New "TSDZ2 Torque Sensor Central Motor"

[Bigwheel]

It is so hard to diagnose problems like yours sysrq from afar. But of the 25 TSDZ2’s that have gone through the shop here none have exhibited any sort of a jerky TA.

The having to stop to reset it is a totally foreign concept. While the assist will stop if it is over run by the parameters of the mode it’s in due to increase in speed meaning it’s time to shift up, when it activates again it is always very fluid.The only thing I can think of is to replace the controller.

As far as the noise you are chasing other than the meshing of the gears, as long as the spindle interface to the bb is super tight, and the blue gear isn’t spitting chunks which makes a grinding noise, I can’t help you there either.

 

[Farther]

Finished my TSDZ2 conversion, but can't get the speedo to work. I think I have the magnet close enough, any ideas?

Also, what's with the gap on the under side of the motor? Just noticed that.

 

[Drum]

Hi Farther,

As far as the speedo is concerned, I would try moving the magnet on the spoke so it crosses near the middle of the sensor rather than the end. Not speaking from experience, just what I would try in your situation.. worth a try.

The motor doesn't look right, as though a wire has been caught between the motor housing and the end cover when it was assembled. I would loosen the motor cover and investigate, or if you bought the motor locally, take it back to the seller and show him.

 

[Marin]

Finished my TSDZ2 conversion, but can't get the speedo to work. I think I have the magnet close enough, any ideas?

Also, what's with the gap on the under side of the motor? Just noticed that.

sensor.jpg

gap2.jpg

On the sensor see the arrow on the top in your photo, line up the magnet with that.

The gap in the case lets moisture out, what it is covering is an empty chamber no wires or controller in there.

 

[casainho]

Good news!! Finally we got pedal cadence

The first part of the video shows torque value and on second part (after I click on LCD power button with the hand that is holding the camera) shows cadence (but the value seems to be half of the cadence).

Next step is to calc the human power and also the Watts Hour of human power. I wish we could implement the disable of some information, so user can decide which information prefer - like cycling only over battery voltage, trip battery discharged total Watts/hour, trip human Watts Hour. I think we need to think on a "design" for all this information.

KT LCD3 showing the following data from TSDZ2 motor:
- battery voltage
- motor power in watts
- motor current
- trip total Watts/hour
- pedal torque value
- pedal cadence
- bicycle wheel speed
- motor assist level
- brakes state
- battery state of charge

[youtube]KsarWeFCn5A[/youtube]

 

[hobbyvac]

For some people with mystery symptoms you might look at the magnetic speed sensor. My assist would cut in and out or just seem reduced at times. I read that the magnet can be too close so I adjusted and it seemed to help but not cure the problem. I finally disconnected it and the assist now works fine. Obviously I no longer have a speedometer but someone also reported that it also removes the speed limiter. I tried a couple more times to re-position the magnet and problems always came back.

This should be one of the first tests you do simply because its so easy,.. just unplug it.

 

[sysrq]

It is so hard to diagnose problems like yours sysrq from afar. But of the 25 TSDZ2’s that have gone through the shop here none have exhibited any sort of a jerky TA.

The having to stop to reset it is a totally foreign concept. While the assist will stop if it is over run by the parameters of the mode it’s in due to increase in speed meaning it’s time to shift up, when it activates again it is always very fluid.The only thing I can think of is to replace the controller.

As far as the noise you are chasing other than the meshing of the gears, as long as the spindle interface to the bb is super tight, and the blue gear isn’t spitting chunks which makes a grinding noise, I can’t help you there either.

''I am not that convinced that the controller is making the best use of the torque sensor with the current software...''
http://www.pedelecs.co.uk/forum/threads/tongsheng-tsdz2-with-torque-sensor-equivalent-to-bosch-or-kalkhoff-impulse-evo.24391/

 

[Waynemarlow]

The symptoms of erratic power on and off and surging has occurred a couple of times on my unit and both times the spoke mounted magnet has turned and become too close to the sensor. Straightened it up and turned off and on the unit and its gone back to normal mode.

Yes the early turning off of the unit when stood talking to your mates, whilst the stragglers catch up, is really irritating.

 

[Stein-65]

Hi, and thank you for letting me in to the forum.

Planning to convert a 3 wheeled coaster brake handicap bike, to electric.
A 250 watt TSDZ2 coaster version seems like a good choice.
When ordering i got feedback from seller ( Ali - Ebird store) that the brake lever with engine cut-off and thumb throttle not would work with the TSDZ2 coaster engine.

Any experience ?
To my understanding the brake lever with engine cut-of is needed to be legal in Europe.

Best regards.


Stein

 

[wpenner]

Hi, and thank you for letting me in to the forum.

Planning to convert a 3 wheeled coaster brake handicap bike, to electric.
A 250 watt TSDZ2 coaster version seems like a good choice.
When ordering i got feedback from seller ( Ali - Ebird store) that the brake lever with engine cut-off and thumb throttle not would work with the TSDZ2 coaster engine.

Any experience ?
To my understanding the brake lever with engine cut-of is needed to be legal in Europe.

Best regards.


Stein

I didn't know the speed sensor affected the drive(i have not finished installing my motor yet). Do you think having more than one magnet might help the TSDZ2 perform smoother? I noticed there is an option in the LCD settings for number of magnets.

 

[Drum]

Hi Stein,
Definitely you shouldn't use a throttle with the coaster brake version of this motor. If the motor is driving the chainring, it is also driving the pedals, so using a throttle would force the pedals to rotate.. not safe. See my post a week or so ago for more details.

With these motors, stopping pedaling stops power delivery quickly. If there is no throttle, there is no real need for brake cut-offs.

Not sure about your legal requirements though. I think lots of these motors have been fitted in Europe.. maybe someone closer to you could comment on the legal requirements.

 

[sysrq]

The symptoms of erratic power on and off and surging has occurred a couple of times on my unit and both times the spoke mounted magnet has turned and become too close to the sensor. Straightened it up and turned off and on the unit and its gone back to normal mode.

If it's a reed switch then the only thing that comes to mind is debounce.
http://www.vespalabs.org/projects/scooter-computer-ii/sensors/reed-switch-debounce
https://forum.arduino.cc/index.php?topic=476457.0
index.php
http://forum.arduino.cc/index.php?topic=14397.0

Yes the early turning off of the unit when stood talking to your mates, whilst the stragglers catch up, is really irritating.

It's mainly while ridding on unassisted mode, when suddenly you need to switch it on an incline or a change in wind direction with a twist grip switch then you realize the display is gone dark and you are forced to stop the bike and start all over again. Or when one rides trough urban area then the best thing to do is to switch it off so you don't draw too much attention due to motor noise, then when you need it on you find that it has been completely turned off itself at the worst possible time.

 

[hurzhurz]

I noticed there is an option in the LCD settings for number of magnets.

The LCD doesn't send the number of magnets to the motor controller.
So the display would probably show the correct speed with multiple magnets, but the motor would think that you drive at e.g. double the speed...

 

[AZUR]

For some people with mystery symptoms you might look at the magnetic speed sensor. My assist would cut in and out or just seem reduced at times. I read that the magnet can be too close so I adjusted and it seemed to help but not cure the problem. I finally disconnected it and the assist now works fine. Obviously I no longer have a speedometer but someone also reported that it also removes the speed limiter. I tried a couple more times to re-position the magnet and problems always came back.

This should be one of the first tests you do simply because its so easy,.. just unplug it.

By coincidence yesterday I went for a ride and I forgot to install the magnet in the wheel. And the strange feelings I did report in the post dated May 28, 2018 11:56 am (intermittent assist) did not appear again. Maybe hobbyvac did found the answer. I can´t wait until I have some time to make some test about this subject again.

 

[Stein-65]

Hi Stein,
Definitely you shouldn't use a throttle with the coaster brake version of this motor. If the motor is driving the chainring, it is also driving the pedals, so using a throttle would force the pedals to rotate.. not safe. See my post a week or so ago for more details.

With these motors, stopping pedaling stops power delivery quickly. If there is no throttle, there is no real need for brake cut-offs.

Not sure about your legal requirements though. I think lots of these motors have been fitted in Europe.. maybe someone closer to you could comment on the legal requirements.

Thanks Drum for sharing you knowledge.
The intention for the throttle was to have "walk assist", not for driving.
Will the cut of brake handle not function at all ?
Could of course use a relay to cut the power to the engine, but since there is a speed limiter in the motor, and cut off switch possibilities on the regular version, it is tempting to try.
Have any other tried brake levers with engine cut of on the coaster version of TSDZ2 ?

/stein

 

[casainho]

Do you know how much torque in Watts you put on the pedals??

KT LCD3 for the first time shows the torque value I put on the pedals, in Watts and so is direct comparable with the motor Watts that is being used.
Pedals torque is calculated using pedals cadence and the force done on the arm cranks (also uses the cranck lenght of TSDZ2 for the calculation).

I verified I am comfortable by doing a continuous value of 150 up to 200 watts of torque on the pedals while the motor is pulling from 500 up to 900 watts from the battery, depending on the assist levels I choose.

KT LCD3 showing the following data from TSDZ2 motor:
- battery voltage
- motor power in watts
- rider pedal power in watts
- motor current
- trip total Watts/hour
- pedal torque value
- pedal cadence
- bicycle wheel speed
- motor assist level
- brakes state
- battery state of charge

On the LCD, at middle right is shown the motor watts and at middle bottom is shown the watts I put on the pedals:

[youtube]ZrrRYpc-wkY[/youtube]

Read more here: https://opensourceebikefirmware.bitbucket.io/kunteng_lcd3

 

[casainho]

Do you know how much torque in Watts you put on the pedals??

KT LCD3 for the first time shows the torque value I put on the pedals, in Watts and so is direct comparable with the motor Watts that is being used.

I verified I am comfortable by doing a continuous value of 150 up to 200 watts of torque on the pedals while the motor is pulling from 500 up to 900 watts from the battery, depending on the assist levels I choose.

Whereas an average cyclist can produce 100 watts, Armstrong can average 400 watts over the final hour of a Tour de France stage.

So seems I am in a better shape than average cyclicist but far from Armstrong

 

[Hillhater]

So seems I am in a better shape than average cyclicist but far from Armstrong

Maybe its because he has(had) a better "diet" than you ! :wink:
PS.. The sprinters and track specialists, can push out up close to 2000W

 

[sysrq]

erian, Why would you put an electric assist on a racing bike? A TDSZ2 system with a large battery could nearly double the weight. The bike would NEVER have racing bike characteristics again. In my view a hybrid type bike, with suspension, would be a better choice. The assist will compensate for the heavier bike weight, plus boost you to your cruising speed. I have the XH-18 small controller. It has a flaw, or design feature, that shuts it off after about 5 minutes of inactivity. Have to have your feet off the pedals to get the torque sensor to initialize properly when you power off/on to reset the controller.

If you are a "spinner" remember that the TDSZ2 boost stops at about 90rpm. You will have to have cluster gearing that will let you reach desired speed at less than 90rpm. May have to change the standard chainring to a 52 tooth, or larger.

I only have about 500 miles on my system - no problems so far.

I don't know why would this be useful design feature. Haven't seen any bicycle speedometer which would require fiddling with buttons to prevent it from shutting itself off while riding and forcing you to stop and get off from the bike.

 

[John and Cecil]

Has anyone using 52v (14s) batteries noticed that over time you can charge them a little bit higher? When we first got our motor 55.8v was the absolute ceiling, but the past couple of days I have been able to run at up to 56.0v. We have about 350 miles on it so far.

Update - I was able to get it to work this morning at 56.1v/56.2v. It definitely seems to be handling a little more power now.

 

[AWD]

Also, what's with the gap on the under side of the motor? Just noticed that.



gap2.jpg

The epdm plate is there to basically block off empty space in the housing below the blue gear and the motor armature gear as mentioned. It never seats fully as designed because the wires leading out of the controller pass behind it and angle up towards a cutout in the rim. They don't lay flat enough to allow for the cover to fully seat against the rim but also as mentioned it can be seen as a way for that area to drain. Although the reason it gets water in there is because of the opening.....

In order to remove the two fasteners you need to remove the crank arm and chain ring to access them. But as previously mentioned in fact there are wires directly behind leading out from the controller it but they are sealed well from the factory. I know this because I have had to remove a good amount of them....so I guess my advice would be to not worry about it.

 

[emr]

Has anyone using 52v (14s) batteries noticed that over time you can charge them a little bit higher? When we first got our motor 55.8v was the absolute ceiling, but the past couple of days I have been able to run at up to 56.0v. We have about 350 miles on it so far.

Update - I was able to get it to work this morning at 56.1v/56.2v. It definitely seems to be handling a little more power now.

Mine has been over 58v when fully charged since new. My little understanding and reading, has suggested that the individual batteries in the package may not all be charged to the same level, and that they become more 'balanced' after each charge. This may depend on the quality of the charger.

 

[Rydon]

On the LCD, at middle right is shown the motor watts and at middle bottom is shown the watts I put on the pedals:

Read more here: https://opensourceebikefirmware.bitbucket.io/kunteng_lcd3

That is the coolest thing I have seen on a low-cost display! Awesome work Casainho!

Some of us are using the TSDZ2 to do conversions for handcycles where arm strength is much less than leg strength and the ratio of human torque to motor watts needs to be adjusted. Sometimes it is difficult for them to get the power they need with the strength they have in their arms. Will this be a parameter that can be adjusted?

 

[casainho]

On the LCD, at middle right is shown the motor watts and at middle bottom is shown the watts I put on the pedals:

Read more here: https://opensourceebikefirmware.bitbucket.io/kunteng_lcd3

That is the coolest thing I have seen on a low-cost display! Awesome work Casainho!

Some of us are using the TSDZ2 to do conversions for handcycles where arm strength is much less than leg strength and the ratio of human torque to motor watts needs to be adjusted. Sometimes it is difficult for them to get the power they need with the strength they have in their arms. Will this be a parameter that can be adjusted?

Thanks for the feedback!

Handcycles? why and how are the final builds? I am just being seeing pictures of regular ebike builds...

I remember to have such an issue with my son of 8 years old, where he can't do the much force with his legs as me. We need to adapt for different persons - for instance, my girlfriend is smaller than me, needs higher sensibility.
For my son, and this was for KT motor controllers firmware and bottom bracket torque sensor, I ended up to use the torque sensor as a PAS for him: he could do force to startup the ebike without the need to rotate pedals but after that, PAS on the torque sensor was also driving the motor so he don't need to make much force on pedals... this way works better for him.

When I was calibrating the power output of pedals, I verified that for every 10kg of force I did the TSDZ2 firmware sends a value of 30 and that is linear over all range. So 3 units for 1kg, 1 unit for 0.33kg. The minimum to detect I would say is 4/5 units, so maybe like 1.5 kg of force to startup and after steps of 0.33kg.

And I don't expect we can change that setting with original firmware. The only option is when we develop our own firmware. I am starting to focus on it and I do not expect we can do more improvements on original.

But maybe the sensor can even have more resolution, at least this is what I feel after being testing it when I started to develop our own firmware and made the torque sensor work:
[youtube]1xiMbCerRKw[/youtube]

 

[echew]

My engine just arrived! Quick shipping from recycles-ebike.com!

Very exciting!

So my immediate first thought is - what are the brake levers for? Do I have to get rid of my existing brakes (I have hydraulics) or are they supplementary?