Is hold_current required for auto tune?

[spikeygg]

I've been looking at tuning the drivers and according to the klipper docs, hold_current should no longer be used. However, the instructions in this GitHub repo's readme say:

• Currents (run current, hold current, homing current if using a Klipper version that supports the latter)

I also found this post from Sineos on the klipper discourse server from 2021 which cites that there are negative impacts from using both run_current and hold_current.

I installed the klipper_tmc_autotune and ran it without adding the hold_current to my configuration and everything seems to be working... I just don't want to be missing something important here.

[robustini]

hold_current is not recommended any longer.
It has been found that the switch between run_current and hold_current has a negative impact on precision.
So this script should not use it at all.

[Sonnenwerk]

Had the same question this day. Anyone willing to adapt the docs ?

[andrewmcgr]

The klipper docs are correct *for klipper's default settings* and not correct for a properly tuned driver. You can deduce this from the chip datasheets. If you can demonstrate that I'm wrong with actual data (say, hook up an encoder to a motor and track what actually happens at adequate precision), sure, but otherwise it's staying how it is.

…

On Fri, Jan 3, 2025 at 5:43 AM Sonnenwerk ***@***.***> wrote: Had the same question this day. Anyone willing to adapt the docs ? — Reply to this email directly, view it on GitHub <#192 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AANJYEV2B3GOJKPYQQXPYC32IWCF5AVCNFSM6AAAAABMWOLYVWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKNRYGIYTAMBXGA> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[Sonnenwerk]

Ah ! You are right !