Attempt: TSDZ2 torque sensor repair, not work :-( any ideas?

[andrea_104kg]

The idea that came to me is to buy two coils and replace them on an old torque sensor.
The first thing to do is to remove the small screws that hold the plastic ring of the pedal sensor.
Then, with a hot air gun, the magnetic ring easily detaches from the iron plate.
Then, take the two rings, one of which needs to be modified to glue it to the old torque sensor.
First, the two steel clamps need to be removed. I tried to force them with pliers on the old ring, but everything broke. Then I tried to grind them with a wheel and... the heat easily removes them in just a few seconds!
Then I glued the new coil to the torque sensor support and connected the wires to the Hall sensor
Unfortunately, it doesn't work :-( the firmware (version 201b with lcd03) gives error "2" which means "torque sensor error". At this point, I think that the 2 coils, even if apparently identical, are not the same, maybe they have different windings and the controller feels that there is something wrong... Any ideas?

 

[Elinx]

@andrea_104kg
You did it again not choose the easy solution, but without any knowledge what to expect you experiment with parts.
The hall sensor with the magnets are the torquesensor, the coilpair acting as a transformer for wireless data transfer.
You didn't say if these coils are original for Tsdz2 or from another source.
I ask because I can imagine that the coilspecs. must be same as before.

@stancecoke has done some investigation of the Tsdz2 torquesensor on the German pedelecforum and did give some values.
Maybe you can use this information for measuring some things.

Youtube

 

[andrea_104kg]

Yes, the two coils are the original coils of the TSDZ2. They are two internal coils of the torsion meter. The premise of my attempt is that the 2 coils of the torque sensor were identical. Therefore, instead of replacing the entire sensor, I thought it would be sufficient to replace the 2 coils. I know that coils are just a means to transmit the data of the torque sensor on the tube. They are only used because in a rotating body, it cannot be directly connected with a wire that would twist with the rotational motion.
I don't understand why you say I don't know what to expect. Obviously using original coils I expect the torquemeter to work as it originally did but sadly it doesn't :-(

 

[endlessolli]

I had once a faulty torque sensor and it turned out that it was the hall sensor itself was broken. Details of the repair I posted on github (scroll all the way to the bottom:

How to calibrate the torque sensor

Contribute to OpenSourceEBike/TSDZ2_wiki development by creating an account on GitHub.

github.com

Hope ths helps and good luck!

 

[Elinx]

... I thought it would be sufficient to replace the 2 coils......
I don't understand why you say I don't know what to expect. ...

If these coils are original, not broken and mounted the right way it is supposed to work again if the hallsensor does work too as @endlessolli pointed.

Copy/past from Github Wiki

How to repair a torque sensor​

• If your Torque Sensor is not working, it could be due to a faulty Hall Sensor. Here is how I replaced it:
• Peel off the black stuff to investigate (in my case the Hall Sensor pins / solder pads were torn off.)
  
  • Be extra careful to not harm any of the tiny SMD parts underneath.
• The Hall Sensor is a std "49E / OH49E / SS49E" Hall sensor as found on ebay.
• Just solder it back on - for details, see linked picture below:

With the remark "don't know what to expect", I refer to your Tsdz2 experiments (also on Jobike) which are not always the easiest, but with an original approach, but the result of which is unknown. Such as welding the clutch bearing, sawing open the housing for cooling, etc.

 

[andrea_104kg]

Unfortunately, I had 2 broken torsimeters. I tried with 2 hall sensor boards, I doubt that both chips are broken (possible but unlikely), but the result is always the same, error 2 on the display. I did many tests:
1- no sensors attached to the control unit, error 2.
2- attached only the inner ring to the control unit, no error but obviously one coil is not enough.
3- complete assembly of the torsimeter, error 2.
4- replacement of the hall sensor board and reassembly of the torsimeter, again error 2.
The coils are new and original, so I don't know what to think..."

 

[endlessolli]

... the hall sensor is imho the weak spot: As it sits (depending on hardwarecalibration) very close to the magnets or even touches it and the magnets are moving relative to the hall sensor, it is easy to shear off the sensor or the soldering pads it is soldered to during std. usage
(And you would not be able to notice, as its all happening under the black goo....)
I would give is a shot and try a replacement of the hall sensor on one of your sensor-boards. These sensors cost almost nothing, so its worth a try.
(but one needs good eyes and a steady hand when soldering....)
I can not think of anything else other than this or the proper function and distance of the coils to another or a wiring / connection issue....

 

[Sturmey]

Yes, the two coils are the original coils of the TSDZ2. They are two internal coils of the torsion meter. The premise of my attempt is that the 2 coils of the torque sensor were identical. Therefore, instead of replacing the entire sensor, I thought it would be sufficient to replace the 2 coils. I know that coils are just a means to transmit the data of the torque sensor on the tube. They are only used because in a rotating body, it cannot be directly connected with a wire that would twist with the rotational motion.
I don't understand why you say I don't know what to expect. Obviously using original coils I expect the torquemeter to work as it originally did but sadly it doesn't :-(

As well as transmitting data, the coils also transmit a small amount of power for the hall sensor in the spindle (like a transformer). Is it possible that the coils turns/turns ratio is different from the original setup as this could possibly result in wrong voltage or insufficient power going to hall?

 

[Elinx]

... Is it possible that the coils turns/turns ratio is different from the original setup as this could possibly result in wrong voltage or insufficient power going to hall?

Also the right distance between the coils has infuence.

 

[andrea_104kg]

At this point, I think that the 2 coils, even if apparently identical, are not the same, maybe they have different windings, also may be that the hall sensor it is broken, i tried to change one of them but it's too small for my capacity of soldering :-(

 

[andrea_104kg]

Also the right distance between the coils has infuence.

the distance are the same, the coil are original... Maybe someone else more competent can get it to work.

 

[endlessolli]

At this point, I think that the 2 coils, even if apparently identical, are not the same, maybe they have different windings, also may be that the hall sensor it is broken, i tried to change one of them but it's too small for my capacity of soldering :-(

Hi again;
if you want, you can send the 2 little sensor boards to me and I try to do the soldering for you (no cost). These are tiny, so they could be send in a std envelope. I am in germany, you in italy, I assume, so cost is low. It would be a shame, if this would go in the trash only because of this. (I also can not belief that the 2 coils are different. They look the same also in terms of wire gauge / winding count when I looked at them last time) Let me know, we can exchange addresses etc via pm.

 

[DontCallMeShirley]

I had once a faulty torque sensor and it turned out that it was the hall sensor itself was broken. Details of the repair I posted on github (scroll all the way to the bottom:

How to calibrate the torque sensor

Contribute to OpenSourceEBike/TSDZ2_wiki development by creating an account on GitHub.

github.com

Hope ths helps and good luck!

Hey there! I was researching the forums concerning a faulty torque sensor and stumbled upon this suggestion.

What were your symptoms of the faulty torque sensor? I've accessed my hidden TE-settings and seen that the values are static at around 25 or 26 no matter the torque being applied. Do you think this could be the hall sensor, or what that return no torque values at all?

I'm contemplating buying a new torque sensor but would rather avoid the extra cost if its just the hall sensor

 

[endlessolli]

Hey there! I was researching the forums concerning a faulty torque sensor and stumbled upon this suggestion.

What were your symptoms of the faulty torque sensor? I've accessed my hidden TE-settings and seen that the values are static at around 25 or 26 no matter the torque being applied. Do you think this could be the hall sensor, or what that return no torque values at all?

I'm contemplating buying a new torque sensor but would rather avoid the extra cost if its just the hall sensor

The Symptoms were non-changing torque values. However, to my best knowledge, the syptoms are similar to other hardware-failures of the sensor - i.e. Red / Black cable ripped / disconnected.
Bevore buying anything new, I would dismantle the motor to find the root cause.

 

[DontCallMeShirley]

The Symptoms were non-changing torque values. However, to my best knowledge, the syptoms are similar to other hardware-failures of the sensor - i.e. Red / Black cable ripped / disconnected.
Bevore buying anything new, I would dismantle the motor to find the root cause.

I see. I've opened the motor and checked for continuity with my multimeter without finding any errors. Torque sensor looks fine visually. I don't know how else I can troubleshoot the torque sensor.

Maybe I'll just order a hall sensor and try the replacement. Worst case I'll just buy a new torque sensor.
Thanks!

 

[endlessolli]

I see. I've opened the motor and checked for continuity with my multimeter without finding any errors. Torque sensor looks fine visually. I don't know how else I can troubleshoot the torque sensor.

Maybe I'll just order a hall sensor and try the replacement. Worst case I'll just buy a new torque sensor.
Thanks!

Sometimes the wires to the sensor are not soldered properly to the pcb, i.e. are disconnected but just held in place only by the black goo.
(When removing the black stuff be extra careful to not destroy / rip off the tiny smd parts from the sensor…)

 

[richj8990]

One day a stat geek needs to tally up all of the Bafang BBSxx complaints vs Tongsheng and then compare that to the total sales and see how many percentages of complaints there are per sale. If it's 0.1, 0.2, 0.3 complaints per sale, and then see how lopsided the complaint ratio is toward Tongsheng. I know for sure it's more Tongsheng complaints per sale, but do not know what percentage more it is, if it's 30% more, or 100% more, not sure.

 

[DontCallMeShirley]

One day a stat geek needs to tally up all of the Bafang BBSxx complaints vs Tongsheng and then compare that to the total sales and see how many percentages of complaints there are per sale. If it's 0.1, 0.2, 0.3 complaints per sale, and then see how lopsided the complaint ratio is toward Tongsheng. I know for sure it's more Tongsheng complaints per sale, but do not know what percentage more it is, if it's 30% more, or 100% more, not sure.

Haha yeah my old BBS01 was much easier to deal with than the Tongsheng. But I just love the torque sensor on this thing (when it works)...

 

[tomjasz]

but do not know

Repeat after me. “I will not diss tsdz2”

 

[docw009]

One day a stat geek needs to tally up all of the Bafang BBSxx complaints vs Tongsheng and then compare that to the total sales and see how many percentages of complaints there are per sale.

Probably 10X higher for Tong Sheng. Maybe half the TSDZ2 complaints are pilot error in setup/install or maybe about its lower power, but the other half are about it breaking,.

This is an interesting thread though, Who would have thought the T/S is wireless.

 

[Slowchris11]

If the driver has more than a certain wheigt and helps the bike move by putting force on the pedal, the Tongsheng may breaks during first use. If you do not work the pedal that hard, because you are feather wheight or lazy, it takes some time to break the single bearing and give the axle to much play. In such cases you usually destroy the coils for the sensor.

If you are a good mechanic, you will see there is at last one bearing missing on the axle.
100% construction fault from one who has not learned how to construct something the right way and does not understand static. That is why we have technical university in the West. In China you have to be a good communist to lead a construction team and be the deciding technician...

The magnet ring has very weak magnetism. If you heat it up for removeal, it may weaken even more and induction get to weak to power the rotating part. When the coils are not mechanicaly damaged, just measure them, don't remove. If they are not shorted or open loop, consider changing the sensor PCB.

 

[endlessolli]

That is why we have technical university in the West. In China you have to be a good communist to lead a construction team and be the deciding technician...
.

This arrogance against „the chinese“ /„China“ really only shows your ignorance. China was the cheap workbench for the West - a long time ago. Meanwhile China is agead of everybody else in i.e. battery technology and many other advanced technical things. See i.e. also the microcontroller ESP32. Such a powerful and versatile thing - for a couple of bucks: designed and made in china.
And the ignorance like yours will lead to a future, where we “the West” are only needed as the paying consumer..,, and when we are out of money, we are not needed at all….
The shortcomings of the TSDZ2 stem from the fact that it “good enough” for the target marked. And due to the fact that it is dirt cheap.

 

[amberwolf]

If ya'll want to have a political discussion, please move it over here:

Off Topic Discussions

endless-sphere.com