[BUG] Maybe bad TMC initialization?

[GMagician]

I'm facing a weird situation that I'm trying to analize.

Cnf: TMC2209 with sensorless home

It's seems that if I power up system via USB and then turnon 24V power (and it seems sometimes, not always) when I start homing, x moves to home, reach endstop and then if stutter without ever terminating home.

If I turn on 24V and then connect usb, it seems this never happens.

Please note:

1. that threshold is "almost" calibrated and sometime X or Y start/stop immediately without doing correct home (then threshold is not for sure too low)
2. On shield there is a diode to power stepsticks vmot via 3.3V when 24V is not present

Is there some initialization/calibration TMC do when powered up then can't be fullfilled when missing 24V?

[ellensp]

And how did you indicate to the firmware that you enabled 24v late and it needed to reinitialize?

[boelle]

please use discord for discussions

[GMagician]

@ellensp my suspect is that when vmot comes to TMC something happens to it, or that when it's powered with low voltage it doesn't initialize correctly....

[GMagician]

@ellensp maybe I found how to replycate this, and it really seems related to missing 24V.

How to generate it:
Power 24V, connect USB:
homing is ok
Remove 24V but keep usb:
homing is bad
Don't know what happens but going on to investigate

[GMagician]

Ok, m122 give an asnwer and there is a problem that I'm not sure it can be solved...after removing 24V and reapplyig 24V I see that vsense and msteps changes
from 0.11 to .325 and mstep from 16 to 8

[GMagician]

Here is m122 before removing 24V (Z2 has wrong current, I just notices now)
15:05:37.387 : X Y Z Z2 E
15:05:37.387 : Address 0 0 0 0 0
15:05:37.418 : Enabled false false false false false
15:05:37.418 : Set current 900 900 900 1000 250
15:05:37.441 : RMS current 887 887 887 994 244
15:05:37.472 : MAX current 1251 1251 1251 1402 344
15:05:37.472 : Run current 28/31 28/31 28/31 17/31 7/31
15:05:37.472 : Hold current 14/31 14/31 14/31 8/31 3/31
15:05:37.503 : CS actual 14/31 14/31 14/31 8/31 3/31
15:05:37.525 : PWM scale 16 16 16 10 4
15:05:37.556 : vsense 1=.18 1=.18 1=.18 0=.325 1=.18
15:05:37.572 : stealthChop true true true true true
15:05:37.603 : msteps 16 16 16 16 16
15:05:37.625 : tstep max max max max max
15:05:37.625 : PWM thresh.
15:05:37.641 : [mm/s]
15:05:37.657 : OT prewarn false false false false false
15:05:37.657 : triggered
15:05:37.672 : OTP false false false false false
15:05:37.688 : off time 4 4 4 4 4
15:05:37.719 : blank time 24 24 24 24 24
15:05:37.719 : hysteresis
15:05:37.741 : -end 2 2 2 2 2
15:05:37.773 : -start 1 1 1 1 1
15:05:37.773 : Stallguard thrs 80 84 0 0 0
15:05:37.773 : DRVSTATUS X Y Z Z2 E
15:05:37.804 : sg_result 0 0 0 0 0
15:05:37.826 : stst
15:05:37.842 : olb
15:05:37.873 : ola
15:05:37.888 : s2gb
15:05:37.920 : s2ga
15:05:37.942 : otpw
15:05:37.957 : ot
15:05:37.989 : 157C
15:05:38.004 : 150C
15:05:38.026 : 143C
15:05:38.058 : 120C
15:05:38.058 : s2vsa
15:05:38.058 : s2vsb
15:05:38.058 : Driver registers:
15:05:38.073 : X 0xC0:0E:00:00
15:05:38.073 : Y 0xC0:0E:00:00
15:05:38.089 : Z 0xC0:0E:00:00
15:05:38.089 : Z2 0xC0:08:00:00
15:05:38.089 : E 0xC0:03:00:00
15:05:38.105 : Testing X connection... OK
15:05:38.105 : Testing Y connection... OK
15:05:38.127 : Testing Z connection... OK
15:05:38.127 : Testing Z2 connection... OK
15:05:38.142 : Testing E connection... OK

and this is after I removed and reapplyerd 24V (report only differences)

15:06:57.700 : RMS current 1602 1602 1602 994 441
15:06:57.738 : MAX current 2259 2259 2259 1402 622
15:06:57.769 : CS actual 31/31 31/31 31/31 31/31 31/31
15:06:57.784 : PWM scale 36 36 36 36 36
15:06:57.816 : vsense 0=.325 0=.325 0=.325 0=.325 0=.325
15:06:57.869 : msteps 8 8 8 8 8
15:06:57.954 : off time 3 3 3 3 3
15:06:57.985 : blank time 36 36 36 36 36
15:06:58.016 : -end -3 -3 -3 -3 -3
15:06:58.038 : -start 6 6 6 6 6
15:06:58.038 : Stallguard thrs 80 84 0 0 0
15:06:58.333 : Driver registers:
15:06:58.339 : X 0xC0:1F:00:00
15:06:58.339 : Y 0xC0:1F:00:00
15:06:58.355 : Z 0xC0:1F:00:00
15:06:58.355 : Z2 0xC0:1F:00:00
15:06:58.370 : E 0xC0:1F:00:00

[GMagician]

@ellensp I did another test. TMC reset when I remove 24V, not when I give it back...maybe is something related to my shield board....investigating
@boelle this may be a serious problem, if not related to my H/W, but solution will not be easy (not possible to know when 24V is missing)

[GMagician]

@Bob-the-Kuhn sorry to bother you but you are the one with the best knowledge on tmc. What may be the reason of the reset registers reported above with relationship to 24V (vmot) removing? Thanks for your attention

[Evg33]

M502
M501
in startup code can help

[GMagician]

@Evg33 keeping 3.3V coming from usb will not reset Cpu so data is not reloaded from eeprom (by the way I did M502 & m500 a tons and redone just now but nothing changes)

[GMagician]

@Bob-the-Kuhn, @ellensp, @boelle & @thinkyhead Hi everybody, sorry to spam this but I think it may be interesting to know.
When TMC2209 is powered from power supply (12V or 24V) and this power supply will miss for some reason and tmc will be powered with 3.3V (also on on vmot, and this is usually done to prevent firing it when usb powers chip but vmot is absent), this will generate this weird issue. Then using stepstick protectors (like watterott ones) will for sure save tmc but will give another issue.

I think it's important to know. I don't use discord because it's too dispersive. Hope this will not bother you, in such base please forgive me.

[GMagician]

ops I forgot...Powering vmot with 5V (with a protection diode) solves

[Evg33]

M502
M501
for reinit TMC
TMC powered from the PSU, not powered from USB.

[GMagician]

@Evg33 ok, now I understood why you said about M502 & M500 but I found a better solution and this solution may help everyone.

As stated before issue "born" when you switch off PSU supply and keep usb. Then you may turn on psu again but chips have lost cnf (note switch off with usb on. If you power with usb and then power PSU everything is ok). This is because vmot on my board (or with stepstick protector) is powered by vcc that it's 3.3V. On my board I have diodes on 3.3V and 5V to act as "stepstick protector" in this case I'm able to send to vmot a 5V when psu is off. With this solution tmc dont' lose configuration and once PSU is on again I'm able to work without M50X

Don't know why tmc2209 act this way and if only 2209 are affected or not (I suspect also 2208)

[Evg33]

All non standalone TMC with same problem.
2209,2208,2130,... etc.

My boards powered by USB from RPi3B+(octoprint) via USB and have relay module (220V to MainPSU).

[GMagician]

Than if you use usb 5V with a diode to vmot you solve this issue

[github-actions]

This issue has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.