TSDZ2 mid drive with 860C, 850C or SW102 displays only -- Flexible OpenSource firmware (Casainho code only)

[Swannking]

I updated the motor firmware to v1.1.0 and was running it for sometimes until I was experiencing pre-mature cut off from my 52V battery. I noticed the motor and battery completely shut down at about 48V. The last time it happened, I unplugged the battery and plugged it back. The battery came back to live reading about 48V. It happened again 2 more times on my way home. I figured the high current draw must had triggered the battery to cut off even the cut off voltage was set at 42V. I went back to the v1.0.0 and the issue has not recurred yet.

[PieterDerckx]

Hi all, maybe this topic has been raised somewhere but i could not find. Hope someone can support me!

I finally upgraded both TSDZ2 and 850C to the latest OS version, but no startup, after 20sec the 'error brakes' is shown. As far is i could find this points to a communiction error. So i checked a lot of older versions and v05.0 beta 3 is the one that starts. Later versions did not at least.
When i swithc on the v0.5.0 beta 3 values are 'running'; values for battery, speed and power are changing constantly, even without activating anything

I used the DIY cable for programming the TSDZ2, must be OK as both flashing and veryfying shows no errors.
Regarding the 850C i'm not that sure as i (assume) here we just do the flashing, no veryfying.... but flashing just starts for any version after press on the (remote) power button of the 850C. Is there a way to check the flashing otherwise?
For programming the 850 C i used the DIY version to acc. https://github.com/OpenSource-EBike-fir ... bootloader

Hope this sounds familiar to one of the users.
For now i would like to have 'any' version to switch back to be able to use my velo (that is where i installed it).

The communication speed were increased to twice the speed and so possible long cables with not so good contacts may give this issues. Luckily there are almost no one with this issue. Please check your cable connections.

Hi casainho, thanks for your feedback. I checked the DIY bootloader, reworked it with new wiring but still the same effect and only uo to version 0.5.0 beta 2 is flashed "ok", but starting this version shows changing values like before, unfortunatly.
I tried to flash the original firmware (APT) using the same DIY box with APT burn tool...no succes..So i guess the DIY box is not OK.
If i understand correct, the (yellow) Bafang bootloader box should work perfectly. So i better arrange one of those, correct?

[Peacepirate]

Alrighty,
finally i installed the newest firmware 1.1.0 to my mothers bike. thanks for the implementation of the sw102 display. Allthough the whole system crashes if entering the last two menu config options the display works fine.
But i wonder in which funny world this firmware is an improvement compared to the stock firmware apart from the display implementation. Are you kidding me??? Months of work and programming and this is the result??? For real? I tested the firmware before giving it to my mother for hours incl. torque calibration and so on, put all possible values to max and i have to pedal 720 degress before getting max motor support? when i push the pedal from standstill on stock firmware i get max support right from the start and i feel immediatelly the power. one has to be a real display enthusiast to consider this firmware an improvement compared to stock firmware. i appreciate the display compatibility but not for the price of a castrated motor software. i was so naive to believe that open source firmware adds options and power and does not weaken the power. just for curiosity i gave my mother the the bike for testing. Her short request after driving to the market and back: delete it!

[casainho]

i was so naive to believe that open source firmware adds options and power and does not weaken the power. just for curiosity i gave my mother the the bike for testing. Her short request after driving to the market and back: delete it!

Very interesting experience! I wonder why the stock firmware is what it is, maybe because this motor is intended for users like your mother that do not want the advanced features like motor support up to pedal cadence of 120 RPM, like the big brands motors have. Maybe this motor is intended for a city EBike and nothing more.

[HughF]

Alrighty,
finally i installed the newest firmware 1.1.0 to my mothers bike. thanks for the implementation of the sw102 display. Allthough the whole system crashes if entering the last two menu config options the display works fine.
But i wonder in which funny world this firmware is an improvement compared to the stock firmware apart from the display implementation. Are you kidding me??? Months of work and programming and this is the result??? For real? I tested the firmware before giving it to my mother for hours incl. torque calibration and so on, put all possible values to max and i have to pedal 720 degress before getting max motor support? when i push the pedal from standstill on stock firmware i get max support right from the start and i feel immediatelly the power. one has to be a real display enthusiast to consider this firmware an improvement compared to stock firmware. i appreciate the display compatibility but not for the price of a castrated motor software. i was so naive to believe that open source firmware adds options and power and does not weaken the power. just for curiosity i gave my mother the the bike for testing. Her short request after driving to the market and back: delete it!

You need AWOP and Startup Boost = disabled otherwise the open source firmware is a bit lame off the line, won't disagree...

AWOP makes the real difference IMHO.

[jeff.page.rides]

Alrighty,
finally i installed the newest firmware 1.1.0 to my mothers bike. thanks for the implementation of the sw102 display. Allthough the whole system crashes if entering the last two menu config options the display works fine.
But i wonder in which funny world this firmware is an improvement compared to the stock firmware apart from the display implementation. Are you kidding me??? Months of work and programming and this is the result??? For real? I tested the firmware before giving it to my mother for hours incl. torque calibration and so on, put all possible values to max and i have to pedal 720 degress before getting max motor support? when i push the pedal from standstill on stock firmware i get max support right from the start and i feel immediatelly the power. one has to be a real display enthusiast to consider this firmware an improvement compared to stock firmware. i appreciate the display compatibility but not for the price of a castrated motor software. i was so naive to believe that open source firmware adds options and power and does not weaken the power. just for curiosity i gave my mother the the bike for testing. Her short request after driving to the market and back: delete it!

You need AWOP and Startup Boost = disabled otherwise the open source firmware is a bit lame off the line, won't disagree...

AWOP makes the real difference IMHO.

What is AWOP?

[HughF]

You need AWOP and Startup Boost = disabled otherwise the open source firmware is a bit lame off the line, won't disagree...

AWOP makes the real difference IMHO.

What is AWOP?

AWOP is the Assistance WithOut Pedal rotation function

[ilu]

I have a problem which I cannot figure out. I have 36V motor with 52V battery and I was using 1.0.0 OSF with SW102 display for a while and everything was working well.. Then at first motor would stop functioning at random times for a while, sometimes coming back after restarting, sometimes after pedalling for a bit.

Then the motor went completely dead. Display was working fine, and was showing adequate readings except for the hall sensor numbers, which were rotating in random order. I thought it was a fault in one of the sensors, but changing the hall sensor assebly from another similar motor did not resolve the issue, nor did changing the whole motor. The motor would not work and the hall numbers still random. When activating the motor with walk assist or throttle, there is a faint 'click' sound from the motor. Also some current goes to the coils as the resistance to turning the motor by hand is increased when it's activated.

Then I tried changing back to stock firmware and display, and surprisingly with those the motor is working well. Is there something fundamentally different how the current is delivered to the motor that could make the difference? Or is it possible that this is a fault in the display? I even tried changing the controller to a new one but still the behaviour is the same.

[skestans]

Alrighty,
finally i installed the newest firmware 1.1.0 to my mothers bike. thanks for the implementation of the sw102 display. Allthough the whole system crashes if entering the last two menu config options the display works fine.
But i wonder in which funny world this firmware is an improvement compared to the stock firmware apart from the display implementation. Are you kidding me??? Months of work and programming and this is the result??? For real? I tested the firmware before giving it to my mother for hours incl. torque calibration and so on, put all possible values to max and i have to pedal 720 degress before getting max motor support? when i push the pedal from standstill on stock firmware i get max support right from the start and i feel immediatelly the power. one has to be a real display enthusiast to consider this firmware an improvement compared to stock firmware. i appreciate the display compatibility but not for the price of a castrated motor software. i was so naive to believe that open source firmware adds options and power and does not weaken the power. just for curiosity i gave my mother the the bike for testing. Her short request after driving to the market and back: delete it!

You should try and get your money back for the open source firmware if you’re not happy with it. You’re owed that much, never mind how many man hours were put into making this firmware at no cost to anyone using it.

[skestans]

AWOP is the Assistance WithOut Pedal rotation function

I personally don’t like AWOP because I always have a foot on the pedal while holding the brakes at red lights. With AWOP I’d need brake sensors and I have hydraulic breaks so there are no factory made elegant levers that has them built in. And the extra pedaling to start off is really not that bad in my opinion, especially if you know how to ride a bike and downshift all the way before stopping.

[casainho]

AWOP is the Assistance WithOut Pedal rotation function

I personally don’t like AWOP because I always have a foot on the pedal while holding the brakes at red lights. With AWOP I’d need brake sensors and I have hydraulic breaks so there are no factory made elegant levers that has them built in. And the extra pedaling to start off is really not that bad in my opinion, especially if you know how to ride a bike and downshift all the way before stopping.

What is great about our firmware we developed is that we implemented that feature in a way that the user can quick enable or disable it, so, everyone will be happy

[casainho]

I have a problem which I cannot figure out. I have 36V motor with 52V battery and I was using 1.0.0 OSF with SW102 display for a while and everything was working well.. Then at first motor would stop functioning at random times for a while, sometimes coming back after restarting, sometimes after pedalling for a bit.

Then the motor went completely dead. Display was working fine, and was showing adequate readings except for the hall sensor numbers, which were rotating in random order. I thought it was a fault in one of the sensors, but changing the hall sensor assebly from another similar motor did not resolve the issue, nor did changing the whole motor. The motor would not work and the hall numbers still random. When activating the motor with walk assist or throttle, there is a faint 'click' sound from the motor. Also some current goes to the coils as the resistance to turning the motor by hand is increased when it's activated.

Then I tried changing back to stock firmware and display, and surprisingly with those the motor is working well. Is there something fundamentally different how the current is delivered to the motor that could make the difference? Or is it possible that this is a fault in the display? I even tried changing the controller to a new one but still the behaviour is the same.

That is a very strange issue. If you already changed the motor controller, I would say the issue is with the hall sensors or motor itself.
I don't think the display would be the source of that issue.

Also some current goes to the coils as the resistance to turning the motor by hand is increased when it's activated.
When the motor is activated, it should be rotating, the energy applied to coils always force movement unless the hall sensors signals are not correct and then the movement will not happen.

Make sure the hall sensor cables are working, use the display to debug and make sure all the signals are working as expected.

[HughF]

AWOP is the Assistance WithOut Pedal rotation function

I personally don’t like AWOP because I always have a foot on the pedal while holding the brakes at red lights. With AWOP I’d need brake sensors and I have hydraulic breaks so there are no factory made elegant levers that has them built in. And the extra pedaling to start off is really not that bad in my opinion, especially if you know how to ride a bike and downshift all the way before stopping.

You don't need brake sensors. Just calibrate your torque sensor properly and hold the bike on the brakes. It's what I do at the lights. I'm quicker off the lights than most scooters up to about 15mph when in a high assist level.

I also have full hydraulic discs on my hardtail mtb, I would never bother with brake sensors, don't see the point in them.

[Alfie]

Hi

I have just purchased the 860c display for my TSDZ2. I have built the ST link and can see the controller firmware with it, so all good there.

The one question I have is do I flash the 860c with the same ST Link and configuration or do I need to make a different one? The wiki is a bit sketch about this.

My unit is 6 pin.

Any help would be much appreciated.

[ilu]

That is a very strange issue. If you already changed the motor controller, I would say the issue is with the hall sensors or motor itself.
I don't think the display would be the source of that issue.

Also some current goes to the coils as the resistance to turning the motor by hand is increased when it's activated.
When the motor is activated, it should be rotating, the energy applied to coils always force movement unless the hall sensors signals are not correct and then the movement will not happen.

Make sure the hall sensor cables are working, use the display to debug and make sure all the signals are working as expected.

I checked the hall sensor voltages, which were fine. Then I realized that actually the hall sequence in the display is repeating, I think before I turned the motor too fast and was missing some numbers. In the beginning of the malfunction the display was showing error code (something about torque sensor but apparently relating to hall sensors, at least in version 1.0.0) which led me to think the problem would be there.

Anyway, at one point I changed the whole motor to my bike with the 52V battery and was troubleshooting this motor with a 36V battery. And I forgot to change the low cut-off voltage from the display settings. So I lowered that and now the motor is working fine. Although I still have no idea what was the problem in the first place. But at least now I know to double check all of the settings before starting to look for the fault.

As a curious side note, when the voltage is below the cut-off and walk assist is activated, something is still happening with the motor, as the resistance to turning the motor is increased and also becomes smoother, ie. losing the usual bumpiness.

[skestans]

AWOP is the Assistance WithOut Pedal rotation function

I personally don’t like AWOP because I always have a foot on the pedal while holding the brakes at red lights. With AWOP I’d need brake sensors and I have hydraulic breaks so there are no factory made elegant levers that has them built in. And the extra pedaling to start off is really not that bad in my opinion, especially if you know how to ride a bike and downshift all the way before stopping.

You don't need brake sensors. Just calibrate your torque sensor properly and hold the bike on the brakes. It's what I do at the lights. I'm quicker off the lights than most scooters up to about 15mph when in a high assist level.

I also have full hydraulic discs on my hardtail mtb, I would never bother with brake sensors, don't see the point in them.

I don’t remember if I had the issue before or after calibrating. Maybe I’ll try switching AWOP on again and see.

You’re saying you have AWOP, no brake sensors, and you’re able to rest one foot on the group one foot on the pedal while braking waiting for the light and there is no torque from the motor? For me it was always engaging the motor and the brakes kept it from moving the bike but that’s not good in the long run so AWOP was switched off.

[HughF]

I rest my foot on the pedal, the motor is ready to go and there is a little torque there. I just hold it on the brakes. As soon as the lights change, bang, I am off like a rocket.

[skestans]

I rest my foot on the pedal, the motor is ready to go and there is a little torque there. I just hold it on the brakes. As soon as the lights change, bang, I am off like a rocket.

Well I tried and I don’t like it. The motor forces against the brakes which can’t be good for the gears. And it’s very hard to shift up or down without a loud CRUNCH noise from shifting under full load whereas without AWOP there is a window where assist stops totally and I can shift under minimal load without slowing down. Back to AWOP off for me.

[HughF]

Fair enough... I consider my blue gear to be expendable, so holding it on the brakes is no issue for me. I also have a hub gear which shifts up under power without issue. Downshifts are the problem, haha

[Mr.Flibble]

Can anyone explain what an ADC is and what an ADC step is please?

I am still trying to get rid of this over run, and the config instructions say :
"Min current ADC step: 1
Min current value (in ADC steps) while pedaling, even if not pressing the pedals. This is nice to keep motor engaged and ready even if rotating pedals with minimal force that it is not even detected by the torque sensor."

Should I change it to 0, or higher than 1 to stop the "motor engaged and ready even if rotating pedals with minimal force that it is not even detected by the torque sensor." ?

Thx in anticipation.

[Waynemarlow]

I'm not sure its at all possible to prevent a minimum of over run, there is a lag between power input sensors and power input to the motor, enough to probably get a small part rotation on occassion. If you fit a brake cut out it will stop instantly that over run.

Certainly the version 20.1 was the best for overrun that I have tried, even better than the stock factory if you over volt the engine, 52 volts seems to really magnify this problem. One of my motors is a real pain in that the torque sensor is so sensitive that you only have to give a peddle a nudge accidently at say a gate and the over run will move the bike. My other engine is good as a mouse and you can abuse the pedals without a murmur. The OS software can reset that parameter somewhat.

Also if you read and try the latest Bosch and Shimano engines, they are all advocating that last little bit of over run as an enhancement to climbing over obstacles and the likes and indeed in practice I would have to concur, but I suspect its the factory units having the same problem, there is jsut some legacy from the delay of reading the torque sensor. Once you get your head around the knowledge that there is a slight over run, you can use it to some advantage in steep off road sections much like the trials riders do to launch up faces with a quick blast of power.

[Mr.Flibble]

I didn't explain very well.
It's the backing off the power to change gear I am looking at here. Currently it's full power unless I stop pending for so long I almost stop if going uphill.
Being able to turn the cranks with little to no pressure while changing gear would be good, just like you could with the original firmware.

[abrainer]

Hi Casainho,
you are looking for a source for the new TSDZ2 controllers with 32 bit XMC MCU.
I have ordered some from https://enerprof.de/motor-kits/display- ... sion?c=234. It is the new version for 25€. I already disassembled one of them and measured the pinout of the MCU. Are you interested in it? Then I would translate the document and send it to you.
Best regards,
abrainer

[casainho]

Hi Casainho,
you are looking for a source for the new TSDZ2 controllers with 32 bit XMC MCU.
I have ordered some from https://enerprof.de/motor-kits/display- ... sion?c=234. It is the new version for 25€. I already disassembled one of them and measured the pinout of the MCU. Are you interested in it? Then I would translate the document and send it to you.
Best regards,
abrainer

Hi.

Thanks. I am busy on the TSDZ2 wireless project. I think there is no reason to use the new motor controller while the previous one is on selling for the same cheap price.

But please share here public all the information you have and others may use it for development.

[PieterDerckx]

Hi all, maybe this topic has been raised somewhere but i could not find. Hope someone can support me!

I finally upgraded both TSDZ2 and 850C to the latest OS version, but no startup, after 20sec the 'error brakes' is shown. As far is i could find this points to a communiction error. So i checked a lot of older versions and v05.0 beta 3 is the one that starts. Later versions did not at least.
When i swithc on the v0.5.0 beta 3 values are 'running'; values for battery, speed and power are changing constantly, even without activating anything

I used the DIY cable for programming the TSDZ2, must be OK as both flashing and veryfying shows no errors.
Regarding the 850C i'm not that sure as i (assume) here we just do the flashing, no veryfying.... but flashing just starts for any version after press on the (remote) power button of the 850C. Is there a way to check the flashing otherwise?
For programming the 850 C i used the DIY version to acc. https://github.com/OpenSource-EBike-fir ... bootloader

Hope this sounds familiar to one of the users.
For now i would like to have 'any' version to switch back to be able to use my velo (that is where i installed it).

The communication speed were increased to twice the speed and so possible long cables with not so good contacts may give this issues. Luckily there are almost no one with this issue. Please check your cable connections.

Hi casainho, thanks for your feedback. I checked the DIY bootloader, reworked it with new wiring but still the same effect and only uo to version 0.5.0 beta 2 is flashed "ok", but starting this version shows changing values like before, unfortunatly.
I tried to flash the original firmware (APT) using the same DIY box with APT burn tool...no succes..So i guess the DIY box is not OK.
If i understand correct, the (yellow) Bafang bootloader box should work perfectly. So i better arrange one of those, correct?

I flashed using the original APT tool, but still no succes
Luckily I can reflash the original APT SW, but that means i'm back to start....
Seems like my 850C is not able to 'accept' the OSF. Did anyone have a similar experience?
What i notice is that the original SW is larger, taking more time to flash. Any thoughts?