Hi all!
I read this post on Hackaday.com about how the DRV8825 is terrible at microstepping on slow speeds - is this causing anyone problems?
I haven’t finished my build yet (just started ;)) but I got curious and googled around, which lead me to this blog - where an electrical engineer sort out what’s causing the problem and how to work around it! Totally worth reading even if you don’t have a problem with choppy microstepping!!
Just add 8 cheap diodes to each driver, and the problem is solved… (or you can put the driver in fast decay mode - but that’ll hurt your ears when the motor is not moving!)
This makes me wonder if it really makes a difference to go with the 1/32 instead of 1/16-mode?
/Kalle
If you notice an issue I feel fast decay is the way to go.
The one thing that isn’t clear is how hard the stepper is being driven. If you run the stepper at max amps This is probably when this happens, You can hear the steppers they sound angry that is why I never run steppers at full power (and they get hot as hell). Regardless of driver being used.
We use smaller pulleys (for greater torque) which helps a lot. You can see the issue pop up in some 3D prints as a surface pattern (some people are using larger pulleys to try and gain a speed advantage which makes this worse). I am not sure if it is positional error showing up or the extruder stepper error. But to ease your mind the “jump” is less than half a full step but I will use half to make the math an over estimate.
We use 200 steps per mm of travel on the x and Y so each step is 0.005mm so the “error” would cause a positional difference of 0.0025mm at worst and is not cumulative. We use 4535.44 steps per mm (depending on leadscrew) so that error would be 0.00011mm.
In summary (hahah sounds official) If I did the math right and understand the article correctly, I would be super stoked if we noticed an error of that size, that would mean we have one hell of an accurate machine!
Thanks for quick reply!
Isn’t 200 steps per mm based on 1/32? at least the calculator changes to 6.25 steps per mm if full step is chosen, and that would lead to an error of 1/12.5 mm which is still only 0.08 mm =)
/Kalle
The error is only present during micro stepping. You can’t run the machine at full stepping it causes all kinds of issues and it is not enough resolution.
