New "TSDZ2 Torque Sensor Central Motor"

Waynemarlow said:
The only time we have seen motor current draw problems was both times a battery fault. Turned out to be a poor weld on some of the cells which was restricting the current, as soon as you stopped or at rest the voltage would recover back to the norm.

that would cause a drastic voltage drop of the battery under current load?

k, next time, I'll setup one of the displays to show batt voltage and current along with motor current and human watts


Swannking said:
May be the torque sensor is failing. I would go into the technical menu which shows the value of the sensor while u are riding and try to see what’s going on with the value when that happens. I highly recommend for u to do the sensor calibration. I used a cheap luggage scale.

isn't the 'human power' display directly related to the torque sensor (after being run through the calibration table) times the cadence ? 'human power' was reading fine, like hovering around 150 or 200 watts or whatever while pedaling briskly, while motor power was 0, then suddenly motor power jumps back to whatever it should be.


casainho said:
If the human power is well calculated (you can test it with motor disabled at assist level 0), motor should not give different currents / power than what you expect. I can´t imagine why that would happen. There no reports of such issue.

I hope you can figure out the issue...

And don´t forget to update to latest firmware version.

I need to figure out why my ST-Link didn't work when I tried setting it up.
 
casainho said:
...

And don´t forget to update to latest firmware version.

oh. when I upgrade 1.0.0 to latest, does that reset all my settings, so I need to go through all the menus and reset everything?
 
LeftCoastNurd said:
casainho said:
...

And don´t forget to update to latest firmware version.

oh. when I upgrade 1.0.0 to latest, does that reset all my settings, so I need to go through all the menus and reset everything?

Yes it does reset everything. It may be a dumb question. Did you also update the firmware of the motor to the corresponding .hex file?
 
Swannking said:
Yes it does reset everything. It may be a dumb question. Did you also update the firmware of the motor to the corresponding .hex file?

as of right now, I'm still running the 1.0.0 firmware that eco-ebikes flashed before they shipped it to me. everything was working fine until a couple rides ago, and on the last ride yesterday it misbehaved a fair bit more than the ride before that.

I was kinda fried today, so I didn't go anywhere, and rain is in the forecast so it might be a few days before I ride it again.

I have the stlink and 850/860c bootloader box, both purchased with the motor from eco-ebikes, my one atteempt at setting up the software using my laptop did not work. when I tried to 'read' the existing firmware into a file, it said it could not communicate ...
Code:
Hardware configuration set to <ST-LINK on USB with STM8S105x6 plugged in and SWIM protocol>.
Blank check before programming (when available) is OFF
Verify after programming is ON.
Clear memory before loading a file is ON.
Erase device memory before programming (when available) is OFF
Display a popup window when an error occurs is ON.
Log activity is OFF
> Reading  PROGRAM MEMORY area...
Error : Cannot communicate with the tool.
Wrong tool selection or check tool power supply or check that a previous session is closed.
Error : < PROGRAM MEMORY reading failed.
Error : < Operation aborted.
 
LeftCoastNurd said:
Swannking said:
Yes it does reset everything. It may be a dumb question. Did you also update the firmware of the motor to the corresponding .hex file?

as of right now, I'm still running the 1.0.0 firmware that eco-ebikes flashed before they shipped it to me. everything was working fine until a couple rides ago, and on the last ride yesterday it misbehaved a fair bit more than the ride before that.

I was kinda fried today, so I didn't go anywhere, and rain is in the forecast so it might be a few days before I ride it again.

I have the stlink and 850/860c bootloader box, both purchased with the motor from eco-ebikes, my one atteempt at setting up the software using my laptop did not work. when I tried to 'read' the existing firmware into a file, it said it could not communicate ...
Code:
Hardware configuration set to <ST-LINK on USB with STM8S105x6 plugged in and SWIM protocol>.
Blank check before programming (when available) is OFF
Verify after programming is ON.
Clear memory before loading a file is ON.
Erase device memory before programming (when available) is OFF
Display a popup window when an error occurs is ON.
Log activity is OFF
> Reading  PROGRAM MEMORY area...
Error : Cannot communicate with the tool.
Wrong tool selection or check tool power supply or check that a previous session is closed.
Error : < PROGRAM MEMORY reading failed.
Error : < Operation aborted.

Did u finish this step “ select the 'Option Byte' tab, and then choose 'Current tab' from the Read Menu to read and fill the option byte values from your TSDZ2.”
 
James Broadhurst said:
I'm thinking of building a 48v coaster brake motor (they only supply a 36v) by using a normal 48v and locking by some means or other the sprague clutch. That should work?

Probably be easier to swap in a 48v motor to the coaster version... "Some means" is kind of hand-wavy.

But, why coaster brake? They are the least effective brake available on a bicycle. E-bikes go fast, and need good brakes.
 
James Broadhurst said:
I'm thinking of building a 48v coaster brake motor (they only supply a 36v) by using a normal 48v and locking by some means or other the sprague clutch. That should work?

The normal motors use helical teeth in the main gear and reduction. The coaster brake motors use straight cut. So keep that in mind. You cannot just swap the gear out. Also keep in mind that the coaster brake programming is different too. So you need the coaster brake type firmware in the program memory tab for sure.

What I do, and what I recommend - get a normal 36v coaster brake motor, and switch out the inside motor to 48v, then reprogram it. FYI you cannot reprogram the coaster brake motor the same way as you do other motors, by uploading the profile, it will cause an error and won't work. You have to go in and manually change the voltage in data memory. Recently they also changed something else and now it's a pain to do. For some reason I cannot get them to go 52v anymore like I used to, but can set them up 48v still no problem. Let me know if you need any assistance I can walk you through it.
 
eyebyesickle said:
FYI you cannot reprogram the coaster brake motor the same way as you do other motors, by uploading the profile, it will cause an error and won't work. You have to go in and manually change the voltage in data memory. Recently they also changed something else and now it's a pain to do. For some reason I cannot get them to go 52v anymore like I used to, but can set them up 48v still no problem. Let me know if you need any assistance I can walk you through it.
Hmmm ... that's exactly what I did. I loaded v19.0 with a 48v battery and all is well.
 
eyebyesickle said:
James Broadhurst said:
I'm thinking of building a 48v coaster brake motor (they only supply a 36v) by using a normal 48v and locking by some means or other the sprague clutch. That should work?

The normal motors use helical teeth in the main gear and reduction. The coaster brake motors use straight cut. So keep that in mind. You cannot just swap the gear out. Also keep in mind that the coaster brake programming is different too. So you need the coaster brake type firmware in the program memory tab for sure.

What I do, and what I recommend - get a normal 36v coaster brake motor, and switch out the inside motor to 48v, then reprogram it. FYI you cannot reprogram the coaster brake motor the same way as you do other motors, by uploading the profile, it will cause an error and won't work. You have to go in and manually change the voltage in data memory. Recently they also changed something else and now it's a pain to do. For some reason I cannot get them to go 52v anymore like I used to, but can set them up 48v still no problem. Let me know if you need any assistance I can walk you through it.

I've used both 48 Volt & 36 Volt coaster brake Motors with the same firmware and all the same settings other than setting it to a 48 Volt motor or 36 Volt motor all else being equal and the same. I use a 52V battery. I like the 36V better than the 48V both use a 52V battery.
 
I would like to give an update of the TSDZ2 EBike wireless project: https://github.com/OpenSource-EBike-firmware/TSDZ2_wireless

I finally got the mobile app working (Bluetooth communication), for doing the configurations. As for display, it will be wireless communications (ANT+ LEV) and it is optional, it will be the Garmin Edge as seen on the picture.

Now I need to do copy-paste work to implement the full configurations, as seen on the video and then I will be able to finally install the wireless board on my TSDZ2 and remove the big 860C display :)

Mobile app configurations (WIP) -- may thanks to the TSDZ2-ESP32 from where I reused the mobile app:
[youtube]RCztvzQ0O6A[/youtube]

The wireless board connected to TSDZ2 display connector:
image.png


We also have our own DIY wireless remote reusing the VLCD5 remote. But since we are using EBike wireless standards, any other commercial remote will work. A video from @rananna developer showing the wireless remote controlling changing his EBike assist level as also changing the page on his Garmin Edge display:
[youtube]s7URIMVzcwc[/youtube]
 
hmmm, trying to load 1.1.0 in, using STVP... eco-ebike's instructions talk about a separate program and data .s19 file, but the github for 1.1.0 only has a single .hex file... do I load this into the program memory, and leave the data memory alone? or do I load the same hex file into both program and data and rely on the addresses of the programming be out of range for hte data ?
 
LeftCoastNurd said:
hmmm, trying to load 1.1.0 in, using STVP... eco-ebike's instructions talk about a separate program and data .s19 file, but the github for 1.1.0 only has a single .hex file... do I load this into the program memory, and leave the data memory alone? or do I load the same hex file into both program and data and rely on the addresses of the programming be out of range for hte data ?

Just load the .hex to the program memory.
 
ilu said:
Just load the .hex to the program memory.

indeed, that worked. i actually tried loading the hex into the data memory but all rows of the hex were in the 0x8000 range of the program memory, so were rejected.

anyways, v1.1.0 is up and running. now I need to go for a decently long ride, but not today.
 
Hi all - first post here.
Having trouble with the "Flash SW102" step shown here:
https://github.com/OpenSource-EBike-firmware/TSDZ2_wiki/wiki/Flash-the-firmware-on-SW102

I can see the SW102 in nRF Connect and nRF Toolbox iOS apps when the SW102 is turned on with the display active.
I can't see the SW102 in either app once I put it into bootloader mode (hold both buttons for 8 secs, display goes blank).
Both apps say that my SW102 doesn't support DFU.

I have taken a sequence of photos and screenshots in the hope that someone here can help me out - many thanks.
 

Attachments

  • 1 SW102 On.JPG
    1 SW102 On.JPG
    69.9 KB · Views: 1,887
  • 2 nRF Toolbox.PNG
    2 nRF Toolbox.PNG
    24.9 KB · Views: 1,887
  • 3 nRF Toolbox No DFU.PNG
    3 nRF Toolbox No DFU.PNG
    20.3 KB · Views: 1,887
  • 4 nRF Connect.PNG
    4 nRF Connect.PNG
    49.9 KB · Views: 1,887
  • 5 nRF Connected.PNG
    5 nRF Connected.PNG
    47.2 KB · Views: 1,887
  • 6 nRF Connect Info 1.PNG
    6 nRF Connect Info 1.PNG
    49.4 KB · Views: 1,887
  • 7 nRF Connect Info 2.PNG
    7 nRF Connect Info 2.PNG
    45.2 KB · Views: 1,887
  • 8 nRF Connect Info 3.PNG
    8 nRF Connect Info 3.PNG
    42.6 KB · Views: 1,887
  • 9 nRF Connect Info 4.PNG
    9 nRF Connect Info 4.PNG
    46.2 KB · Views: 1,887
  • 10 nRF Connect No DFU.PNG
    10 nRF Connect No DFU.PNG
    37.8 KB · Views: 1,888
  • 11 Held Buttons Bootloader Mode.JPG
    11 Held Buttons Bootloader Mode.JPG
    148.9 KB · Views: 1,888
  • 12 nRF Toolbox Not Found.PNG
    12 nRF Toolbox Not Found.PNG
    11.1 KB · Views: 1,888
  • 13 nRF Connect Not Found.PNG
    13 nRF Connect Not Found.PNG
    40.3 KB · Views: 1,887
More news about TSDZ2 wireless:

Wireless firmware update bootloader is ready!!

@rananna did the work and the hex files, sources and technical description are here: https://github.com/OpenSource-EBike-firmware/TSDZ2_wireless-bootloader

To resume: firmware can be updated using Bluetooth connection on a smartphone or a PC and to start the bootloader / firmware update, user will need to do one of the following options:
- long press the button in the TSDZ2 wireless board at startup (on the remote wireless, long press all buttons)
- use NRFconnect on smartphone or PC to write 1 to Bluetooth specific characteristic to enable the bootloader

Here is a video @rananna did explaining how to update firmware on a PC and using the Bluetooth connection:
[youtube]E0GXf4Y68pM[/youtube]

I hope this video can give some clues to @DaveL on the usual process to update firmware on the SW102 display and the process is similar.
 
James Broadhurst said:
eyebyesickle said:
FYI you cannot reprogram the coaster brake motor the same way as you do other motors, by uploading the profile, it will cause an error and won't work. You have to go in and manually change the voltage in data memory. Recently they also changed something else and now it's a pain to do. For some reason I cannot get them to go 52v anymore like I used to, but can set them up 48v still no problem. Let me know if you need any assistance I can walk you through it.
Hmmm ... that's exactly what I did. I loaded v19.0 with a 48v battery and all is well.

Didn't realize you were using OSF. With stock programming it doesn't work out like that.
 
James Broadhurst said:
Hmmm ... that's exactly what I did. I loaded v19.0 with a 48v battery and all is well.

isn't 1.1.0 the current OSF version ? mine, bought with the OSF preinstalled by eco-ebikes, came with 1.0.0 and I recently upgraded it to 1.1.0.... or is this something specific to the coaster brake version?
 
This seems to be how Tongsheng does the torque sensor hardware calibration (more pictures here):

index.php


And after we need to do the calibration on the firmware, only then we will have the best responsive TSDZ2 motor: https://github.com/OpenSource-EBike-firmware/TSDZ2_wiki/wiki/How-to-calibrate-the-torque-sensor
 
eyebyesickle said:
izeman said:
As my chainline is soooo out of straight I thought of the Alfine SG-S7001-11 (11 gears, 409% from lowest to highest gear, ebike approved). Any comments on this before I buy (the whole set is almost 450€ with gear lever and stuff ...)


Screenshot from 2020-02-24 20-30-05.png

Looks nice. I haven't been paying attention to IGH much lately. It seems many are ebike rated now... A bit stronger internally.

There will be some 50T chain rings for TSDZ with 10mm offset in maybe a month, FYI, but nothing bigger than the 42T 10mm offset for now...

Can anybody tell me if there are already bigger (50 or 52) chainrings with offset on the market?

Gert
 
Figured you guys might find this useful.



Building on najbyk's work I made some aluminum heatsink shims that replace the thermal pads between the motor endplate and gearbox housing. I actually made a crude test version a few months back and it performed well, I was never able to get the temperature above 42C (48V motor 52V battery, max current limits). Since installing the final version I've only had one warm test day which while I wasn't watching the temps super close it seemed to peak out in the low 30s after some steep climbs. I made a few of these and threw them up for free here, so if anyone wants one just cover the shipping and it's yours. If people are interested I might make some more. Edit 1/7/21 - due to popular demand I've made some more of these, available at the link above


 
Have a TSDZ2 with the wonderful open source firmware and a gear-change sensor. I found that the delay plus ramp up after a gear change was really annoying, so came up with a quick solution using a microcontroller to adjust how much time the gear sensor triggered the brake circuit. Couldn't be bothered to design up my own board, so just used a very common ATtiny85 board available from many places. I got mine from Jaycar here in Australia.
https://www.jaycar.com.au/duinotech-attiny85-micro-usb-development-board/p/XC3940

The beauty of these is they are programmable using the Arduino environment without any specific programmer. Just connect to a computer via a USB/microUSB cable. Kind of like using a sledgehammer to crack a walnut, but much easier than designing a discreet circuit, and has the flexibility to have the timing easily adjusted in software.

I cut the gear sensor cable and wired it through the ATtiny board. Both +5V wires soldered together to the 5v terminal on the board. Both grounds together to the ground on the board. Trigger from sensor to P2, output to brake cutoff circuit from P0. Code file is attached. Excuse the poor code! Should have enough comments to make sense though.

I programmed the board to run in it's 1MHz mode in the Arduino IDE, and unsoldered the boards power LED. Without LED and running at 1MHz the draw should only be a bit over 1mA. Even more power savings could be gained by putting the processor to sleep and using an interrupt to wake, but I figure with only 1mA draw it's not worth it.

After testing I found that 100ms of brake "blip" was enough for smooth shifting, given that I have a fairly slow max ramp set in the OSF.

Bit of heat shrink around the outside and I'm super happy.
 

Attachments

  • code.txt
    1.5 KB · Views: 25
I'm having a problem with my motor, I can often rotate the pedal a quarter of a turn before it feels any resistance also if I push down hard it feels like something inside the motor is turning or slipping without the gear moving. Is this the sprag clutch failing?
 
Blacklite said:
Have a TSDZ2 with the wonderful open source firmware and a gear-change sensor. I found that the delay plus ramp up after a gear change was really annoying, so came up with a quick solution using a microcontroller to adjust how much time the gear sensor triggered the brake circuit. Couldn't be bothered to design up my own board, so just used a very common ATtiny85 board available from many places. I got mine from Jaycar here in Australia.
https://www.jaycar.com.au/duinotech-attiny85-micro-usb-development-board/p/XC3940

Very cool. What gear shift detector are you using?
 
Back
Top