Accuracy of setting stepper current in UART mode #1024
Replies: 1 comment 5 replies
-
|
Did you change up the rsense values in the configs and recompile the firmware for testing? I had noticed my Y motor has been a fair bit hotter (I think it's because I'm printing faster?) but dropping the current by 30ma seems to have resolved that issue for me. I wonder if the marlin guys on discord have any insights, they have a dedicated trinamics channel... |
Beta Was this translation helpful? Give feedback.
-
|
Yes I did, as I concluded the UART control is pretty complex and in StealthChop it uses an automatic tuning which is based on the CS value not on the Run current. I measured the temperature for the standalone mode then compared to that using UART with various Rsense and current values. I had read somewhere that the max current was defined by the trim pot but it turns out it isn’t. It does seem there is a wide scope for the current without issues, my Y axis settled around 47C. I ended up leaving the Rsense at 0.11 and setting the current for X&Z at 800 and Y at 730. |
Beta Was this translation helpful? Give feedback.
-
|
it's interesting what you say about E current setting in the adv.h, when I initially did the mod for the extruder and flashed the firmware, I used the default current setting 800ma, when I did the other 3 drivers, I had the correct current values for adv.h but E did actually come out at 800ma again, I thought it was a one off, set it via M906 E990 and forgot about it until you mentioned it. I'll see what I can find out from the marlin people, we may get some insights into how to set the drivers up, it's worth looking at how other boards with TMC uart drivers are configured too. From the ender 3 adv.h for SKR mini E3 v2 using TMC2209 chips in uart mode: Same for Y/Z with E being set to 650 |
Beta Was this translation helpful? Give feedback.
-
|
Yes very interesting, the 580mA is much closer to the actual RMS current that the V2 motors are running at, I believe that the E3 does actually use 0.11R sense resistors, I found a schematic with the values on. A more precise setting for the V2 would be 580mA with Sense of 0.15, which gives a similar result to 800 with 0.11. I didn’t change mine because of the E current problem, 800 with 0.15 would overheat the extruder motor. I will update the document to correct my errors and include the temperature measurements when I have done some more investigations. I too will try to track down the bug in E setting. If more users are going there be using UART mode it would be good to know that we are configuring it appropriately. |
Beta Was this translation helpful? Give feedback.
-
|
@Reggi3 I have completed the testing and as expected setting the Rsense to the required 0.15 results in the correct currents being set. This is the final report. I have looked to find the E0 bug but it is so complex it is impossible! |
Beta Was this translation helpful? Give feedback.
-
|
Well done, that's fantastic, thank you! Looks like I'll adding some information to my PDF too then :-) |
Beta Was this translation helpful? Give feedback.
-
Following from the Jyers gets Linear #707 I modified my 4.2.7 board for full UART and all appeared to perform well. However I started to investigate the current setting using Marlin. After digging into the code and the data sheet I found that there are some issues which I believe results in the currents not being set as specified.
Edit
After extensive tests measuring the temperatures of the drivers it appears that the current setting is even more complex than I thought. There is additional PWM scaling being implemented which affects the calculations.
However the conclusion is that even though the Rsense value in Marlin (0.11) is not the actual value used on the Creality boards the 800mA nominal driver currents that are quoted have also been calculated using the same Rsense. So the current settings do appear to correspond and in fact the current can be set higher than the standalone values.
Beta Was this translation helpful? Give feedback.
All reactions