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

eyebyesickle said:

Making progress, hoping to hear good news by next week. I have another route to go if I don't get permission by then, but things appear to be working out.

You are currently using the 850c system for yourself, correct? Any new previews of the display screen to show off? I am trying not to crack open my 850C as is!!! Im too OCD for that mess :lol:

Click to expand...

eyebyesickle said:

Any ETA casainho? :wink:

Click to expand...

eyebyesickle said:

FYI - the newest version with imperial unit/throttle/etc support IS NOT AVAILABLE YET for the bootloader, but is expected shortly. The bootloader DOES NOT USE THE SAME .bin FILE THAT YOU USE TO MANUALLY UPDATE!!!

Any ETA casainho? :wink:

Click to expand...

casainho said:

eyebyesickle said:

Any ETA casainho? :wink:

Click to expand...

Today I hope.

So, the wiki now needs a lot of updates, I hope to do it on next days. I will link to your site, to your documentation. Also it is very important to tell users the shops that sells the displays with the cables and/or already with firmware flash and that box, etc so I will also put the links to it.

I will keep on wiki the previous way to open and install the firmware as it is needed for development bit the first choice will obviously be this of using original bootloader.

Great work!!!

Click to expand...

buba said:

vadda said:

On the A5 I found that the voltage reported by the LCD3 compared to that
measured with the Tester is 0.4V lower.

Click to expand...

But in the end, the TSDZ2 will always be sensitive to heating. And going on long climbs is the worst thing you can do due to the combination of rolling friction, aerodynamic drag and the climb itself: you need to convert the energy in your battery to potential energy when climbing. And the faster you go the faster the conversion needs to happen and the more power you are pulling from the battery through the motor. And there is no way escaping that you need the power on those climbs. Which is a product of torque and angular velocity -> Torque means current -> Current means increased motor temperature.

So in short, I think it can be improved in the future. For now:

- Do not use the Experimental Motor modes as I think they are not as efficient
- Use a high nominal voltage battery. Your motor will be able to rotate faster in a lower gear and you could have the same power with less current but faster angular velocity. There is a small gain with higher voltages due to less transmission losses in cables and so on.
- Be as light as possible when climbing as this drastically reduces the power needed
- Maintain a low rolling resistance on the bike and keep drag at a minimum
- Reduce speed slightly as this also makes a big difference

All these suggestions have little effect on their own. But combined you might notice a small difference!

Click to expand...

geeksville said:

btw eyebyesickle,

I think the "open source" aspect of this software could be a nice marketing tool for you (and even for the Chinese manufacturers you are working with). Because it can address doubts customers might have about buying from you as a "small" manufacturer. With all of the code out there they don't need to worry about support if you were to go away. And for TongSheng it gives them a differentiator vs the the (currently more popular) Bafang ecosystem.

Also possibly useful for positive hype if you ever do something like a kickstarter. (Because you can point backers at code that exists right now and is already working on vehicles. And the openness of it is nice for some people)

And if you do talk the SF102 folks into doing a semi-custom build you might ask them if they can stuff the nRF51422 instead of the nRF51822. Because it seems to be the same CPU but they just paid the licensing fees for ANT protocol. Which would make it possible for the SF102 to talk to garmin GPSes (for super seamless uploading of cadence and human pedal power). I bet the COGS difference is less than $0.50 per unit.

Just my 2 c.

Click to expand...

eyebyesickle said:

geeksville said:

btw eyebyesickle,

I think the "open source" aspect of this software could be a nice marketing tool for you (and even for the Chinese manufacturers you are working with). Because it can address doubts customers might have about buying from you as a "small" manufacturer. With all of the code out there they don't need to worry about support if you were to go away. And for TongSheng it gives them a differentiator vs the the (currently more popular) Bafang ecosystem.

Also possibly useful for positive hype if you ever do something like a kickstarter. (Because you can point backers at code that exists right now and is already working on vehicles. And the openness of it is nice for some people)

And if you do talk the SF102 folks into doing a semi-custom build you might ask them if they can stuff the nRF51422 instead of the nRF51822. Because it seems to be the same CPU but they just paid the licensing fees for ANT protocol. Which would make it possible for the SF102 to talk to garmin GPSes (for super seamless uploading of cadence and human pedal power). I bet the COGS difference is less than $0.50 per unit.

Just my 2 c.

Click to expand...

Makes sense... that isn't something I would really be considering unless the OSF developers and everyone involved wanted in... I do this mostly because I like it... although lately it seems things have developed a bit more - I actually value this whole process because it is forcing me to learn new things as well!

I will see what we can do about the nRF51422 instead of the nRF51822- that sounds possible!

Thanks for all your hard work!

Click to expand...

elem said:

buba said:

elem said:

hello, have an issue with V0.20, suspected ancient value not cleared even if i writed all 00 before programming .

Click to expand...

Okay, everything is working as expected I just need to explain a couple of things

When you install the 0.20.0 it will load the default values, it will not be all zeros. Instead, the default values will be loaded and you can change those to whatever you want.

elem said:

ok, motor stop at about 250w, like if street mode was activate, was testing torque mode, changed to power mode, it s the same .

Click to expand...

You need to set the motor power limit. Disregard the Street Mode and disable it for now. Just set the power limit to what you want it to be. It is in the Basic Setup sub menu.

elem said:

trying to activate street mode with different value ( 35 km/h and 450 w max ) it s the same
trying to reset to defaut value, was impossible, a value of 42 was there and impossible to change ...

Click to expand...

Disable the Street Mode for now so you get comfortable with configuring the normal system. And later on it is no problem at all to enable Street Mode.

Please follow the instructions on the wiki on how to reset to defualt values. The system will not power down as it did previously. It will just smoothly set all values to default.

elem said:

well, with my short test it s seem the start delay noticied in V 0.19 as gone

Click to expand...

I am glad to hear that! For MTB riding, Startup assist without pedal rotation is very good if you properly configure that. But maybe wait until you get comfortable with the 0.20.0.

elem said:

also tested in my garage the torque on pedal, it s different, you probably modified something

in menu 11 ( previously 9 in v 0.19 )
2 torque sensor ADC : static 217, left and right with my weight ( 64 kg ) 340, dynamic ( pedaling ) hiting 320 wich is less than static and very different than V 0.19
3 pedal cadence hitting 125 max
4 PWN duty cycle : from 80 to 120( about )
5 ERPS hitting 360 ( about )

okay, i ll try to clear controler memory and reprogramm it .

in case of problm, is it possible to use V0.20.7 for LCD3 and for example V0.20.6 for controler ?

Click to expand...

You can clear it but it will load the default values again. I think you thought the old values were there. I am pretty sure there are no bugs so you can use the Alpha 7 on both the motor controller and display.

So take a look at the wiki and I assure you everything will work!

https://github.com/OpenSource-EBike-firmware/TSDZ2_wiki/wiki/0.20.0-(DEVELOPMENT)-%7C-KT-LCD3-%7C-TSDZ2-%7C-Manual

Click to expand...

Ok, i was using V0.20 alpha 7 , the one for LCD3 and the one for TSDZ2, i know the standard process .
fact, it's seem the default value you used are good
had only to change max power to 550 w ( my standard set up ), nomber of cells ( 12 ), max curent ( 15 )
eMTB configured to 3 ( instead of 2 from default value )

If you said alpha 7 work it s okay i thinck it s true

why i thinck there is some value from previous install ( v 0.19 ) is because when i came to power mode menu i found the same value i used in v 0.19 and i m pretty sure it s not your default value ( 0.3/0.6/0.9/1.5/2.1 this was my set up for different lvl in v 0.19 )

Click to expand...

elem said:

i just tried to clear my controleur ( may be not necessary, as i remenber you said it was the last time needed ! ) and reprogram it
sadly i have the same problem, and the same value for " reset default value " = 42, is it a timer to the reset ? haven t see it decrementing !?

Edit : sorry found the reason : Reset to default values. Enter the sub menu so that the number 42 starts blinking. Hold the POWER button a short amount of time and the system will reset to default values.

soo many things to learn

and also my own value from my previous install .

soo i thinck that it s some stored value in my LCD3 and i need to reprogramming it twice !?

sorry to disturb buba, all seems good when i was programming, no errors, nothing special, and i m using often my st link because i m also using marcoq code witch need to reprogram each time you want to improve a set up .

Click to expand...

jimmyfergus said:

I'm sad to I can't try out 0.20, because before I got it flashed I cracked my LCD3 display (so I'm awaiting a new display. Worse I ordered a SW102, which isn't yet ported so I'll either have to wait or try to port myself).

In the meantime, when I first heard of the eMTB exponential multiplier I thought it was applied on top of Torque or Power proportional. I think I was wrong about that though - it's it's own mode.

However, couldn't / shouldn't it be just an filter on the other modes? It could just be a factor applied to the result (or the torque input). Assuming the loop can afford another multiplication operation? That way I could adjust assist level for how much I feel like contributing, but still get the responsive adaptive nature of the exponential curve.

Thanks as always for all the work devs - I'm loving what you're doing.

Click to expand...

btslo said:

How do you "wipe" before upgrading? I just flashed over 0.19 and it seems to be working okay, had to setup everything again.

Click to expand...

elem said:

yes buba you are right, i ihanck it was my own value but was littely different .

Soo, seem ok, all run right, gone back to V0.19 and will go ti V0.20 later after testing all the improvment in V0.19, in fact too many things happenned in the same time, changing my engine to a 48V version, changing my metal gear to a new nylon one because the sound of the metal gear is stressing me , moving faster from V0.19 ( only 3/4 ride with 36V motor in exp mode ) to 0.20, only one ride with V0.19 in standard mode ( 0) with the 48V engine, and going to V0.20 directly with activating exp mode witch was only tested with lvl 1 in my garage .

Soo all seem ok, i haven t any experience with 48V engine, it s different than 36V motor

i liked to be able to use exprimentale mode with 36v motor because my natural pedal frequency is about 70 rpm and it was cool to be able run the 36v at 95 rpm or more witch i usely hit with my harmonic/muscular ( without motor ) bike .

36v motor is good as it seem it accept easily experimental mode, 48v motor is good because the useable range of use ( human rpm ) is better, when it s a problem climbing in trial mode at 40 to 60 rpm with 36v motor it s easy with 48v motor .

the bad thing : i can t get the best of the two "world", the good thing, even if 48v can t reach the same rpm than 36v motor using experimental mode, it s just necessary to remenber that we are " assisted " and we could get the best of the motor learning how use it in the best way .

well, it seem V0.19 it s a very good release in fact, very stable and also the latency at start up may be the wrong way to preserve the nylon gear, will post as soon as i ll move to V0.20 and thancks all to the hard way done,
i can t make donation to all invloved in the devellopment but you casainho have done a great job, merci beaucoup .

Click to expand...

elem said:

sorry buba but i thinck V0.20 isn t in cause, it s probably a technical issue an only "circonstancial" ( in french : un concours de circonstance, don t know in english !? )

and, last thing, why i suspect a partialy demagnetised, because even if i was abble to hit power ( from battery ) it s seem my motor don t give the same power for the same use of the wattage ( what a hard job to translate french speacking to english language, sorry .... )

have fun, carpe diem

Click to expand...

eetulaa said:

buba said:

vadda said:

On the A5 I found that the voltage reported by the LCD3 compared to that
measured with the Tester is 0.4V lower.

Click to expand...

But in the end, the TSDZ2 will always be sensitive to heating. And going on long climbs is the worst thing you can do due to the combination of rolling friction, aerodynamic drag and the climb itself: you need to convert the energy in your battery to potential energy when climbing. And the faster you go the faster the conversion needs to happen and the more power you are pulling from the battery through the motor. And there is no way escaping that you need the power on those climbs. Which is a product of torque and angular velocity -> Torque means current -> Current means increased motor temperature.

So in short, I think it can be improved in the future. For now:

- Do not use the Experimental Motor modes as I think they are not as efficient
- Use a high nominal voltage battery. Your motor will be able to rotate faster in a lower gear and you could have the same power with less current but faster angular velocity. There is a small gain with higher voltages due to less transmission losses in cables and so on.
- Be as light as possible when climbing as this drastically reduces the power needed
- Maintain a low rolling resistance on the bike and keep drag at a minimum
- Reduce speed slightly as this also makes a big difference

All these suggestions have little effect on their own. But combined you might notice a small difference!

Click to expand...

Does the experimental high cadence mode affect the way the motor is controlled below the old cut off rpm?
I think the high cadence mode was one of the most welcome add compared to original firmware when used a 36v battery.

Most useful i've found it in steep climbs on trails. With it you can shift small enough gear before start to climb and still get to gain good speed to start to climb.

Click to expand...

thineight said:

Hello, hope you are enjoying your bike over these days.
Yesterday I had a 1000m climb with the v0.20alfa7.
I did not setup the advanced cadence yet, and I used the power mode because I did not have time to try out the rest of the options.

The overall feeling is very positive, very good one! :thumb: :thumb:

I have 4 levels plus eMTB and used pretty much the eMTB mode which, with the actual configuration (factor 3), could develop around 300-350w on the steepest part (20-25%). The 4th assistance could reach 400-450w which is the limit I set for my 36v motor.
As the eMTB mode is not then able to develop the full power under heavy load, I should try changing the eMTB parameter (from 3-->4) to increase the power, or leave as it is to keep a safe margin against overheat and, in case of need of extra power, switch to the max assistance to get the 450w for a limited period.

Even if I try myself to keep a safe margin against overheat, I will wait the instructions of the thermal pads installation to install it and finally improve the overall thermal performance. Next time I open it up for greasing I will add the thermal pads as well.

Click to expand...

thineight said:

Buba, yesterday I had to push the bike on the steepest rough slopes and I had to use the walk assist. Unfortunately it is not really usable on that condition due to the shakes, and you consequently loose the assist. We discussed this matter before, but I really think we should implement something to improve that behaviour otherwise it is nearly impossible to use the walk assist when you need it the most.
We discussed about a "denounce" filter or a always-on (not desired by casainho for safety reasons). Do you think we have room to improve that function a bit?
Maybe walk assist, once enabled, can be disabled after 1-2 second of no button pressed? I think it will be safe enough, what do you guys think?

Thanks and looking forward to see the new firmware-display improvements.

Click to expand...

casainho said:

So my list for a new version of SW102:
- ANT
- RTC clock just like 850C
- USB charger just like 850C (I travel in mountains and my phone, computer and 4G router, all of them charge on the 850C USB charger, this technology equipment are very important to me in this travels)
- more flash memory, maybe 512 kbytes (twice current value because having ANT we would need more space for more features, connect to other ANT sensors, etc)

Could be a quickstarter for presales of like 1000 units, $60 each unit for final user. Then with money of presales, you could ask the production in the Chinese company of that 1000 units.

Click to expand...

andrea_104kg said:

Today I spent an hour on the test of version 20, unfortunately alpha5 (I couldn't upgrade but I think it changes little)
Obviously every test is conditioned by a thousand factors
1) I live in a hilly area, so only descents and climbs, no plains
2) I am old and fat :-(, 55 years old, 173 cm tall and currently 101kg in weight
3) every torsimeter seems to be unfortunately different :-(
so my impressions must be taken for what they are, simple impressions.
torque mode: how the original software suffers from oscillations during pedaling. It is much better than the original software but the power is not delivered constantly during pedaling, it is probably more real, even the power delivered by the legs is not constant.
Emtb mode: amplifies the defect of the torque mode, that is the power oscillation is even bigger and I don't like it.
I have tried various levels of sensitivity but the result does not change.
I'm not worried at all because power mode is SIMPLY FANTASTIC.
With a x3 multiplier I get a perfect, wonderful, exceptional emtb level.
With the pressure of my legs I can obtain a very precise power progression, from less than 100w to over 800w (48v 750w motor) in an extremely linear and pleasant way. The power goes up gradually, sweetly, with an impressive ease of adjustment.
For me only this change is worth a nobel prize
So buba, I'll propose you in Stockholm
Thank you!

Click to expand...

buba said:

thineight said:

thineight said:

Buba, yesterday I had to push the bike on the steepest rough slopes and I had to use the walk assist. Unfortunately it is not really usable on that condition due to the shakes, and you consequently loose the assist. We discussed this matter before, but I really think we should implement something to improve that behaviour otherwise it is nearly impossible to use the walk assist when you need it the most.
We discussed about a "denounce" filter or a always-on (not desired by casainho for safety reasons). Do you think we have room to improve that function a bit?
Maybe walk assist, once enabled, can be disabled after 1-2 second of no button pressed? I think it will be safe enough, what do you guys think?

Thanks and looking forward to see the new firmware-display improvements.

Click to expand...

Will take a look and come up with something!

Click to expand...

+1 :thumb: I think this would be useful as well as its easy to have your fingers bounced of the button for a short time and loose all walk assist when you are needing it most.

Click to expand...

eyebyesickle said:

Forgive me for spamming across the TSDZ2 threads guys, but I am really excited about this, and really put a lot of work in. No no no, nothing compared to anyone programming etc... but it is all relative, and I am extremely challenged sometimes :lol: so... I see this as a major thing, bringing the availability of the OSF TSDZ2 to a much wider range of people who would not even be able to think about cracking a 850C open. Anyway, let me know if anything looks like it needs correcting etc - I hope everything is good, took me long enough

Click to expand...