TSDZ2 OSF for all displays, VLCD5-VLCD6-XH18, LCD3, 860C-850C-SW102.

[Waynemarlow]

You are more advanced than I am. Because you have already done some tests with the mbrusa firmware. I’m just going to start testing.

Some people are unaware, that in certain situations, the "power delay issue" can be dangerous.

I became aware, in September 2020, of the mspider65 version. I never had much interest in testing it. Before moving on to any wireless solution, first I want to make sure that the motor firmware works well, without major problems.

Thanks.

Azur what you are saying is very wrong and in some way I feel that your comments should be removed by the Moderator. The firmware is only dangerous in your perception, you have never even used the current version and yet you are willing to comment that the firmware can be dangerous.

You have a choice, move your comments to the V1.0 thread which you are talking about and don't use V1.0 if you feel you are unsafe. Revert back to the factory firmware, no one is explictly making you download the firmware.

Your comments are not valid, move to another thread of your own making please and let the rest of us congratulate MBrusa on his work.

 

[AZUR]

I have already withdrawn the statement.

 

[Waynemarlow]

I have already withdrawn the statement.

But why keep banging on about something which has no relavence to this thread, your comments are unwelcome and out of place.

 

[AZUR]

.
@mbrusa,

Until I start testing your firmware, I will not post any more on this forum.

In order for the other members of the forum to realize that the problem is important, I leave here the information that mspider65, put in the mspidr65 forum.

Sorry mbrusa, if I disturbed your work,

Thanks

I think this will probably be the last major version I release as I am satisfied with the result by now and I don't think there are any other major improvements to be made.
The season is improving and now I will use my time mainly to enjoy the bike.

I had forgotten that there was still the annoying problem of the power delay on restarts.
Here then a new version that solves, or at least significantly improves, the situation.

New v12 version of the controller firmware (only controller fw need to be updated):
https://github.com/TSDZ2-ESP32/TSDZ2-ESP32-Wiki/blob/master/bin/STM8/TSDZ2-v12.zip

For those interested in the changes made, just see the changes made with the last commit on github.

 

[HughF]

Azur, are you aware that each fork of the firmware has it's own thread? You have been quite good at avoiding my posts where I asked that...

 

[Elinx]

...
Until I start testing your firmware, I will not post any more on this forum....

You had to do this 25 posts ago, instead of spamming this thread.
I hope to see your feedback if you have fully tested v0.20.1C

 

[Peacepirate]

But you still don't understand.

You're just introducing noise.

Azur

So can I ask with all your persistence in repeatedly bringing the subject up that you personally think there is a problem, how you personally are going to solve the problem ? It’s obvious that you cannot code nor fully understand the electrical problems that you perceive.

If I was in MBrusa’s situation of some random guy repeatedly complaining on a thread that I had created and a problem, if at all a problem, was something that I was probably not to concerned about as very few others are concerned by, I would probably pack my coding skills up for the summer and go and enjoy my cycling Any further coding gains I can test with others on a more quiet basis without all the hassles of maintaining a thread on ES.

Thx Waynemarlow,
i could‘t describe better what you pointed out about this irritating guy...

 

[StormTD5]

I have tried diffrent mode on startup, weak assist on every mode...

.....
Calibrating the torque sensor
Hope its the torque sensor that is the trouble...

If the sensitivity of the torque sensor is low, a calibration could help a bit, but is not ideal.

Now I have made an calibration an with now weight the read was 23 and with 85kg rider the reading was 27. I'll guess that very low.
I have used the readings and no change. With stock FW it works.
I have noticed when I'm start the bike with not with full charge it can feel sluggish and with a couple of restarts it just wake up and work properly(stock FW)
The other bike I have is very fun to ride! (hybrid mode)

 

[Elinx]

....
Now I have made an calibration an with now weight the read was 23 and with 85kg rider the reading was 27. I'll guess that very low.
I have used the readings and no change......
The other bike I have is very fun to ride!..

I think your readings are about 230 and 270 (you must read the last digit too).
If this is the case, the range is only 40 and yes that is very low. (normal range is about 120 to 150)
But if you insert 230 as min. and 270 max. the tsdz2 should behave a little bit better.
For comparision you could measure the min.max. torque values of the other (good working) bike too.

...a couple of restarts it just wake up and work properly(stock FW)

You always must startup tsdz2 without feet on the pedals for setup the zero load of the torque sensor.

 

[StormTD5]

....
Now I have made an calibration an with now weight the read was 23 and with 85kg rider the reading was 27. I'll guess that very low.
I have used the readings and no change......
The other bike I have is very fun to ride!..

I think your readings are about 230 and 270 (you must read the last digit too).
If this is the case, the range is only 40 and yes that is very low. (normal range is about 120 to 150)
But if you insert 230 as min. and 270 max. the tsdz2 should behave a little bit better.
For comparision you could measure the min.max. torque values of the other (good working) bike too.

I blame old eyes :lol: OK I'll try that :thumb:

The good working bike was about 150 and max 260. I'll guess its little better. Have not used the calibration values yet.

 

[Elinx]

.....

I blame old eyes :lol: OK I'll try that ...

The good working bike was about 150 and max 260. I'll guess its little better. ...

A range of 110 isn't perfect, but as you had seen, remarkable better

FYI
It is possible to improve the torque values with hardware calibration, but you need some mechanical skills

 

[StormTD5]

.....

I blame old eyes :lol: OK I'll try that ...

The good working bike was about 150 and max 260. I'll guess its little better. ...

A range of 110 isn't perfect, but as you had seen, remarkable better

FYI
It is possible to improve the torque values with hardware calibration, but you need some mechanical skills

Hardware calibration, well I'll guess I may have to do that but I'll probably have to change the display first to a more easy to work with.
Which display can you recommend? I'm more a welder than soldering person but it's always fun to learn new skills 8)

 

[Elinx]

.....

Hardware calibration, well I'll guess I may have to do that but I'll probably have to change the display first to a more easy to work with.
Which display can you recommend?...

You are right that stock display's are not the easiest to handle if you want to change some setting, but with some practice it will work. I don't know about vlcd5, because I have XH18. (If you change stock for stock you haven't to solder.)

If you don't like this handling and you don't care about minimalist display, you have the choice for the latest colored 860C.
That one is also useable for Casainho's builds and the recent Blacklite build (based on mspider65 v12).
Stock display you can use with mspider65 versions too. Handling settings, wireless with a smartphone.
I don't recomment the KT LCD3, because this display is a bit isolated. You only can use this for v020 builds.

If you want minimalist display (like vlcd6 you have now) you can choose also SW102, but for first time flash you have to break the case open. If that is no problem for you, SW102 has same benefits as 860C with regard to other builds.

 

[mallesepp]

The 860C display is the best for me. For view as well as settings

 

[casainho]

The 860C display is the best for me. For view as well as settings

860C and SW102 are the most advanced displays, see here a table comparison:

https://github.com/OpenSourceEBike/TSDZ2_wiki/wiki/Displays-comparison

 

[JohnAnanda]

The SW102 is a nice option if you want a small thing.
The developers did an excellent work for the user interface with such a tiny screen.
The main drawback for me, besides having to open a case that is not meant to be opened, is that it lacks luminosity in daylight and is difficult to see sometimes.
The buttons also lack tactile feedback when pressed, but this can be ameliorated when the case is opened for flashing.

 

[casainho]

The SW102 is a nice option if you want a small thing.
The developers did an excellent work for the user interface with such a tiny screen.
The main drawback for me, besides having to open a case that is not meant to be opened, is that it lacks luminosity in daylight and is difficult to see sometimes.
The buttons also lack tactile feedback when pressed, but this can be ameliorated when the case is opened for flashing.

And SW102 has Bluetooth, that if other developers are willing to develop, we could make it working with our mobile app. The TSDZ2 wireless board uses the same microcontroller as SW102 so would be a matter to copy the firmware I developed for the wireless board.

 

[JohnAnanda]

Speaking about the SW102, I noticed it reads voltage 0.3V higher than all my other measuring tools.
I haven't seen an option to correct it, like in the version for original displays.
Can I just increase the reset voltage and the low cut-off voltage by 0.3V in order to have a correct estimation of SOC ?

 

[mbrusa]

Of course that's what you have to do.
With the stock displays there is a correction option because the controller error is added to the conversion error of the data sent to the display.

 

[mbrusa]

@mctubster, I realized that you asked questions that I have not answered.

Some feedback for TSDZ2-v20.1C

I have been running the firmware with stock settings, only changes originally to config were setting 36v battery and capacity, street mode.

I am using Hybrid mode on a heavy Cargo bike. The responsiveness with default settings is an improvement over V1.1, moderately faster to detect and start power and also less overrun. Startup using torque mode is perfect.

I just change one setting - Motor acceleration from the default of 5 (after installing the firmware) to 35 (as recommended for 36V motor and 36V battery)

Wow, much more responsive. I assume it has increased the current ramp only?

Yes, increasing "Motor acceleration" increases the current ramp. You can try with values greater than those recommended, gradually and carefully.

I don't really have an issue with overrun personally because of the weight of the bike. In the late 2020 firmware for Bosch gen 4 motors they deliberately introduced some overrun (potentially only in emtb mode) and I'm sure they are doing high speed calculations - but essentially they will overrun for a short period of time and I assume have a much faster current ramp to cover a mountain biker pausing pedaling briefly to avoid a pedal strike.

I don't believe this is happening but could be a suggestion for a future improvement - have a faster current ramp if the bike is over a certain speed - that is if someone is already moving over a threshold speed ramp the current faster so the motor can catchup and start applying power faster. The Power only mode I believe has the boost option doing this in a way?

Within the predefined limits this is already the case, the current ramp is calculated as a function of the speed, the higher the speed the shorter the ramp.
Startup boost does not intervene directly on the ramp, but on the adc value of the torque sensor, it is proportional to the thrust on the pedals, so with the same parameters the effect can be controlled with the thrust.

Also wanted to mention I also reproduced the battery voltage replacing the time bug (860c). The setting does not appear to save, so on next power on is back to displaying the time.

Anyway loving the new firmware.

I saw this bug, already fixed.
I needed to reduce the number of variables in eeprom, so all bool variables are saved in bit instead of byte. I had entered this which is not bool, it happens ...

 

[casainho]

I saw this bug, already fixed.
I needed to reduce the number of variables in eeprom, so all bool variables are saved in bit instead of byte. I had entered this which is not bool, it happens ...

mbrusa, would you like to support the TSDZ2 wireless board? the wireless board acts like a display, the firmware is similar to the one of SW102 display, with the advantage that it connects already to our mobile app. And the only other new features are ANT+ LEV for wireless connection of cycling GPS displays and the wireless remote. There is also support for connection to a regular wired remote, for the ones who do not like to use the wireless remote. I don´t have free time to keep developing the motor firmware, wireless firmware, displays firmware, mobile app and GPS displays apps!

860C and SW102 firmware data fields / structure were mainly developed by other developer that is not active anymore on this project - all this firmwares and softwares were a BIG work collaboration from different developers, some that are not active anymore on this project.

 

[Mr.Flibble]

Thanks for all your hard work mbrusa.

It feels like halfway between 1.1 and the original firmware - a little more wooden than 1.1 and not quite so natural at high cadence, but a nice compromise.
The over run is still there (but greatly reduced compared to 1.1) but it winds down rather than going full belt, making it easier to change gear without wreaking my drive train or grinding to a halt.
Ghost peddling is non existent in Emtb mode

I haven't got to a really steep hill yet, but the support while climbing seems much better at low cadence than 1.1.

The only problem I have come across is, that twice in hybrid mode the bike has gone off by its self and not stopped (even when I lifted the back wheel off the ground) until I turned the bike off
My guess would be that the crappy cheap Chinese clutch thing in the motor drags, as the peddles move slightly by themselves when the rear wheel rotates. This sets the cadence sensor off, which makes the motor go, which turns the wheel, which makes the peddles turn, which sets the cadence sensor off, which ......

Thanks again.

 

[StormTD5]

.....

Hardware calibration, well I'll guess I may have to do that but I'll probably have to change the display first to a more easy to work with.
Which display can you recommend?...

You are right that stock display's are not the easiest to handle if you want to change some setting, but with some practice it will work. I don't know about vlcd5, because I have XH18. (If you change stock for stock you haven't to solder.).....

I'll stick to the stock display, have three bikes with tongsheng motors. I just have to learn how to work with the display
I did Flash my dougters coaster bike, and Wow! The acceleration just at level 2! And at level 4 :lol: :lol: so fun!!
Totally different bike
Thank you all who made this possible! What an effort!!

 

[Anikea]

Torque Sensor hardware calibration
After driving a few thousand km with the last firmware from mbrusa, I recently got another TSDZ2. This one behaved very differently compared to my other one. If there was little human power, the engine provided immediate support and was far too strong for me, but with more human power, there was hardly any additional engine output.

The software calibration of the torque sensor gave me 166/245 (just 79 difference). That's why I took my time today and tried out the hardware calibration. Although I had read several times that this is actually not recommended. Or maybe that was the case because I had read that in a post by Cassainho and in his Firmware the sensor is not only calibrated with the beginning and end value, but also the curve shape is taken into account.

Before that I had documented the curve shape of my sensor with a few values ​​and it was very unfavorable. As my subjective driving experience, the curve showed a strong increase at the beginning and then only a little increase. Far from linearity.

The calibration itselfe was less complex than I had planned because I saved the motor dismantle and disconnect the sensor from the controller. I then set the sensor to an initial value of about 70 and reassembled everything . After the installation I got an E02 error and thought crap . But that was probably only because my initial value was now so far away from the programmed value that the controller had assumed a defective sensor when initializing. After I reflashed the engine with the new values, everything worked fine.

The result already impressed me on the first lap. The response behavior in terms of personal performance has been much better. The Torque curve that I created after the calibration shows the difference. It is much more lienar than it was the case before.

The left graph is after the calibration, on the right it is before.

 

[Elinx]

.... have three bikes with tongsheng motors. ....
...Wow! The acceleration just at level 2! .....

Curious about the torque range of this third bike.

Torque Sensor hardware calibration....

The left graph is after the calibration, on the right it is before.

From a torque range 79 to 162 is a good result.
Also the improvement from the lineair part, from about 10 to 30 kg should help a lot.