Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
 

Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)  

Page 42 / 49
  RSS
CybrSage
(@cybrsage)
Honorable Member

Awesome!  Thanks!

Posted : 03/10/2019 11:08 pm
guy.k2
(@guy-k2)
Noble Member

Users should be aware that my 0.9 branch https://github.com/guykuo/Prusa-Firmware/tree/0.9-Degree-Stepper-Support is based on the Prusa MK3 branch, the one that gets continually changed by Prusa as they update the firmware. We are now at a 3.8.1RC, well beyond the release 3.8.0.

If you need something that is based closer to 3.8.0, I do have a separate snapshot available. It was taken before a big chunk of 3.8.1 Prusa changes. https://github.com/guykuo/Prusa-Firmware/tree/0.9-MK3-snapshot-sept-21-2019

Personally, I run the most recent 0.9 branch, not the old snapshot.

Posted : 03/10/2019 11:14 pm
nikolaistolstoy@gmail.com
(@nikolaistolstoy)
Eminent Member

@guy-k2

Hey there. Got a bit backed up with jobs and so have still been running your .9 3.7.0 branch. Also a bit backed up on this thread!

If I was to update to the current release in your .9 branch is there anything I would have to do differently? I remember some discussion of going to a different linear advance, do I need to re-measure K values? Also I think I will have to re-adjust e-steps as you reduced microsteps for geared extruders and I have a Bondtech. Anything other things I might have missed for when I do the update?

Thanks so much for all the effort you put into keeping this all update. I would be stuck with old firmware with my .9 and Bondtech combination if not for you.

Thanks so much

 

PS would you recommend the 3.8.0 ish branch or 3.8.1 for someone printing for clients?

This post was modified 10 months ago by nikolaistolstoy@gmail.com
Posted : 04/10/2019 1:59 am
guy.k2
(@guy-k2)
Noble Member

I would go with the 3.8.1 RC based one. It has behaved well and looking through the changes made since my snapshot branch, they appear reasonable and even fix a few bugs. For instance one thing recently added was a better mechanism for fully clearing the EEPROM.

Coming from 3.7x I recommend....

1. Absolutely do a factory reset with data clear after flashing to 3.8.x  Failure to do so, has created many weird problems. Erase all the data, don't just do a plain factory reset without erase.

2. Follow the micro-stepping and e-steps instructions in my branch directions. The order of operations matters. M350 must be BEFORE M92. Otherwise, M350 setting micr0-stepping will alter your previously set e-steps value.

K-factor  for LA 1.5 is definitely different. It's going to end up very low like 0.05 to 0.10. Mine came out to be 0.075. I set M900 in my custom startup code to set the value with each print. You can determine your k-factor with the test pattern generator at http://marlinfw.org/tools/lin_advance/k-factor.html

 

Posted : 04/10/2019 2:17 am
CybrSage liked
SigmaRelief
(@sigmarelief)
Active Member

I'm taking the 0.9 deg plunge after running your firmware in 1.8deg mode and am looking into the details of how to set the linearity correction to 1.130 for the OMC motors.  I have been following your tuning discussions since the beginning this spring and saw there was talk about adjusting the math used for linearity correction, but it is not immediately clear what needs to be changed as the variants file only appears to enable linearity, but does not set levels.  I believe the settings exist in tmc2130.cpp and the " tmc2130_set_wave(X_AXIS, 247, tmc2130_wave_fac[X_AXIS]);" looks promising, but I'm getting lost in the scaling function and can't determine what (or if) 247 needs to be replaced with to effectively set this.  I am still on your 3.8.0 branch, I'm sure there is a reason there isn't a #define for the individual motor type, but at least pulling the linearity variables into the variants file with a breakdown of the math would be a huge help for OMC users.

Posted : 16/10/2019 2:40 am
SigmaRelief
(@sigmarelief)
Active Member

Well that edit didn't go through...  I think I may just need to set "uint8_t tmc2130_wave_fac[4] = {1.130, 1.130, 0, 0};", but it would be great to note this in code and possibly pull to your helpful variants files.

Posted : 16/10/2019 3:00 am
vintagepc
(@vintagepc)
Noble Member

You know that can be set through the settings menu, right...?

Posted : 16/10/2019 1:41 pm
guy.k2
(@guy-k2)
Noble Member

can be set through the settings menu

Exactly. Also, it's not just a matter of simply setting the linearity correction value. We also need to store the value in EEPROM. Setting it through the settings menu stores the value so it survives. The exact value will vary a little bit for each motor.

Posted : 16/10/2019 2:54 pm
SigmaRelief
(@sigmarelief)
Active Member

Interesting, with all of the other settings one may want to tune that are not exposed to the user, this is a welcome change.

Posted : 17/10/2019 2:52 am
SpecialT
(@specialt)
Eminent Member

hi,

i have been looking for a 0,9 stepper motor for the last few days and have now found 2 that i could get. the question is whether the 1.6 ampere is too much with one motor. the other stepper motor has 5.6V for that.

 

https://de.nanotec.com/fileadmin/files/Datenblaetter/Schrittmotoren/ST4209/L/ST4209L1704-A.pdf

https://de.nanotec.com/fileadmin/files/Datenblaetter/Schrittmotoren/ST4209/L/ST4209L1206-A.pdf

Posted : 30/10/2019 9:02 pm
guy.k2
(@guy-k2)
Noble Member

I would expect TMC2130 overheating with a 1.6 amp motor. EINSY does not like much more that 1.0 amps on the motors.

Posted : 30/10/2019 10:24 pm
SpecialT
(@specialt)
Eminent Member

mhhh. the other Stepper has 1,2 amp ^^. its hard to find any good HQ Stepper in germany without import.

How much Ncm did you thin the 0,9 should have?

This post was modified 9 months ago by SpecialT
Posted : 31/10/2019 6:59 am
christoph.e5
(@christoph-e5)
Trusted Member

What's wrong with ordering the OMC ones? I got mine via amazon.de within a week.

Posted : 31/10/2019 12:46 pm
SpecialT
(@specialt)
Eminent Member

I'm not a fan of the cheap stepping motors. That's why I didn't want to save money, because these are decisive for the quality of the print.

Posted : 31/10/2019 4:53 pm
SpecialT
(@specialt)
Eminent Member

i Think these would work great "https://de.nanotec.com/fileadmin/files/Datenblaetter/Schrittmotoren/ST4209/L/ST4209L1206-A.pdf".

 

Posted : 31/10/2019 5:01 pm
christoph.e5
(@christoph-e5)
Trusted Member

I'm eager to see your results. For OMC we have those, so it's a known "good solution" regarding to the VFA issue. Others might or might not solve the VFA issue aside from maybe being higher quality steppers. Any data you can add will only benefit the comunity!

Posted : 31/10/2019 5:05 pm
vintagepc
(@vintagepc)
Noble Member

After having done some serious prints as of late with a variety of features in shiny black filament, I'm curious what others' experiences are with VFAs on curved surfaces. Parts that are X/Y planar are beautifully smooth, but when printing a cylinder I am still seeing VFAs to some degree or another on surfaces involving both X&Y motion. Have others experienced similar. or does it normally also go away in these situations? 

 

(OMC 0.9s w/ linearity correction turned on in the menu)

This post was modified 9 months ago by vintagepc
Posted : 04/11/2019 2:52 am
jondamora
(@jondamora)
Active Member

Has anyone tried these?

https://www.amazon.com/gp/product/B00W98OYE4/ref=ppx_yo_dt_b_asin_title_o07_s01?ie=UTF8&psc=1

They are $20, 0.9A and 0.9 degree steppers.  I have 2, and only installed one on my extruder so far, as I had that taken apart to install some other upgrades.  However I am having some trouble getting them configured correctly in the firmware.  I uncommented the line 

#define DEFAULT_AXIS_STEPS_PER_UNIT {100,100,3200/8,560}

but I still only get half the extruder steps I am supposed to.  I also commented the original line with the 280 in it.  Is there anywhere else I have to change?

Posted : 04/11/2019 3:08 am
jondamora
(@jondamora)
Active Member

Note on the above post, I do not have the mmu installed so I was hoping for the extra resolution by not reducing the microstepping, let me know if I am wrong in that.  Thank you for your work.

Posted : 04/11/2019 3:15 am
chocki
(@chocki)
Prominent Member

I think you will find that this is actually a limitation of the 8 bit Einsy board.

A curve is not actually a curve as such but many small straight lines which create the curve and this is what you are seeing, jerk values have a big influence on this as well..

I'm sure Guy can explain it better.

This post was modified 9 months ago by chocki
Normal people believe that if it ain’t broke, don’t fix it. Engineers believe that if it ain’t broke, it doesn’t have enough features yet....
Posted : 04/11/2019 8:30 am
Page 42 / 49
Share:

Please Login or Register