TSDZ2 open source firmware only for KT-LCD3 (v0.19.0 / v.0.20.0beta1)

[mbrusa]

I didn't notice any difference in the start-up time after I shift still at 2 seconds or 2 revolution.

I'm sorry I expected an improvement ...
However, I continue to look for a solution.

I think the perfect setup would be that when you shift the motor stops completely like it does and then within a half a second it starts ramping up power and by the time you hit 1-2 seconds it's at full power again.

It would be perfect, the problem is that the brake sensors must also work correctly.

It could be me but it seems like with these last couple of hex files I have to shift early because power levels off at 70 RPMs and by the time you hit 80 it has half the power and by 90 it's almost zero.

Thanks for all your efforts!

The change does not affect the maximum cadence.
If you have the feeling of a different response, didn't you enable the motor in experimental high cadence mode before?

 

[jeff.page.rides]

I didn't notice any difference in the start-up time after I shift still at 2 seconds or 2 revolution.

I'm sorry I expected an improvement ...
However, I continue to look for a solution.

I think the perfect setup would be that when you shift the motor stops completely like it does and then within a half a second it starts ramping up power and by the time you hit 1-2 seconds it's at full power again.

It would be perfect, the problem is that the brake sensors must also work correctly.

It could be me but it seems like with these last couple of hex files I have to shift early because power levels off at 70 RPMs and by the time you hit 80 it has half the power and by 90 it's almost zero.

Thanks for all your efforts!

The change does not affect the maximum cadence.
If you have the feeling of a different response, didn't you enable the motor in experimental high cadence mode before?

I did try the high cadence mode a month or so ago and I did notice that I could pedal past 95 but that was the only thing that I noticed. I'll try it again and see if I notice more this time. For some reason it just seems that the power drops off really fast past 70 even though I can still pedal up to 95 there's no power left.
Keep me posted if you wanna try something else glad to test it.
Thanks

 

[jeff.page.rides]

I didn't notice any difference in the start-up time after I shift still at 2 seconds or 2 revolution.

I'm sorry I expected an improvement ...
However, I continue to look for a solution.

I think the perfect setup would be that when you shift the motor stops completely like it does and then within a half a second it starts ramping up power and by the time you hit 1-2 seconds it's at full power again.

It would be perfect, the problem is that the brake sensors must also work correctly.

It could be me but it seems like with these last couple of hex files I have to shift early because power levels off at 70 RPMs and by the time you hit 80 it has half the power and by 90 it's almost zero.

Thanks for all your efforts!

The change does not affect the maximum cadence.
If you have the feeling of a different response, didn't you enable the motor in experimental high cadence mode before?

I did try the high cadence mode a month or so ago and I did notice that I could pedal past 95 but that was the only thing that I noticed. I'll try it again and see if I notice more this time. For some reason, it just seems that the power drops off really fast past 70 even though I can still pedal up to 95 there's no power left.
Keep me posted if you want to try something else. I will test it.
Thanks

 

[casainho]

I would like to thanks to mbrusa for the development of the solution for the motor overrun. I just did a new firmware release that includes the solution and it is clearly a good one!! I remember Buba doing a different solution on V0.20 that would be the best when user did a calibration of the PAS sensor of TSDZ2 but this new solution is small in code and do not require any calibration, which makes the user life much easier!!

On this new version I also implemented the street mode as also the max power quick configuration, features that were available on V0.20 for some time ago (they were not a priority for me as I currently do not use them much).

A new nice feature is the battery pack internal resistance automatic calculation and this makes the battery SOC working with very good accuracy.

 

[mbrusa]

Thanks to you for the project that you have started and that you are carrying out with always interesting improvements.
I am happy that motor overrun solution can be useful to everyone.
I started to get interested in the firmware for fun, but then it became a challenge to myself.
With TSDZ2 I keep both body and mind in exercise ...

 

[jeff.page.rides]

If you have the feeling of a different response, didn't you enable the motor in experimental high cadence mode before?


I did try the high cadence mode a month or so ago and I did notice that I could pedal past 95 but that was the only thing that I noticed. I'll try it again and see if I notice more this time. For some reason, it just seems that the power drops off really fast past 70 even though I can still pedal up to 95 there's no power left.
Keep me posted if you want to try something else. I will test it.
Thanks

I tried 48 volt high cadence mode and I really liked it from 70 RPM to 90 RPM it pulls a lot better. And I can pedal up to 110 RPM

 

[mbrusa]

...
Keep me posted if you want to try something else. I will test it.
Thanks

I tried 48 volt high cadence mode and I really liked it from 70 RPM to 90 RPM it pulls a lot better. And I can pedal up to 110 RPM

This is not a solution to the gearsensor problem but it can improve.
If it gets worse or you have other problems, you can always go back.
With LCD3, try setting "Startup assist without pedal rotation" with a value between 10 and 30.
It is for immediate assistance at the start, but also at restart after gearsensor stop.

 

[jeff.page.rides]

...
Keep me posted if you want to try something else. I will test it.
Thanks

I tried 48 volt high cadence mode and I really liked it from 70 RPM to 90 RPM it pulls a lot better. And I can pedal up to 110 RPM

This is not a solution to the gearsensor problem but it can improve.
If it gets worse or you have other problems, you can always go back.
With LCD3, try setting "Startup assist without pedal rotation" with a value between 10 and 30.
It is for immediate assistance at the start, but also at restart after gearsensor stop.

last year I used startup assist without pedal rotation and it worked great on dirt trails, but I haven't been on dirt trails for 4 months. It didn't seem like it has worked for a while but I thought maybe I didn't have it turned on. So I looked through my settings today and found I did have it turned on and it was set at 35% but it doesn't work. It did work in beta 20 before we made all the other changes. So in the last changes that were made it stopped power assist without pedal rotation from working.
Can we find out what the problem is and get it working again it would be great before I start riding in the dirt again?
I was also thinking if it was possible to turn up the sensitivity to EMTB past 20 I might not have to preset my torque sensor by pulling back on the pedals when I start it?

 

[mbrusa]

Really "startup assist without pedal rotation" no longer works?
I didn't notice, now I check.
For eMTB I have to understand better.
Do you have a coaster brake motor?
At power on, do you have to pull the pedals backwards?
Sensitivity 20 is already very high, you probably don't use the whole range.
Keep in mind that eMTB assistance is limited by the maximum current or maximum power values set.

 

[jeff.page.rides]

Really "startup assist without pedal rotation" no longer works?

Correct it doesn't work anymore.

I didn't notice, now I check.
For eMTB, I have to understand better.
Do you have a coaster brake motor?

Yes, I have a coaster brake motor.

At power-on, do you have to pull the pedals backward?

Yes, I pulled backward on the pedals when I turn on the system.

Sensitivity 20 is already very high, you probably don't use the whole range.
Keep in mind that eMTB assistance is limited by the maximum current or maximum power values set.

I set emtb at 20, max watts at 700w, max amp at 17a. The Max watts it shows as I ride is 680w
Thanks

 

[mbrusa]

Really "startup assist without pedal rotation" no longer works?

Correct it doesn't work anymore.

I fixed the "startup assist without pedal rotation" problem, it actually didn't work anymore.

Sensitivity 20 is already very high, you probably don't use the whole range.
Keep in mind that eMTB assistance is limited by the maximum current or maximum power values set.

I set emtb at 20, max watts at 700w, max amp at 17a. The Max watts it shows as I ride is 680w
Thanks

Increasing the sensitivity does not increase the power, you are already at the maximum, but you would get there with less effort.
Is that what you want?
I could eliminate sensitivities 1 and 2, move the others by 2 positions and add two new sensitivities, 19 and 20.
Can it go well?

 

[Waynemarlow]

I set emtb at 20, max watts at 700w, max amp at 17a. The Max watts it shows as I ride is 680w
Thanks

Jeff are you getting into over heating problems, we have a number of climbs here which require a steady 550W or so and the temp of the unit really climbs in that time ? or have you modified the motor for better cooling ?

 

[mbrusa]

No, if there are heating problems, they remain!
They can be solved only by limiting the power or by installing the temperature sensor or with other cooling systems.
The question is whether he wants to have the current maximum power with less effort. Obviously the heating problems increase.

 

[jeff.page.rides]

Really "startup assist without pedal rotation" no longer works?

Correct it doesn't work anymore.

I fixed the "startup assist without pedal rotation" problem, it actually didn't work anymore.

Sensitivity 20 is already very high, you probably don't use the whole range.
Keep in mind that eMTB assistance is limited by the maximum current or maximum power values set.

I set emtb at 20, max watts at 700w, max amp at 17a. The Max watts it shows as I ride is 680w
Thanks

Increasing the sensitivity does not increase the power, you are already at the maximum, but you would get there with less effort.
Is that what you want?
I could eliminate sensitivities 1 and 2, move the others by 2 positions and add two new sensitivities, 19 and 20.
Can it go well?

You fixed the "startup assist without pedal rotation" problem! That's great!
As the weather and the temperature warms I turned down my Amps and Watts. But during the winter or anytime below 60 degrees Fahrenheit I can run 700 watts and 17 Amps and it works fine, no overheating.
As far as the sensitivity goes, I would just like to be able to adjust the torque sensor sensitivity in the software rather than starting the system with a negative sensitivity setting by pulling back on the cranks as the system boots up.
Thanks!

 

[mbrusa]

I try to see if I can do anything.
Can you tell me the values of "Torque sensor ADC value" in LCD3 Advanced Technical Data.
Value by pulling back on the cranks as the system boots up.
Value without pulling back the cranks.

 

[jeff.page.rides]

I try to see if I can do anything.
Can you tell me the values of "Torque sensor ADC value" in LCD3 Advanced Technical Data.
Value by pulling back on the cranks as the system boots up.
Value without pulling back the cranks.

The ADC with no load on the cranks is 130. Pulling back with both arms and 152 cranks as hard as I do when I turn on the TSDZ2 is 100 ADC. If we can have a setting that you could adjust the sensitivity from -5 to - 35 that would be awesome. Especially for those that have a standard TSDZ2 and can't adjust it by pulling back as it boots up.

I appreciate very much all that you're doing MBRUSA and also all the testing that THINEIGHT is doing, THANKS!

 

[mbrusa]

One question,pulling back on the cranks as the system boots up, when they release, does the bike not start alone without pedaling?

 

[casainho]

I try to see if I can do anything.
Can you tell me the values of "Torque sensor ADC value" in LCD3 Advanced Technical Data.
Value by pulling back on the cranks as the system boots up.
Value without pulling back the cranks.

The ADC with no load on the cranks is 130. Pulling back with both arms and 152 cranks as hard as I do when I turn on the TSDZ2 is 100 ADC. If we can have a setting that you could adjust the sensitivity from -5 to - 35 that would be awesome. Especially for those that have a standard TSDZ2 and can't adjust it by pulling back as it boots up.

I appreciate very much all that you're doing MBRUSA and also all the testing that THINEIGHT is doing, THANKS!

On the latest firmware on my branch, the coast brake is already reported to work by one user (you guys can look at the code). Soon I will make a release were the amount of negative torque to detect a brake, will be configured by user, then I will consider the coast brake version is fully working.

 

[jeff.page.rides]

One question,pulling back on the cranks as the system boots up, when they release, does the bike not start alone without pedaling?

No, the bike doesn't start on its own, it starts just the same as if I hadn't pulled back when we boot the system. The difference is once you start to move it's a lot smoother has a lot more power and continues to provide power with Less input.

 

[famichiki]

I haven't been on here for a long time. Good to see development continuing and that Mbrusa has fixed the motor overrun issue. Before the world went crazy, I intended to work on this further and publish my code.. but my priorities obviously changed.

Here it is for reference, perhaps it is different to Mbrusa's approach? It totally removes the overrun and without any jarring effect.

The main change is in motor.c here:

https://github.com/OpenSource-EBike-firmware/TSDZ2-Smart-EBike/blob/v0.20.0-beta.1/src/controller/motor.c#L775-L776

Replace this:

// set the ticks counter limit depending on current wheel speed
ui16_cadence_sensor_ticks_counter_min = ui16_cadence_sensor_ticks_counter_min_speed_adjusted;

With this:

// set the ticks counter limit depending on cadence RPM
if (!ui16_cadence_sensor_ticks)
{
    ui16_cadence_sensor_ticks_counter_min = CADENCE_SENSOR_TICKS_COUNTER_MIN;
}
else
{
    ui16_cadence_sensor_ticks_counter_min = ui16_cadence_sensor_ticks;
}

There are some supporting modifications that will need to be done to ebike_app.c which I can explain if anyone is interested.

In addition, to elimate the speed dependent ramp down both PWM_DUTY_CYCLE_RAMP_DOWN_INVERSE_STEP_DEFAULT and PWM_DUTY_CYCLE_RAMP_DOWN_INVERSE_STEP_MIN need to be set to identical values in main.h

 

[mbrusa]

Welcome back famichiki

My first test was also similar to yours, based on ui16_cadence_sensor_ticks_counter_min.
It was pretty good but I couldn't get it to work in advanced mode.
So I preferred to create a separate code that works in parallel with the existing one.

 

[mbrusa]

One question,pulling back on the cranks as the system boots up, when they release, does the bike not start alone without pedaling?

No, the bike doesn't start on its own, it starts just the same as if I hadn't pulled back when we boot the system. The difference is once you start to move it's a lot smoother has a lot more power and continues to provide power with Less input.

Hi Jeff
Today I tried to add a negative offset to the ADC torque value also to the version for stock displays.
It's very good, it's a pleasant feeling, especially at the start.
Only drawback, it increases the overrun time.
Didn't you notice this?
It's not a problem?

 

[jeff.page.rides]

One question,pulling back on the cranks as the system boots up, when they release, does the bike not start alone without pedaling?

No, the bike doesn't start on its own, it starts just the same as if I hadn't pulled back when we boot the system. The difference is once you start to move it's a lot smoother has a lot more power and continues to provide power with Less input.

Hi Jeff
Today I tried to add a negative offset to the ADC torque value also to the version for stock displays.
It's very good, it's a pleasant feeling, especially at the start.
The only drawback, it increases the overrun time.
Didn't you notice this?
It's not a problem?

I don't notice any extra overrun. My one bike does use a VLCD5 so if you can send me a copy of the latest file to flash for the motor all have it flashed and see how it works.
When you have a chance to update the latest file for the LCD3 let me know and I'll flash that one too.
Thanks!

 

[jbalat]

Hey guys keep the development going.

Just wondering if anyone is on top of the field weakening code that casainho has just introduced in his 860 version?

Is it possible to just use that firmware on the ktlcd3 with some minor mods or do we need to understand the changes and copy them over to 20beta1?

I can start doing compares with his older and newer versions and see if I can understand it.

 

[Elinx]

.....
I can start doing compares with his older and newer versions ...........

I had about the same idea for replacing the higher cadence option for this, but I am no programmer and only have done a comparision to see the code.
The comparision for fieldweakening I could do was these.

That is a comparision with the last build and isolates the changes needed for field weakening
If I compare the 5 files with Total Commander with v0.20, the differences are huge, I lost overview.
So I hope someone can made some "chocolate" of this.