Tsdz2 firmware open source adapted to vlcd5, vlcd6 and xh18

I have just tried the newest version 19a/036, and it worked great for a few km's.
But after stopping at a store, I turned off the battery when I went in, and when I came out, nothing works.

The display (vlcd5) almost seems frozen, no buttons are working, not info, not +- assistance level.. speedometer does not work. Also the engine did not assist anything.

oeNFNc1m.jpg


I have tried reflashing older version, but the error persists.. Could I have bricked my vlcd5 somehow?


EDIT: The odometer was not 9999miles, and I had my speed in km/h, before this problem appeared..
 
flufferty said:
I have just tried the newest version 19a/036, and it worked great for a few km's.
But after stopping at a store, I turned off the battery when I went in, and when I came out, nothing works.

The display (vlcd5) almost seems frozen, no buttons are working, not info, not +- assistance level.. speedometer does not work. Also the engine did not assist anything.

oeNFNc1m.jpg


I have tried reflashing older version, but the error persists.. Could I have bricked my vlcd5 somehow?


EDIT: The odometer was not 9999miles, and I had my speed in km/h, before this problem appeared..

Did you install a temperature sensor(possibly incorrectly?). Someone else had the wrong wiring for the temp sensor and their display looked like that. Try reconnecting main cable connection to LCD, might be bad connection on data wire. Check for damage to the main cable wire.

Others have had success by using using stlink to write 00s to their controller, then reprogram when they were having random errors.
 
wpenner said:
Did you install a temperature sensor(possibly incorrectly?). Someone else had the wrong wiring for the temp sensor and their display looked like that. Try reconnecting main cable connection to LCD, might be bad connection on data wire. Check for damage to the main cable wire.

Others have had success by using using stlink to write 00s to their controller, then reprogram when they were having random errors.

No, no temperature sensor or external mods, only flashing of the firmware. I have checked inside all cables and all pins look good. Do you happen to have eeprom/flash zeroed-files? Im running on Mac so I use the stm8flash app, not the gui seen in screenshots.

I unplugged the speed sensor and the buttons worked again, I could enable off-road for example, so the flash seems to be working, but after giving power either by walkassistance or pedalling the display "locks" again and only backlight on/off works..

UPDATE/EDIT:
I manned up and took a vlcd5 from my other bike, and tried it, and everything works again. The speedsensor is correct, the firmware answers to enable/disable commands, and walkassist/pedalling works again.

Considering I only used 19a/036 for ~20kms and 3 stops, and this happened, I am suspicious of the firmware..
(before this I have used earlier vlcd5-marcoq for months..)

I guess this is the time to buy a 850c display.. :lol:
 
flufferty said:
wpenner said:
Did you install a temperature sensor(possibly incorrectly?). Someone else had the wrong wiring for the temp sensor and their display looked like that. Try reconnecting main cable connection to LCD, might be bad connection on data wire. Check for damage to the main cable wire.

Others have had success by using using stlink to write 00s to their controller, then reprogram when they were having random errors.

No, no temperature sensor or external mods, only flashing of the firmware. I have checked inside all cables and all pins look good. Do you happen to have eeprom/flash zeroed-files? Im running on Mac so I use the stm8flash app, not the gui seen in screenshots.

I unplugged the speed sensor and the buttons worked again, I could enable off-road for example, so the flash seems to be working, but after giving power either by walkassistance or pedalling the display "locks" again and only backlight on/off works..

UPDATE/EDIT:
I manned up and took a vlcd5 from my other bike, and tried it, and everything works again. The speedsensor is correct, the firmware answers to enable/disable commands, and walkassist/pedalling works again.

Considering I only used 19a/036 for ~20kms and 3 stops, and this happened, I am suspicious of the firmware..
(before this I have used earlier vlcd5-marcoq for months..)

I guess this is the time to buy a 850c display.. :lol:

Flashing the firmware does not write anything to the VCLD5. Fairly confident there is a problem with your display, or broken data wire.
 
wpenner said:
Flashing the firmware does not write anything to the VCLD5. Fairly confident there is a problem with your display, or broken data wire.

I know we don't write to the VLCD5. What I meant was the feeling of weird coincidence. I run the marcoq-firmware for months without a problem, and the new version, on my first day, the display goes broken. Just wondering if it's possible to send weird packets of speed/data and actually put the display in this state. I think the 9999odo implies the fault happened in the communication, not originating from the display.

Of course, it can all be just a coincidence, and my display would have broken down this exact day even if I ran stock all time.
 
flufferty said:
wpenner said:
Flashing the firmware does not write anything to the VCLD5. Fairly confident there is a problem with your display, or broken data wire.

I know we don't write to the VLCD5. What I meant was the feeling of weird coincidence. I run the marcoq-firmware for months without a problem, and the new version, on my first day, the display goes broken. Just wondering if it's possible to send weird packets of speed/data and actually put the display in this state. I think the 9999odo implies the fault happened in the communication, not originating from the display.

Of course, it can all be just a coincidence, and my display would have broken down this exact day even if I ran stock all time.
Do you use three o four cables to update the firmware? It's possible that you corrupt the firmware with only 3 cables....

Enviado desde mi SM-G510 mediante Tapatalk

 
ashrambo said:
Do you use three o four cables to update the firmware? It's possible that you corrupt the firmware with only 3 cables....

I'm yet to make up a cable but is the 4th one needed VCC/5V ? The wiki says it's optional so in what situation could it corrupt the firmware?

Here is the list of wires: The connections to STLinkV2 clone:

purple wire connect to STLinkV2 RST pin (Potentially optional, did not appear to be needed flashing from Windows)
orange wire connect to STLinkV2 GND pin
black wire connect to STLinkV2 SWIM pin
brown wire connect to STLinkV2 VCC/5V pin (Optional - For those who want to flash without a battery attached)
 
Hi there,

so i flashed my Motor yesterday with the new Firmware from Marcoq and everything works as designed, but

1) i miss a bit more Support in low cadence mode and
2) the support comes a bit late and it needs something much muscle power during start up

which values i have to change to get this realized?
 
ashrambo said:
Do you use three o four cables to update the firmware? It's possible that you corrupt the firmware with only 3 cables....

So it's confirmed by others with corrupt firmware that garbage data can brick the VLCD5?
 
Using Marco version ( TSDZ2_Controller_vM0.19.A_and_TSDZ2_Configurator_Beta_0.3.6_con_codice_0.19_Stable ) at the moment.

One bug I have found is : If I use the throttle the peddle assist does not work any more. The throttle still works but I have to perform a power cycle to get peddle assist back.
 
cliftonarms said:
Using Marco version ( TSDZ2_Controller_vM0.19.A_and_TSDZ2_Configurator_Beta_0.3.6_con_codice_0.19_Stable ) at the moment.

One bug I have found is : If I use the throttle the peddle assist does not work any more. The throttle still works but I have to perform a power cycle to get peddle assist back.

In the next release I will try to fix the bug! :thumb: :thumb:
 
flufferty said:
ashrambo said:
Do you use three o four cables to update the firmware? It's possible that you corrupt the firmware with only 3 cables....

So it's confirmed by others with corrupt firmware that garbage data can brick the VLCD5?

No, i have not seen any other reports of this.
 
Hi guys, tell me please where I can read any useful information about working with VLCD5 with FOS.

For example, how to switch between modes, and maybe I have to set something important in the VLCD5 menu, not just in configurator?

I found how to enter settings menu, hold down the buttons "i" + on/off for 4 seconds and then quickly press "i" 4 times. But all the things there are't pretty clear. Should I set something up there?

The usual pedal assistant worked yesterday, but not today. Maybe it should go into some particular mode?

The motor is connected to the AC / DS 48V 5A power supply. The motor is just staying on the table (for tests, my battery is not ready yet). Perhaps the effort that I create with my hand is very weak and i need to choose the right mode for the assistant to work. Yesterday I explored the menu, and somehow the assistant began to work, and today it is gone. But the walk-assist continues to work normally.

Also in some cases ( when i click on/off few times ) when i incrase assist levels i have E05, E03, E01 errors at every assist incrase click.

PS.

st-link Cannot communicate with the device ( for people who will search for this problem )

I got this error and spend all the day for trying to fix it. Finnaly found out that my st-link v2 programator had some bug, changed it to new one and it started to work immediatelly, also want to notice that my cable was pretty short from start.
 
TSDZ2_Controller_vM0.19.A_and_TSDZ2_Configurator_Beta_0.3.6_con_codice_0.19_Stable

instaled on one of my bike ( other two use : casainho 0.16 and marcoq first release without configurator 0.16 )
as the other work fine exepted backward i would like try it first before improving the one using marcoq 0.16

backward resistance is gone as expected,
found there's too much time for the motor to start when pedaling ( about a full pedaling rotation )

as i changed some parameters may be an issue from me !?

if no, is there a parameter i can modify to have the assist starting quickly ?

see picture for the parameters i changed

configurator V1.jpg

Ooh, and a thing ... i don t use speed sensor, soo may be must change " boost at speed 0 " to " boost at cadence 0 "
may be the reason the issue i mentionned !?
 
brobro said:
Hi guys, tell me please where I can read any useful information about working with VLCD5 with FOS.

For example, how to switch between modes, and maybe I have to set something important in the VLCD5 menu, not just in configurator?

I found how to enter settings menu, hold down the buttons "i" + on/off for 4 seconds and then quickly press "i" 4 times. But all the things there are't pretty clear. Should I set something up there?

The usual pedal assistant worked yesterday, but not today. Maybe it should go into some particular mode?

The motor is connected to the AC / DS 48V 5A power supply. The motor is just staying on the table (for tests, my battery is not ready yet). Perhaps the effort that I create with my hand is very weak and i need to choose the right mode for the assistant to work. Yesterday I explored the menu, and somehow the assistant began to work, and today it is gone. But the walk-assist continues to work normally.

Also in some cases ( when i click on/off few times ) when i incrase assist levels i have E05, E03, E01 errors at every assist incrase click.

PS.

st-link Cannot communicate with the device ( for people who will search for this problem )

I got this error and spend all the day for trying to fix it. Finnaly found out that my st-link v2 programator had some bug, changed it to new one and it started to work immediatelly, also want to notice that my cable was pretty short from start.

Here’s some nice writing about it.
I used Google translator...

https://ashthinkings.blogspot.com/2019/06/kit-motor-electrico-parte-4-marcoq.html
 
marcoq said:
Translated by Google Translator:

Hi guys.
With this publication I officially released version 3.6 of the Java configurator and version M0.19.A for the controller board firmware.
What has been done:
- In the last release released, the eMTB mode started from TOUR assistance (up to TURBO), now with the configurator it is possible to define the starting assistance level "eMTB Start Assist Level", for example you can set it to 10, 20, 50, etc ... but must not exceed the level of SPORT assistance.
- I introduced the parameter eMTB Motor Pull Factor, which can be set from 0.0 to 1.0, this determines "the pull of the motor", or if I put a low value I get that the motor reaches the highest assistance values ​​with "less effort", if I put a low value, I'll have to "commit" a little more with the pedals.
- In the project folder there is an eMTB.xls spreadsheet that can be used to do some "paper" tests.
- I deleted the check to enable the "Last Beta Code" ... now the firmware has been aligned with the stable 19.0 version released by Casainho, of course all the customizations and implementations entered by the undersigned remain.
- I entered the VLCD5 Battery SOC 4 Bars checkbox which allows VLCD5 owners to implement the 4-bar battery status display as on the VLCD6. I did this because it would be a problem to manage the VLCD5 with the accuracy as in VLCD6 but with 6 notches ... I should add some fields in the Java interface .. but now it's full ... and it would be too expensive in time to move all. This solution should solve the problem of "misaligned" display with battery status.
- I have solved some small bugs ... and optimized the startup of the CPU.

On Endless they are discussing the "human power" factor, the calibration of the torsion sensor, and how to obtain a more accurate reading ... furthermore the developers want to release a firmware version that also includes the eMTB mode.
The big advantage of the Java configurator is that you can "experiment" with various settings without having to go through the code ... so do your tests !!!
Good rides to everyone !!!

the attached file ( archive ) corrupted, controler profile in /scr, missed
correct archive found in italian forum
 
Tsdz2Canyon said:
Here’s some nice writing about it.
I used Google translator...

https://ashthinkings.blogspot.com/2019/06/kit-motor-electrico-parte-4-marcoq.html

Thanks a lot Sir! This is exactly what i looked for! Very helpful :flame:
 
elem said:
the attached file ( archive ) corrupted, controler profile in /scr, missed
correct archive found in italian forum

Could you post a link to that please? I wonder why the GitHub repo is no longer being updated.
 
famichiki said:
elem said:
the attached file ( archive ) corrupted, controler profile in /scr, missed
correct archive found in italian forum

Could you post a link to that please? I wonder why the GitHub repo is no longer being updated.

http://www.jobike.it/Public/data/marcoq/2019710183424_TSDZ2_Controller_vM0.19.A_and_TSDZ2_Configurator_Beta_0.3.6_con_codice_0.19_Stable.zip

don t know why marcoq no long update his own github ?

many informations on italian forum, could be amazing if someone make a " compile "

here is his post on the forum ( using translator ) .

Hi guys.
With this publication I officially release version 3.6 of the Java configurator and version M0.19.A for the controller board firmware.
What has been done:
- In the last release released, the eMTB mode started from TOUR assistance (up to TURBO), now with the configurator it is possible to define the starting assistance level 'eMTB Start Assist Level', for example you can set it to 10, 20, 50, etc ... but must not exceed the level of SPORT assistance.
- I introduced the parameter eMTB Motor Pull Factor, which can be set from 0.0 to 1.0, this determines 'the pull of the motor', or if I put a low value I get that the motor reaches the highest assistance values ​​with 'less effort', if I put a low value, I'll have to 'commit' a little more with the pedals.
- In the project folder there is an eMTB.xls spreadsheet that can be used to do some 'paper' tests.
- I deleted the check to enable the 'Last Beta Code' ... now the firmware has been aligned with the stable 19.0 version released by Casainho, of course all the customizations and implementations entered by the undersigned remain.
- I entered the VLCD5 Battery SOC 4 Bars checkbox which allows VLCD5 owners to implement the 4-bar battery status display as on the VLCD6. I did this because it would be a problem to manage the VLCD5 with the accuracy as in VLCD6 but with 6 notches ... I should add some fields in the Java interface .. but now it's full ... and it would be too expensive in time to move all. This solution should solve the problem of 'misaligned' display with battery status.
- I have solved some small bugs ... and optimized the startup of the CPU.

On Endless they are discussing the 'human power' factor, the calibration of the torsion sensor, and how to obtain a more accurate reading ... furthermore the developers want to release a firmware version that also includes the eMTB mode.
The big advantage of the Java configurator is that you can 'experiment' with various settings without having to go through the code ... so do your tests !!!
Good rides to everyone !!!


and the link to the forum where last version wha commited

http://www.jobike.it/forum/topic.asp?TOPIC_ID=76426&whichpage=38

probably other interested informations before this page and after ....
 
I wonder is the reason he does not update his repo due to the copyright and usage warning he now has in his configurator? There seems to be an issue around the closing off of the software and Casainho, who I believe originally started the open source project wants it to remain open and free.
 
Minatauro said:
I wonder is the reason he does not update his repo due to the copyright and usage warning he now has in his configurator? There seems to be an issue around the closing off of the software and Casainho, who I believe originally started the open source project wants it to remain open and free.

probably, but marcoq code is based on casainho/buba one ...
and configurator use also stancecoke code ...

even with the warning, marcoq configurator is free for personnal users .

But marcoq have is own reason, thancks to him for his improvment of casainho code, it s good to use oem display with open source firmware .

began to resolving my issue, it s better if i change soft start from 20 to 5,
will try other option, must learn what to do :lol:soft start 02.jpg
 
Of course, its fantastic that we get the benefit of the hard work of all the developers and testers and hopefully it remains open for all.
Yes the lower value for the soft start timeout is perfect. I have copied Chri 27.5 settings that he put on Jobike forum with the Soft start timeout set to 1 and k soft factor set to .150 and duty cycle ramp up also set to 1. These work great for me and I have a 48v motor.
2019712101854_versione%203.6.jpg
 
Ok, found too chris set up, wanted to test lower value before going to his own ;)

don t know what's k soft factor ? what it's supposed doing ?
will change it too if it's better !
 
Minatauro said:
don t know why marcoq no long update his own github ?

elem said:
But marcoq have is own reason, thanks to him for his improvment of casainho code

I think the simple reason is, that marcoq is not familiar with git. The usage of git is very confusing for beginners :).

Perhaps someone should show him the few steps, how to use the git plugin directly in eclipse :idea:
The usage is much easier than the UserGuide suggests.

regards
stancecoke
 
Back
Top