was mentionned in the thread earlier but nothing more.

The only thing holding me back for this motor is the lack of assist settings (adjust sensitivity to cadence and torque sensors).
 
Chito said:
mm28 really uses a torque sensor, use a spring between the axle and the transmission to the pinion.

Thank you for the photos. But please explain better. I understand the principle, but I don't see a suitable spring in your photos....

regards
stancecoke

Lingbei_6.png
 
stancecoke said:
Chito said:
mm28 really uses a torque sensor, use a spring between the axle and the transmission to the pinion.

Thank you for the photos. But please explain better. I understand the principle, but I don't see a suitable spring in your photos....

regards
stancecoke


Hello stancecoke,
the analysis you did in the images are correct.
Here I put more images of the spring, I have repaired it because it was broken:
https://www.instagram.com/p/B39eVk8hKgt/
https://www.instagram.com/p/B39cwD-BjDN/
https://www.instagram.com/p/B39cxx6hH3t/
https://www.instagram.com/p/B39c0JPhXBf/
I had to drill the shaft to remove the original spring, since it was very tight:
https://www.instagram.com/p/B39c2QXB2LZ/
you have to be very careful when removing the magnet because it can break
https://www.instagram.com/p/B39eSNMBlzV/
fortunately it was easily repaired
https://www.instagram.com/p/B39c5pahErM/

The spring works with little pressure on the pedal, it seems more like a switch than a linear sensor. To start pedaling assistance, it is necessary to put some pressure on the pedals at the beginning

the problem that Metaller (forum user) had, is solved with a fine adjustment in the magnet position so that the pulses are synchronized
 
I saw in a part of the forum that the motor controller uses a stm32 micro controller. Is it correct? If so, it could easily be programmed in arduino IDE with a st link v2. :D
Does anyone know where to find any programming code of any controller? :wink:
having a code, it can be edited or new functions can be programmed and then loaded in mm28 :flame:
The code in programming language would be more useful, for example in C or Java, not the code in hexadecimal
 
Chito said:
it seems more like a switch than a linear sensor

Hm, with some beam theory and some assumptions of dimensions from your new photos, the sensor could work quite linear...
You'll get a displacement of about 1mm if you step on one pedal with 80kg. Of course your spring made from two stripes is much weaker, than just one solid spring.

Lingbei_7.png

Chito said:
Does anyone know where to find any programming code of any controller? :wink:

See my open source project for Lishui Controllers, they use STM32 as well (perhaps it's even a Lishui-Controller, can you find a print LSW.... on the PCB?)
https://github.com/stancecoke/LishuiFOC/wiki

regards
stancecoke
 
stancecoke said:
Chito said:
it seems more like a switch than a linear sensor

Hm, with some beam theory and some assumptions of dimensions from your new photos, the sensor could work quite linear...
You'll get a displacement of about 1mm if you step on one pedal with 80kg. Of course your spring made from two stripes is much weaker, than just one solid spring.

Lingbei_7.png


Chito said:
Does anyone know where to find any programming code of any controller? :wink:

See my open source project for Lishui Controllers, they use STM32 as well (perhaps it's even a Lishui-Controller, can you find a print LSW.... on the PCB?)
https://github.com/stancecoke/LishuiFOC/wiki

regards
stancecoke

Congratulations stancecoke. Very good calculations very good modeling and very good approximation.
Adjusting the parameters a bit: a = 3; b = 10; therefore I = 22.5.
we assume f = 0.9 approximately, but L is more difficult to give an approximate value because measuring directly can give a value between 3 and 12 mm, https://www.instagram.com/p/B393_kshQUm/
To this we must add a component of transverse flexion of the spring so I decided to take experimental measurements and model a little
for the spring I used a = 1.5 mm, therefore I = 2.8 but since they are two springs we have I = 5.6. I measured the force that would be exerted on the pedals so that f is approximately 0.9 and shows w = approximately 6.5 kg
https://www.instagram.com/p/B3920Z4hOSJ/
With these results, adjust in the equation a value of approximately L = 14.2 mm so that the results are consistent with the experimentally measured force. So recalculating for the original spring, we have: w = F / (170/10) = 3fEI / L ^ 3 = (3/17) * (0.9 * 210000 * 22.5) / (14.2) ^ 3 = 262N = 26.2 kg approximately


I haven't seen the motor controller board yet
 
Is it possible to replace the STM32F031C6T6 of the lingbei controller with an STM32F103C8T6?
 
rick.schoenmaker@gmail.com said:
Aquakitty said:
rick.schoenmaker@gmail.com said:
rick.schoenmaker@gmail.com said:
I've been in contact with OKFeet (supplier) and they send me an new Hall Sensor component to bolt on (no soldering). Very helpfull with video instruction for assembly. I'm wondering if more people have trouble with the Hall Sensor or other sensors. Maybe my issues are due to the Dutch weather.

After receiving the Hall-sensor, I opened the motor and delivered a surpise I was afraid of....

At least 50ml of rust water came out. The motor was covered in rust and also elements of the controller. So not only a defect on the hall sensor... After putting it back together I now also get a “Current Error”... The thing is dead I presume

Sadly OKfeet said this is not part of warranty/after sales care. It surely is a manufacturer error on the seals where moisture was able to get in via the ingoing/outgoing wires.

Don’t know what to do now, giving it another chance or seeking an alternative...

You should never assume any of these Chinese mid-drives are waterproof. You have to waterproof them
yourself.

After direct contact with the Lingbei factory I received today a new motor (fairly quick as it was shipped last friday from China). $100 for the motor and $100 postal it fairly cheap as well (OKfeet asked for $236 without shipment).

An extra connection on the main wire from brakes/display/gashandle to the motor was included (maybe for flashing the firmware?).

Now the biggest challenge is to make it waterproof. Still figuring out the best way (silicon kit at the wiring and plastic cover over the motor at least).

So it's it possible to order a motor directly from them than at that price?
I'm liking the sound of this motor and thinking of getting it for my old mountain bike, not sure if I would be best off with 36v battery to make the sensor work better?
Best place to get this motor right now with shipping and import tax to the UK anyone?
Thanks.
 
Chito said:
Is it possible to replace the STM32F031C6T6 of the lingbei controller with an STM32F103C8T6?

Why a "STM32F031C6T6"? In the beginning of this thread a "STM32F101C6T6" is mentioned?!

https://endless-sphere.com/forums/viewtopic.php?f=28&t=94220#p1382795

But all three share the same pinout. The STM32F031C6T6 has only one ADC, so my FOC code can't work with it.

STM32 comparison.PNG

regards
stancecoke
 
rick.schoenmaker@gmail.com said:
Now the biggest challenge is to make it waterproof. Still figuring out the best way (silicon kit at the wiring and plastic cover over the motor at least).
use liquid gasket or rtv at seams and ACF50 on any connections and electronics.
 
Chito said:
The spring works with little pressure on the pedal, it seems more like a switch than a linear sensor. To start pedaling assistance, it is necessary to put some pressure on the pedals at the beginning

the problem that Metaller (forum user) had, is solved with a fine adjustment in the magnet position so that the pulses are synchronized

Thanks to you I finally understood how the torque sensor works in this engine, it is left to understand how to do it without multiple assembly-disassembly of the engine... :) My motor seems to have started to blow again... :cry: apparently the spring has weakened or flipped... I will try to disassemble the old shaft.
 
My 48V 500W Lingbei is acting really funny and some of the symptoms might look similar to what Metaller described earlier.

The torque sensor seems to be doing opposite to what it should. I noticed that the harder I press the pedals, the less like power will be delivered. Only when pedaling very gently the motor will activate and deliver full motor assist. So great on flat, but pretty useless when trying to climb some hills or overtake in the traffic. And my feeling is that overall power delivery when pedaling seems to be pulsating. However, the throttle works fine.

I was wondering if anyone has a clue what the problem might be?

I have been in contact with the seller, and after going back and forth they offered to provide me with the cable to program the motor “with an update”. I am bit puzzled what that update would actually do and worried it might be some sort of a workaround disabling input from a faulty torque sensor and relaying on PAS only.

If the torque sensors in this motor is so fragile, it would mean that it’s not really big advantage over TSDZ2 (which was I considering in the first place).
 
fifthbro said:
My 48V 500W Lingbei is acting really funny and some of the symptoms might look similar to what Metaller described earlier.

The torque sensor seems to be doing opposite to what it should. I noticed that the harder I press the pedals, the less like power will be delivered. Only when pedaling very gently the motor will activate and deliver full motor assist. So great on flat, but pretty useless when trying to climb some hills or overtake in the traffic. And my feeling is that overall power delivery when pedaling seems to be pulsating. However, the throttle works fine.

I was wondering if anyone has a clue what the problem might be?

I have been in contact with the seller, and after going back and forth they offered to provide me with the cable to program the motor “with an update”. I am bit puzzled what that update would actually do and worried it might be some sort of a workaround disabling input from a faulty torque sensor and relaying on PAS only.

If the torque sensors in this motor is so fragile, it would mean that it’s not really big advantage over TSDZ2 (which was I considering in the first place).

What is happening to you is due to an incorrect offset between the sensor magnets. When you feel the intermittency is when you lower the left pedal because the natural torque of the bottom bracket axis helps to further offset the magnet that is more central, however when you lower the right pedal, it does not contribute to offset the magnet due to the torque of the axis because chain transmission is more direct. If you also feel loss of power when you lower your right foot it is because the angular offset between the magnets is very large.
The solution to this problem is to rotate the magnet (the one inside) a little (VERY LITTLE) counterclockwise if you look at it from the side of the chain. Or rotate the magnet that is outside, in favor of the hands of the clock, but it is more difficult.
 
Chito said:
https://youtu.be/Y9HFFLLJLaw
This video explains better


thanks allot , great video
i broke the magnets while taking them out stupid me , do you think if i glue them they can work out?...
 
maitilupas said:
Chito said:
https://youtu.be/Y9HFFLLJLaw
This video explains better


thanks allot , great video
i broke the magnets while taking them out stupid me , do you think if i glue them they can work out?...

my magnet also broke, I joined it with cyanoacrylate and it was perfect
https://www.instagram.com/p/B39eSNMBlzV/
 
Chito said:
my magnet also broke, I joined it with cyanoacrylate and it was perfect
https://www.instagram.com/p/B39eSNMBlzV/

great to know ! thanks allot for the pic, more relief but meanwhile found another problem. i also found out my spring is broken inside and cannot take it out
 
maitilupas said:
Chito said:
my magnet also broke, I joined it with cyanoacrylate and it was perfect
https://www.instagram.com/p/B39eSNMBlzV/

great to know ! thanks allot for the pic, more relief but meanwhile found another problem. i also found out my spring is broken inside and cannot take it out

I had to drill the shaft to remove the original spring, since it was very tight:
https://www.instagram.com/p/B39c2QXB2LZ/
I drilled it with a dremel, a tungsten drill and a tungsten strawberry. Then I hit the spring from behind
 
Back
Top