mbrusa said:
@AZUR, there are many posts where you re-propose the video of your tests with v1.1.0, the first time it was useful, after not.
@Mbrusa, First I apologize if I did something wrong. It was not my intention.
I would also like to thank you for the work you are doing.
Based on mspider65 Version V7 and display firmware made by Casainho and others, you are improving the performance of the TSDZ2 motor.
Great work. Well done. thanks.
I made two different videos.
One showing the "human Power" field and another video showing the "current motor" field. These two tests were done on an uphill road. 800 meters road with 8% slope.
I also made another video on a flat road.
As you refer, it is difficult to replicate the problem, that's why I put those three videos. In those videos I show 3 different test scenarios.
mbrusa said:
You say you want to do it to help me, ok I believe it and I appreciate it, to prove it though, you should at least try this version. Do not you think?
Of course, I will test your firmware and give you my feedback. I will do it as soon as possible.
IF I had no intention of testing your firmware, I was not spending time writing these posts. My language is not English, so it is not easy for me to make these posts.
After testing your version I will also, perhaps, test the V12 mspider65 version.
Do you know if the V12 works with the 850 or 860 display?
I am very pleased to know that you are already trying to solve this problem. And now, we have two experts working on the subject. You and mspider65. Very good !
I hope I can help and contribute to the solution of the problem. That's what motivates me!
The best way to test and replicate the problem is on an uphill road and without stopping the movement. Just stop pedaling, after PWM is at 100, maximum value, for a few moments. And preferably, only restart pedaling after the motor current reaches 0.
The test should be done on a road with a slope of about 8%.
The delays are also greater when the level of assistance is higher.
Also when the slope is greater, the delay is greater and you probably feel the delay.
The test environment that I set up is a good tool to measure the problem. But a Go Pro camera is required to record, at least the tsdz2 screen.
In the test I did on the flat road, (video I put on this forum)
I didn't feel the problem. I just noticed that the problem occurred when I saw the video. It was an accident that I found out on the recording.
mbrusa said:
It mainly occurs after a descent followed by a climb, if you resume pedaling late.
This is a good example. That is why this problem, in some situations, can be dangerous.
mbrusa said:
Conclusion, I have to believe both those who say that the problem is still there, and those who say that it no longer exists.
I understand very well what you mean.
I will try to help, testing your version and if the problem still exists, try to replicate it and if possible record and measure the problem.
mbrusa said:
I try to clarify. A distinction must be made between standing start and resuming pedaling in motion.
If there is a delay from standstill there are settings with which to remedy, always in any case.
Also for the restart in motion there are settings for improve, "Assist without pedaling" which now works also in motion, "Min current adc step" and also "Torque adc offset".
The problem will not be 100% eliminated, it will be less likely to occur.
It mainly occurs after a descent followed by a climb, if you resume pedaling late.
I added "Min current adc step" taken from v1.1.0 and it has improved a lot.
"Min current adc step" is the current value (ADC step) maintained during the movement of the pedals, even without any force detected by the torque sensor.
Only by always enabling "Min current adc step" even without the movement of the pedals the problem never occurs, but this is not acceptable.
Sunday I resumed the tests with the other bike, the one I used last summer, with the same firmware in Power + Boost mode I did not find any delay, plus with "Assist without pedaling" disabled and "Min current adc step" = 0, I haven't had the satisfaction of seeing the problem once.
The only difference is that on this bike, to have an immediate response from a standing start, I have to force the "Torque adc offset" value a lot, and I think this also affects the response in motion. Now I understand why I never saw the problem last summer.
When carrying out the tests, with your version, I will take all of your technical advice into account.
Thanks
Azur