New "TSDZ2 Torque Sensor Central Motor"

[casainho]

I am starting looking at the TSDZ2 motor controller V2 and I see the following differences:
- microcontroller ARM Cortex M0, 32 bits at 64MHz
- that three big R003 resistors are for sure to measure each motor phase current while on the controller V1 there was only one big resistor to measure the battery current

The increasing of processing power and the capability to measure the 3 phase currents means the system can implement FOC. Anyway, we should be able to port the firmware of controller V1 to controller V2 and ignore at start that 3 phase currents, as the sum of the phase currents is equal to the total phase currents that we calculate already on the controller V1 firmware, as total phase current = battery current / PWM duty_cycle:



I also updated the main page of TSDZ2 EBike wireless controller with information for the Garmin Edge / cycling GPS displays: https://opensourceebike.github.io/

The page has now the following topics:
- EBike wireless controller
- Mobile app
- Fully wireless and wired remotes
- Garmin Edge / cycling GPS displays

 

[mctubster]

I remember a while back seeing a post about the crank arms that come with the tsdz2 and how they’re not symmetrical: one side sticks out more than the other from the frame.

I was fine with it until recently where I started going for much longer rides (50–70km). After these longer rides, I’m getting pain in my left knee. I suspect this is due to the uneven crank arms.

What was the solution? I think I remember they could be swapped with another set of cranks that were symmetrical but I don’t remember which cranks it was. Can anyone share the reference for these replacement cranks?

From an earlier post I made

I have been obsessing over Q Factor for some time now. I bought some replacement cranks and measured against the bundled TSDZ2 cranks. Here are the results for my bike with bottom bracket width 70mm.

Original TSDZ2 crank arms (edited to make clearer)
206mm (offset 26mm to the right)

Shimano crank arm drive side (Shimano STePS FC-E6000) (kept original TSDZ2 crank arm non drive side to decrease the offset)
199mm (offset 19mm to the right)

Bafang straight crank arm drive side (AC08-2) (kept original TSDZ2 crank arm non drive side to decrease the offset)
193mm (offset 13mm to the right)

 

[LeftCoastNurd]

yea, I hated the Q of the stock TSDZ2 setup, the right pedal was just stuck out way too far. with the BB right crank that Eco-ebike sold me, its close enough that my foot consistently lands on the pedal and I'm not annoyed while pedaling. I'm just running the stock 42t TSDZ2 chainring with its ring guard, it doesn't snag my pants leg

fyi, I tried a 52-42t front setup, never bothered to get a derailleur working with it, was just manually shifting, tried it for a few days. the 52T was just too tall, I could go fast as heck with 52:13 gearing but the power demands heated it up too fast, so I dropped back to the original 42, and swapped my cassette from a 13-26 to a 11-32 (8 speed), thats a tall enough gear, and a low enough gear. when this 700c hybrid bike was pure-pedal powered, I was running a 48-38-28T front with that 13-26 and when I was in shape, I could go just about anywhere, including climbing the mile long grade I live up that has 8%+ stretches..

 

[Retrorockit]

There is a casette I like from Sunrace. The 11-40T 8 speed. Good range, and the wide gear splits don't seem to bother the wide power band of electric motors. Very inexpensive also. When I swap chains I just swap the casette also. I did need a longer derailer hanger.

For the rapid overheating issue. Does your firmware have the crank rotation delay set to zero. I always felt that this could lead to the motor running in a stalled condition at stop lights with your foot resting on the pedal. If you don't have brake switches this could be possible.

 

[beemac]

Is this a new or old controller? ....

This is the old controller which can be prgrammed with OSF

Here you see how componenst are placed on the new 2020 controller:
NewTsdz2controller.jpg

This is the "old" one:
oldcontroller.jpg

Soon I hope to start documenting the TSDZ2 motor controller V2 and start developing the early firmware for it. Meanwhile where is the repository that I will be using: https://github.com/OpenSourceEBike/TSDZ2_motor_controller_v2

Interesting... where did people get the new controller? Are they available to buy as a spare separate to the motor anywhere yet?

 

[Elinx]

.....
... where did people get the new controller? ....

I think the only store that sells the new 2020 controller is the German EnerProf

 

[casainho]

Just to let you guys know that I just started the very initially firmware development for TSDZ2 motor controller v2 - this version has a way a more capable hardware to implement FOC, which will mean more efficiency motor control and less noise - also the microcontroller manufacturer provides a ready code for FOC, so should be very fast to get it working!!

Here the repository: https://github.com/OpenSourceEBike/TSDZ2_motor_controller_v2

And building the very initial code:

 

[LeftCoastNurd]

Just to let you guys know that I just started the very initially firmware development for TSDZ2 motor controller v2 - this version has a way a more capable hardware to implement FOC, which will mean more efficiency motor control and less noise - also the microcontroller manufacturer provides a ready code for FOC, so should be very fast to get it working!!
...

awesome. um, whats FOC ?

edit:
ah, read the app note on it at the git site.

Field Oriented Control (FOC) motor control is a method to generate three phase sinusoidal signals which can
easily be controlled in frequency and amplitude in order to minimize the current which means to maximize the
efficiency. The basic idea is to transform three phase signals into two rotor-fix signals and vice versa.

interesting. Very interesting.... reminds me of the articles I used to read in Motion magazine on doing motor control using DSP techniques.

 

[LeftCoastNurd]

SO... can an existing TSDZ2 be upgraded with a new CPU board, or do you need a whole new TSDZ2 ?

 

[casainho]

Just to let you guys know that I just started the very initially firmware development for TSDZ2 motor controller v2 - this version has a way a more capable hardware to implement FOC, which will mean more efficiency motor control and less noise - also the microcontroller manufacturer provides a ready code for FOC, so should be very fast to get it working!!

I forgot to mention that if more developers help, I am pretty sure we can have the very first working version of the firmware in 1 month or less!

SO... can an existing TSDZ2 be upgraded with a new CPU board, or do you need a whole new TSDZ2 ?

You need to just buy this motor controller V2 for the same 35€, flashing the new firmware and install it inside your TSDZ2 as the V1 controller. The programmer to flash the firmware is a different one from STLinkV2, but still is a very cheap one of 4€ bought on Ebay.

 

[LeftCoastNurd]

I forgot to mention that if more developers help, I am pretty sure we can have the very first working version of the firmware in 1 month or less!

You need to just buy this motor controller V2 for the same 35€, flashing the new firmware and install it inside your TSDZ2 as the V1 controller. The programmer to flash the firmware is a different one from STLinkV2, but still is a very cheap one of 4€ bought on Ebay.

sigh, my programming/coding days are in my past. in the 80s, I would have been hot to trot on this. but I achieved a stage of burnout where I no longer have the concentration or patience to follow through on even simple programming projects, and now I'm a retired old fahrt.. heck I'm having trouble finding the patience to sort through a pile of stupid website stuff I should be doing for friends and fun.

I will eagerly await something usable, and consider swapping out my motor controller, however! I still need to take my motor apart and pack the housing with some thermal goo I have a jar of, supposed to massively improve the heat conduction.

 

[MatejO]

Hi.

The TSDZ2 with torque sensor looks OK.
But... What can I expect if I mount it on enduro MTB ? Will it take the hill climbing?
I am down to put some effort into the climb. I am just looking for some assistance.

Thank you for your feedback.

 

[LeftCoastNurd]

Hi.

The TSDZ2 with torque sensor looks OK.
But... What can I expect if I mount it on enduro MTB ? Will it take the hill climbing?
I am down to put some effort into the climb. I am just looking for some assistance.

Thank you for your feedback.

i only see thermal issues when I use 3X or more my human power, so the motor is outputting 500-700 watts when I'm outputting like 100-150 watts in a relatively high gear. keep that up for a couple miles of steep hill, and the motor gets too hot. if I select a gear and motor power multiplier such that the motor is only outputting under 300W while I'm doing 100-150 watts, it won't heat up nearly as fast, and thats still a net of 3X your base power.

having seen the temps I've seen, I would not want to run a tsdz2 '750w' 48/52V system without a temp sensor and the open source firmware configured to show the temps on the display panel. there's things you can do to improve the motors cooling, as it comes, its not very good at it. packing the motor housing with a suitable thermal compound is one solution, others involve holes in the motor housing (imho, bad idea on a dirty-bike), or adding heatsink fins (hard to do and look like anything other than a kludge.

 

[HrKlev]

Hi.

The TSDZ2 with torque sensor looks OK.
But... What can I expect if I mount it on enduro MTB ? Will it take the hill climbing?
I am down to put some effort into the climb. I am just looking for some assistance.

Thank you for your feedback.

If you are doing enduro on your enduro bike, its not the way up you should be worried about, its the way down. The crank is not only underdimensioned and made put of butter, it also has a recess for a lock ring right at the critical spot. I would avoid jumps and drops with that motor, it might hurt.

 

[Waynemarlow]

If you are doing enduro on your enduro bike, its not the way up you should be worried about, its the way down. The crank is not only underdimensioned and made put of butter, it also has a recess for a lock ring right at the critical spot. I would avoid jumps and drops with that motor, it might hurt.

Mmmmm what is it with forums and threads where one bad experience written in such terms as above can influence how it’s presumed all operators will experience the same thing. Sorry we do hard XCountry riding that probably to most would be Enduro style riding but just without jumps, and yet to bend or break a crank.

Just the amount of miles we have done and on the number of engines within our group, all without serious mechanical problems, would indicate that if someone is regularly breaking or bending cranks, then it’s that individuals riding style which is taking his engine outside of its design, rather than all engines of its type. Sure if you want to do black or diamond downhills with large jumps then you only have to take one look at the motor and say it’s not for this purpose, so why fit it and then complain that it’s made of butter.

Sometimes we have to make personal engineering design decisions around our own useage requirements. If you want to fit this engine, use it way outside what it was designed for that of a commuter bolt on and then accuse it of being made of butter because you didn’t bother to do basic engineering design assessment, then please don’t put off other users by your own failure to do so.

 

[casainho]

I found a kart that uses TSDZ2 and seems that are people adapting this karts to do real work!!

Berg toys is the manufacturer of this karts: https://www.youtube.com/watch?v=4lXI-ksBXHk

 

[ebbsocalMTB]

Thanks for the posts about my situation folks. Regarding a few responses about my technical abilities or knowledge... I guess it's a good thing that I don't feel like I need to explain myself, and those reading, also don't really care about my explanation. We're on even ground of not giving shits about each other, which is a nice place to be :lol:

The truth is, I'm not an electrical engineer (I hit triangles with calculators for a living), and I've got enough on my plate between racing spec miata, desert moto racing and mountain bike racing... I did the sensible thing and hit the easy button and ordered from eco-ebike on this one. The cost was marginally more then I would have spent procuring all of the different parts from different places and it showed up pre-installed. It's not that I'm not capable of doing the work... It's that this was a logical way to get the right items in a timely manor.

So... Regarding the testing I had done... I did 3 commutes to work to hash out this issue. The first ride I did was at assist level 12 and I largely let the motor do it's thing and didn't pay that much attention to the wattage output of the motor. This was my first work commute and the motor went into thermal throttling at mile 5.1 - 5.2 and hit 180 degrees.

You can see that point by the line/mark in the graphs where my heart rate climbs quickly. This is when I shut the bike off and just pedaled pretty lightly.

The next work commute I did on the same route, I payed much closer attention to the heat/output of the motor. I kept it in assist level 7, and made sure to keep the motor pretty much at the 250-300w mark, except for pulling away from the 2 stoplights I hit. This ride was extremely perplexing because it started to reach thermal max at mile 4, and by mile 5.1, it was at 180 degrees. Same exact ride, same ambient temperatures.

So far, the following potential causes for this have been floated out there:
1) My inability to own and operate this motor correctly. While possible, this seems unlikely to cause this motor to overheat at assist level 6 after 15 minutes
2) Some issue with the current firmware, potentially related to field weakening. Again, seems unlikely, but possible. :|
3) An issue with my particular unit from a mechanical perspective :?
4) An issue with a sensor on my particular unit, temp sensor or otherwise :x
5) A mechanical issue with the bicycle, at this point I'd say this one is dead as I've gone through everything

So... It seems the next steps are the following
1) Reflash to stock firmware. I'm not 100% sure what this will do other then validate that the motor could output power at assist level 6 without overheating. The problem is, I won't know if it's overheating and there won't be a temp sensor reading.
2) Reflash with a previous version of the OSF.
3) Return the unit to Eco-Ebike for them to QA/QC or replace
4) Tear the unit down and check it myself and add any thermal modifications needed.

I still have a few questions:
1) David from eco-ebike has suggested that the field weakening and a high cadence can generate heat faster with the OSF. Does anyone here see any logic to that and are there specific parameters I should be looking out for when riding to avoid this? IE, keep cadance below 90, make sure motor RPM is in window XXX-XXX etc?
2) Is field weakening alone, what is allowing the motor to run higher RPM then the stock firmware, or are there other contributors. I've read about Field Weakening, and understand it generally when it would come to an archaic DC motor... but I'm afraid I'm a bit over my head in 3 phase brushless motors... although I'm drinking from the fire hose of knowledge. My assumption at this point is that field weakening is the electric motor equivalent of retarding and advancing ignition timing on an IC engine, and that there is likely a point of no return in here that may be contributing to my problem.
3) How much resistance should I feel when pedaling the bike in the work stand with the system off? I don't really have a great gauge for what this should feel like, but my only comparison is an x01 eagle drivetrain, which is quite efficient in comparison. Anything basically feels like pedaling through molasses compared to that.

So... I think the best next steps for me here are to try and previous version of OSF, and define a specific route that is repeatable. I need to figure out what the best way to data log the motor output so I can analyze that data. I could just point a gopro at the 860c display, but there has to be better way then things.

 

[HrKlev]

Mmmmm what is it with forums and threads where one bad experience written in such terms as above can influence how it’s presumed all operators will experience the same thing. Sorry we do hard XCountry riding that probably to most would be Enduro style riding but just without jumps, and yet to bend or break a crank.

Just the amount of miles we have done and on the number of engines within our group, all without serious mechanical problems, would indicate that if someone is regularly breaking or bending cranks, then it’s that individuals riding style which is taking his engine outside of its design, rather than all engines of its type. Sure if you want to do black or diamond downhills with large jumps then you only have to take one look at the motor and say it’s not for this purpose, so why fit it and then complain that it’s made of butter.

Sometimes we have to make personal engineering design decisions around our own useage requirements. If you want to fit this engine, use it way outside what it was designed for that of a commuter bolt on and then accuse it of being made of butter because you didn’t bother to do basic engineering design assessment, then please don’t put off other users by your own failure to do so.

I dont know what is up with you every time someone says something negative about this motor? Im not going to join a pissing match about whos doing the hardest riding either. When someone mentions enduro and tsdz2 in same setting, a warning is in its place. I was expecting a lot of issues taking this offroad, but had surprisingly few. Until I bought a second one where the shaft broke, and I was made aware of that design. I would love to have had a warning so I did not have to waste time and money, but I didnt catch it before it was too late.

If it works for you, great! Its still a design flaw, and others might not be so lucky.

A normal square taper crank is 17mm. People break those also. Tsdz2 is 15mm, and the 68-73 version is 14mm in the lock ring groove. The 100mm version is 13.4mm. As for the butter comment, try to put one in a lathe, you will agree. That, plus the underdimensioned shaft is not a great design.

People do what they want, Im just putting facts and a warning in here, no need to get all worked up for it. Go riding instead.

 

[JohnAnanda]

I still have a few questions:
1) David from eco-ebike has suggested that the field weakening and a high cadence can generate heat faster with the OSF. Does anyone here see any logic to that and are there specific parameters I should be looking out for when riding to avoid this? IE, keep cadance below 90, make sure motor RPM is in window XXX-XXX etc?
2) Is field weakening alone, what is allowing the motor to run higher RPM then the stock firmware, or are there other contributors. I've read about Field Weakening, and understand it generally when it would come to an archaic DC motor... but I'm afraid I'm a bit over my head in 3 phase brushless motors... although I'm drinking from the fire hose of knowledge. My assumption at this point is that field weakening is the electric motor equivalent of retarding and advancing ignition timing on an IC engine, and that there is likely a point of no return in here that may be contributing to my problem.
3) How much resistance should I feel when pedaling the bike in the work stand with the system off? I don't really have a great gauge for what this should feel like, but my only comparison is an x01 eagle drivetrain, which is quite efficient in comparison. Anything basically feels like pedaling through molasses compared to that.

1) 2) As Casainho and other developers have explained, field weakening allows for assistance above 4000 rpm for the motor, but at the cost of a lower efficiency (= more heat). That corresponds to around 90 cadence at the crank.
You can have assistance a little above 90 cadence without field weakening if you use a battery voltage higher than your motor, like a 48V or 52V battery on a 36V motor, or 52V battery on a 48V motor.
I suggest you install a firmware version that allows to disable field weakening so you can more clearly see the difference.

3) There's a very significant resistance on mine too. The one way bearing on the blue gear seems to give a lot of drag when engaged, I don't know if that's how it's supposed to be. I will check that when it is time to change it..

 

[Chalo]

A normal square taper crank [spindle] is 17mm. People break those also. Tsdz2 is 15mm, and the 68-73 version is 14mm in the lock ring groove. The 100mm version is 13.4mm. As for the butter comment, try to put one in a lathe, you will agree. That, plus the underdimensioned shaft is not a great design.

Viscount bottom bracket spindles from the 1970s developed a reputation for breaking off at the snap ring groove. They were 17mm in diameter and made in England from good material.

http://pardo.net/bike/pic/fail-001/FAIL-012.html



I think if the TSDZ2 weren't a hokey Chinese aftermarket product, it would already have been subject to product recalls.

 

[Waynemarlow]

I dont know what is up with you every time someone says something negative about this motor?

I'm all up for criticism in its place, but this motor in my view is being unfairly rated at times by posters who are using it in ways it was never intended by the designers and then dissing it because their one example broke for example the crank.

You yourself stated I was expecting a lot of issues taking this offroad, but had surprisingly few. Until I bought a second one where the shaft broke, and I was made aware of that design. So on the basis the first unit was fine and did what you expected but then when the second units shaft broke, all engines of this type according to your logic on the shaft design are butter. What happened to the first units crank, has it too broken ? Are you sure this isn't an an example of isolated manufacturing faults ?

If we take the 6 units we currently have in service where some of us at 95kgs standing on the peddles on all downhill runs, over steps and drops through some pretty rooted terrain, we are not seeing crank damage such as yourself. On your terms of the design being poor, I should be seeing crank failures and as much as I have been expecting a failure as the shaft is not the largest around, we have yet to replace one. Are we operating just inside the design bubble and you are just outside ? Just because we bent 17mm shafts 20 years ago, has metallurgy and modern machining taken that 17mm to a 15mm equivalent, sorry I don't have enough data to prove or disprove that.

Are you sure you are not seeing the damage we used to see on the older first rear suspension mtbs where the BB was to low on full suspension load and the pedals if in the wrong place on landing were grounding, taking the full suspension load as an impact. I do remember those peddle strikes where we used to bend the cranks on all too a regular basis, Do you live in rocky terrain and you are getting hard pedal strikes on the inside of corners ? There's so many other factors to take into account here thats unique to yourself, that in reality without some pretty good evidence other than eyeball adhoc unsupported engineering logic " it must be butter " I would be hesitant to agree with you from our experience of the design.

I would consider your experience as an outlier, it happens in all things where some users get repeated failures and yet the vast majority do not. Read up on the Specialized Brose experience where some users get through 3 engines a year and yet the one who rides in our group has never had a problem in 3 years.

At the price range we are paying for this motor you have to admit you are getting a lot for your money if you treat them as 400 - 500 W assistance motors with torque sensing.

 

[Waynemarlow]

3) Return the unit to Eco-Ebike for them to QA/QC or replace

You got it, what ever you do to this engine, the doubts are now there for you. You would be much better to put the money from the refund into a second hand specifically designed E commuter bike.

 

[MatejO]

Thank you for reply's.

I am more looking in DIY option and to build it strong and light.

p.s. Based on the picture of UK bracket spindles from the 1970s... This looks far from good material composition. Look at the cracks and metal flakes inside the casting!

 

[Chalo]

Based on the picture of UK bracket spindles from the 1970s... This looks far from good material composition. Look at the cracks and metal flakes inside the casting!

That's a surface characteristic of rapid fracture. It's not a reflection of the steel's quality (though cruddy steel is more vulnerable to fatigue cracking followed by fracture than pure steel is).

The Viscount spindle wasn't a casting. It was machined (somewhat crudely) from wrought steel bar. The bike's workmanship wasn't great, but the materials used were good.

 

[HrKlev]

I'm all up for criticism in its place, but this motor in my view is being unfairly rated at times by posters who are using it in ways it was never intended by the designers and then dissing it because their one example broke for example the crank.

Its like 5 different people experiencing the same here: https://endless-sphere.com/forums/viewtopic.php?t=98361 . It is all extended axles, so I think the problem is worse on those, but the design is still pretty much the same on the 68mm version. Enduro bikes usually have wider BB than the old 68mm standard, so it applies to the original post i replied to.

You yourself stated I was expecting a lot of issues taking this offroad, but had surprisingly few. Until I bought a second one where the shaft broke, and I was made aware of that design. So on the basis the first unit was fine and did what you expected but then when the second units shaft broke, all engines of this type according to your logic on the shaft design are butter. What happened to the first units crank, has it too broken ? Are you sure this isn't an an example of isolated manufacturing faults ?

Shortly after the second one broke, the controller fried on my first one (after 980km, so not too bad). After I fund the thread I linked to and saw the design was the same on both type of axles (though, slightly better on the 68mm version), I decided I do not want to risk it and made another setup. Risk is high enough without having a crank shaft snap.

If we take the 6 units we currently have in service where some of us at 95kgs standing on the peddles on all downhill runs, over steps and drops through some pretty rooted terrain, we are not seeing crank damage such as yourself. On your terms of the design being poor, I should be seeing crank failures and as much as I have been expecting a failure as the shaft is not the largest around, we have yet to replace one. Are we operating just inside the design bubble and you are just outside ? Just because we bent 17mm shafts 20 years ago, has metallurgy and modern machining taken that 17mm to a 15mm equivalent, sorry I don't have enough data to prove or disprove that.

Are you sure you are not seeing the damage we used to see on the older first rear suspension mtbs where the BB was to low on full suspension load and the pedals if in the wrong place on landing were grounding, taking the full suspension load as an impact. I do remember those peddle strikes where we used to bend the cranks on all too a regular basis, Do you live in rocky terrain and you are getting hard pedal strikes on the inside of corners ? There's so many other factors to take into account here thats unique to yourself, that in reality without some pretty good evidence other than eyeball adhoc unsupported engineering logic " it must be butter " I would be hesitant to agree with you from our experience of the design.

I would consider your experience as an outlier, it happens in all things where some users get repeated failures and yet the vast majority do not. Read up on the Specialized Brose experience where some users get through 3 engines a year and yet the one who rides in our group has never had a problem in 3 years.

Yes, everything around here is rocks and roots. Maybe I am just outside the bubble. Maybe the guy asking for advie is just outside too? I was just pointing out a major design flaw. A 3mm reduction from the original 17mm design of a square taper BB is HUGE, that is a fact.

This is pure speculations without any soruce, but I highly doubt that any of the other cranks are better than the ones I have seen. Its a cheap chinese motor where cutting costs is the most important factor.

At the price range we are paying for this motor you have to admit you are getting a lot for your money if you treat them as 400 - 500 W assistance motors with torque sensing.

Yes, I liked the motor for what it was, especially with the OSF. If I had not burned the controller, I would probably put it on another bike for road duty. For enduro or any other type of offroad, I would not risk it. What other people do is none of my business, but they do deserve a warning when they ask for advice.