[BUG] TMC2209 with LDO 0.9 VERY noisy/resonant/vibratory

[freighttt]

Ive got a strange issue with a new printer im settting up. i have a set of tmc2209's, on a 24v corexy.

the original 1.8* motors are louder than they should be, but i didnt think much of it. this week i installed a set of LDO 0.9* steppers, and they vibrate so hard that the x gantry sings. they arent missing steps, the printer operates fine, theres just insane resonance, even at low speeds. i spent the last 3 days working on the frame, replacing belts, went through marlin systematically and turned on and off every setting i could think of(LA, square wave, etc). i played with microstepping numbers, voltages, etc, to no avail. i had reached the end of my sanity when i decided to stick the steppers on another machine i have, running 2208s. suprise...smooth as butter.

another thing to note is that turning stealthchop on and off results in less vibration, but not lower noise. stealthchop is actually where most of the noise and vibration comes from, with it off there is less vibration(but still tons) and the 'r2d2' sounds are pretty pronounced. i do not use hybrid threshold, but have tested it on and off when i went through the various options systematically testing them(and combinations of them).

im positive now that this is something wrong with my build, not the motors themselves. in doing some online research, i see some people mentioning issues with PWM? Beyond config/config_adv im pretty lost with coding. i thought id ask here before i go randomly messing with stuff. the video this gent posted sounds pretty similar to what im experiencing.

teemuatlut/TMCStepper#82

i feel this may be a bug because the same motors do not have this problem on a different machine. the vibration can be felt when just holding the motor in your hand, without it being connected to anything or under stress, and again on the other machine the same motor, being fed the same voltage and step count, feels like butter. that leads me to believe this is marlin/tmc not getting along. this resonance/vibration is experienced with both 1.8*, and 0.9*, however, 0.9* exacerbate the issue.

My Configurations

Required: Please include a ZIP file containing your Configuration.h and Configuration_adv.h files.
Marlin-2.0.x.zip

Steps to Reproduce

1. begin print.
2. machine sounds like chewbacca
3. enable/disable stealth chop
4. similar but chewbacca has taken xanax and has calmed down a little with stealthchop off
5. systemically test and eliminate settings like linear advance, square wave, etc
6. still chewbacca
7. install rubber dampers anywhere that vibration could be amplified
8. han solo, please take your dog back, i dont want him anymore

Expected behavior: [What you expect to happen]

normal levels of stepper noise and vibration

Actual behavior: [What actually happens]
video:
imgur.com/a/ZEBlkVX

[Logo2001]

@freighttt
Did you activate SPREAD?
Set also the Stepper to 256 Microsteps and the correct Current in configurations_adv.h (// @section tmc_smart)

[freighttt]

@Logo2001
what do you mean by spread? spreadcycle? i currently have hybrid threshold off so they run either in stealthchop or spreadcycle only.

i currently have it set to 16 microsteps, should i increase it to 256 so the driver doesnt interpolate itself?(is that how it works?)

i will try setting the current via the firmware vs lcd and saving to eeprom.

[Logo2001]

@freighttt
Yes, i mean Spreadcycle. In your Case - Marlin give Steps out for 16 Microsteps. If Spread not activ, the stepper run with 16 microsteps only.
On my mainboard I have to set a jumper to activate the spread function of the TmC2209 !!!
Which Board you are using?
You can also try to deactivate Stealthchop & Hybrid to check is Spreadcycle possible/available.

[Logo2001]

one more question: which Stepper Driver and Version with TMC2209 do you have?

[freighttt]

@Logo2001
i am running BTT 2209 v1.2 on a BTT skr1.3

yes i have them running in spreadcycle, it is the only way to run them smoothly. unfortunately i run a small printing business and i have to run the machine, hopefully this issue isnt causing damage to the motors. it almost feels like a poor signal, but it is worse on stealthchop to where its not usable. spreadcycle is at least smooth enough that it doesnt seem to cause damage to the print. i would assume it was the drivers themselves but when i plug them into another machine, or the motors into another machine running 2208s, everything is silent and operates as it should. i have tried new fresh firmware, with no luck. i also tried and older version to no avail.

here i have found another thread with people having the same issue just last month in may.
https://reprap.org/forum/read.php?415,852017,873784#msg-873784

[Marlor]

Did you remember to reset your eeprom after the adjustment of steps/mm and speed?
Your max speed is halved and steps are double

[freighttt]

@Marlor steps have been doubled. i wasnt aware that speed needed to be halved, 0.9's should have enough torque to handle normal print speed, but i will try dropping the feedrate to 250. am i understanding you correctly though?

even in stealthchop with the horrendous noise, they dont miss steps at 65mm/s, and in spreadcycle they dont have any issues with my current feedrates or acceleration.

[Logo2001]

I don't understand how you can switch it around when you need soldering on the back...

[Logo2001]

only when the jumper is removed, the software can set stealth Chop, Spread Cycle and Hybrid
It is preset on your drivers and you cannot change it by software.
It is written in the manual.
It might confuse the drivers.
But I have not tested the differences.
Fact is that in Stealth Chop Mode my steppers (1,8°) are also loud.

[freighttt]

Y’all have lost me now.

BTT 2209s on an skr1.3 do not require any soldering to operate as standard with full uart control through the lcd provided your jumpers are correct.

My jumpers are correct and I can alternate between spread cycle and stealthchop through the lcd with a push of the button. There is an audible change in motor noise to confirm this. When operating correctly, they would be significantly quieter in stealth chop. In my case, the grinding noise goes away when placing them in spreadcycle, but they don’t get any louder than they already are. Stealthchop is as loud as spread cycle, though the tone and noises are different.

Most easily the difference can be demonstrated by switching between the modes at idle. With stealth chop on, there is no noise at idle with motors enabled, whereas with stealthchop there is some white noise that is audible if held close to your ear. It is under operation that stealth chop seems to send a ‘dirty’ signal causing the motors to be extremely loud and resonant, even when held loose in your hand, despite the motion being seemingly okay.

[Logo2001]

my statement was not entirely accurate.
I should finish reading the datasheet :)
see picture.
My setup definitely runs pure Spread Cycle. (Jumper to VDD)

[Logo2001]

i have tested with M569 and its says "Stealth Chop mode" on all axis.
I no longer understand the world.
Stealthchop is commented out in Marlin config and the Jumper "Spread" on the MB is switched to VDD (measured).
hmmm very suspicious

[freighttt]

@Logo2001 there is no spoon man

[laurienzu]

I have exactly the same problem as you!!! I switched to Marlin on my Sapphire Pro in order to enable UART mode on my new ERYONE TMC2209 drivers but I have some coil whine noise also when motors are not moving. The noise reduces when I disable StealthChop but starts again when motors are running... I tried everything (chopping time, lowering or increasing current) and it's impossible to remove, it only disappear in standalone mode so using the drivers not in UART 🙁
It's exactly as you described, motor works and moves correctly without missing steps but I don't want to have a singing printer (belts resonates a lot propagating the sound).
I hope someone can help us, I would like to use these drivers at the maximum potentialities.

[Logo2001]

@Loreagle
please upload your Config and adv.h files.
I want to see if I can find something for you and freighttt.