I have some comments about the settings for TMC drivers (at least the TMC2209) described in your user guide document pages 262 to 264, and also to some parameters that can be worth testing (like the interpoaltion flag you added in the last firmware version)
In the optimizing settings, page 263 there is in the example code for setting a call to :
mystepper->rms_current( TMC2209CURRENT )
and few lines after
mystepper->irun(21); // irun = CS (current scale setting) on the sheet
this is actually setting two times the current, since rms_current method is calculating and setting the current scale (as well as the vsense if curent scale is too small)
it is then better/simpler to call rms_current function which doing the current scale calculations and setting for you.
Be aware that the TMC2209CURRENT is RMS, which is 1.4 smaller than the max curent found in motor datasheets (for a motor with 400mA current the RMS current is slighlty less than 300mA )
//extract of tmcstepper arduino code:
void TMCStepper::rms_current(uint16_t mA) {
uint8_t CS = 32.01.41421mA/1000.0(Rsense+0.02)/0.325 - 1;
// If Current Scale is too low, turn on high sensitivity R_sense and calculate again
if (CS < 16) {
vsense(true);
CS = 32.01.41421mA/1000.0(Rsense+0.02)/0.180 - 1;
} else { // If CS >= 16, turn off high_sense_r
vsense(false);
}
Also it can be interesting to "play with" the possibility to set a hold current multiplier, which is allowing to have the coil power on but with reduced power/heating (this is the same rms_current call, with a second member being the hold multiplier .
And another setting ( either through the define, or in future version through the management server) can be choosing between the two chopper modes (stealthchop or spreadcycle), allowing to choose between noiseless and torque performance and accuracy depending of the needs
Regards
Joël
Last edit: Collet 2024-10-02
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi Joël
Thanks for the suggestions, I am always open to anything that improves the code and makes things better.
As I think I explained a number of times, the tmc code was not written soly by me and involved others, though I did have some input into it, it was more like a joint effort. I could not test the tmc code because I do not have any tmc modules. I seem to remember a lot of discussion in the beta team whould helped-tested the code for tmc.I am extermely greatful to the help given by Paul and team for this work and documentation supporting it. There would have been none had it not been so.
I takes me longer and longer to write code now, and I know that I am not up to that task looking ahead. Honestly, it will take me several days of re-reading your post for me to grasp it, Not because I am an idiot - but the brain ain't working like it used to.
If you would like to make the changes to the code I am more prepared to drop that in. I also have a good suggestion by Eric concerning the motor board configs and again, it is hard for me to grasp now.
Peehaps it might be better to move the firmware back to github. The reason it moved was because I was trying to keep everything together, rather than having bits here and there. Git hub is not a great tool when one combines it with documents, board designs, etc, etc
regards
Robert
Last edit: brownrb 2024-10-03
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I am also very greatful to you and all people working on this master piece (Paul, Eric ,....). I can help, write some "code tuning"and post here or there and contribute a little (but it will be very little, compared to the work already done) or on a github if you open one.
Nevertheless a big thank you and Hat Off for this and also the others "equipments" (SQM, boards..)
Regards
Joël
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Good morning.
I gave up on the home sensor a long time ago, since it was impossible for me to make it work.
Would these changes make this easier?
Would they work with the LILYGO board?
Thank you!
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
It will not works because the Diag pin is not linked with the CPU.
Even if it was linked, this function is not reliable because of the
slow speed (ref Trinamic date sheet)
Le 22/10/2024 à 08:28, Carlos Izquierdo Vázquez a écrit :
Good morning.
I gave up on the home sensor a long time ago, since it was impossible
for me to make it work.
Would these changes make this easier?
Would they work with the LILYGO board?
Thank you!
Hi Robert, all
I have some comments about the settings for TMC drivers (at least the TMC2209) described in your user guide document pages 262 to 264, and also to some parameters that can be worth testing (like the interpoaltion flag you added in the last firmware version)
In the optimizing settings, page 263 there is in the example code for setting a call to :
mystepper->rms_current( TMC2209CURRENT )
and few lines after
mystepper->irun(21); // irun = CS (current scale setting) on the sheet
this is actually setting two times the current, since rms_current method is calculating and setting the current scale (as well as the vsense if curent scale is too small)
it is then better/simpler to call rms_current function which doing the current scale calculations and setting for you.
Be aware that the TMC2209CURRENT is RMS, which is 1.4 smaller than the max curent found in motor datasheets (for a motor with 400mA current the RMS current is slighlty less than 300mA )
//extract of tmcstepper arduino code:
void TMCStepper::rms_current(uint16_t mA) {
uint8_t CS = 32.01.41421mA/1000.0(Rsense+0.02)/0.325 - 1;
// If Current Scale is too low, turn on high sensitivity R_sense and calculate again
if (CS < 16) {
vsense(true);
CS = 32.01.41421mA/1000.0(Rsense+0.02)/0.180 - 1;
} else { // If CS >= 16, turn off high_sense_r
vsense(false);
}
if (CS > 31)
CS = 31;
irun(CS);
ihold(CS*holdMultiplier);
//val_mA = mA;
}
Also it can be interesting to "play with" the possibility to set a hold current multiplier, which is allowing to have the coil power on but with reduced power/heating (this is the same rms_current call, with a second member being the hold multiplier .
And another setting ( either through the define, or in future version through the management server) can be choosing between the two chopper modes (stealthchop or spreadcycle), allowing to choose between noiseless and torque performance and accuracy depending of the needs
Regards
Joël
Last edit: Collet 2024-10-02
Hi Joël
Thanks for the suggestions, I am always open to anything that improves the code and makes things better.
As I think I explained a number of times, the tmc code was not written soly by me and involved others, though I did have some input into it, it was more like a joint effort. I could not test the tmc code because I do not have any tmc modules. I seem to remember a lot of discussion in the beta team whould helped-tested the code for tmc.I am extermely greatful to the help given by Paul and team for this work and documentation supporting it. There would have been none had it not been so.
I takes me longer and longer to write code now, and I know that I am not up to that task looking ahead. Honestly, it will take me several days of re-reading your post for me to grasp it, Not because I am an idiot - but the brain ain't working like it used to.
If you would like to make the changes to the code I am more prepared to drop that in. I also have a good suggestion by Eric concerning the motor board configs and again, it is hard for me to grasp now.
Peehaps it might be better to move the firmware back to github. The reason it moved was because I was trying to keep everything together, rather than having bits here and there. Git hub is not a great tool when one combines it with documents, board designs, etc, etc
regards
Robert
Last edit: brownrb 2024-10-03
Hi Robert,
I am also very greatful to you and all people working on this master piece (Paul, Eric ,....). I can help, write some "code tuning"and post here or there and contribute a little (but it will be very little, compared to the work already done) or on a github if you open one.
Nevertheless a big thank you and Hat Off for this and also the others "equipments" (SQM, boards..)
Regards
Joël
Here the code with my proposed modifications, for anyone wanting to test.
driverboard.cpp and related tmcstepper_defines.h.
Regards
Joël
Last edit: Collet 2024-10-03
Good morning.
I gave up on the home sensor a long time ago, since it was impossible for me to make it work.
Would these changes make this easier?
Would they work with the LILYGO board?
Thank you!
Hi Carlos.
It will not works because the Diag pin is not linked with the CPU.
Even if it was linked, this function is not reliable because of the
slow speed (ref Trinamic date sheet)
Le 22/10/2024 à 08:28, Carlos Izquierdo Vázquez a écrit :
--
Cet e-mail a été vérifié par le logiciel antivirus d'Avast.
www.avast.com