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

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

Page 22 / 53
  RSS
dobbewan
(@dobbewan)
Trusted Member
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Posted by: AnatomicFlack

Guy,

I see you added an "0.9-degree-with-LA-15" branch to your github.  Is that ready to play with, or something you are still working on?  

 

I noticed this yesterday as I was trying to find time to compile/merge my own as well... Super excited to see the community working on this! Really excited to get off work today and flash this on my bear setup! I am assuming whatever guy is working on is still a "test" but I am willing to try it out and compare with jta and my own. I am not a expert at combining or merging firmware, so I am having a bit of trouble in places. Hopefully this will give me some answers because I really really really want LA 1.5 working on this bondtech pancake seeing that I will have to rewire the pinout to go back to the moons stepper.  I am hoping this will solve the sloppy retractions and clicking like others have claimed 🙂

Posted : 03/07/2019 3:20 pm
Bunny Science
(@bunny-science)
Noble Member
Topic starter answered:
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)

HOLD UP!. No, that branch was created to see if GitHub could merge in the changes. I have not manually done a merge. Wasn't worth the effort until I hear that the LA 1.5 is actually helping with 0.9 motors.

 

Posted : 03/07/2019 3:42 pm
dobbewan
(@dobbewan)
Trusted Member
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Posted by: guy.k2

HOLD UP!. No, that branch was created to see if GitHub could merge in the changes. I have not manually done a merge. Wasn't worth the effort until I hear that the LA 1.5 is actually helping with 0.9 motors.

 

Oh I have not pilfered through it yet!! So no damage done! lol. I just noticed this last night before bed.

From what Chris W is recommending to me after some feedback, is that LA 1.5 would clear up my 0,9 pancake retraction issues and the horrid clicking sounds from that lack of torque. This is why I am really really interested in this. I don't really feel like disassembling the bondtech and rewiring for the correct pins if LA 1.5 corrects the retraction issues like Chris W claims.  I will compare what I have with jta's branch and see how bad I screwed up the manual merge   :/  

 

Posted : 03/07/2019 3:52 pm
CybrSage
(@cybrsage)
Honorable Member
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Posted by: guy.k2

Cybersage,

1. What does M503 report in the Pronterface terminal window?

2. Could you please attach your entire configuration_Prusa.h file? So we can see what you set.

 

 

I 7-zipped the file since I was not sure it would accept it otherwise. 

SENDING:M503
echo:Steps per unit:
echo: M92 X100.00 Y100.00 Z400.00 E415.00
echo:UStep resolution:
echo: M350 X8 Y8 Z16 E32

Wondering if this is another stuck data issue...but full factory reset is a pain in the ass, so I figured I would wait and let you look at it first.

Attachment removed
This post was modified 5 years ago by CybrSage
Posted : 03/07/2019 4:27 pm
Bunny Science
(@bunny-science)
Noble Member
Topic starter answered:
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)

Cybersage, your machine is wining the prize for now following instructions.

Your configuration_prusa.h looks correct to me. It should be setting e-microsteppting to 16.

However, your M503 reports E32 for microstepping.

Posted : 03/07/2019 4:50 pm
CybrSage
(@cybrsage)
Honorable Member
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)

Thunk I should see of Prusa will send me a new board? I will do a full factory reset first.

I will put on the new firmware, make sure it reads 16, put on the old, make sure it reads 32, then back to 16.  If it fails to do that stuff after a full reset, I will hot support.  Sound like a good plan?

Posted : 03/07/2019 4:56 pm
AnatomicFlack
(@anatomicflack)
Eminent Member
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)

Hmmm... I just updated to your July 1st firmware and I'm getting the same result.  I made sure to copy over the new configure from the variant folder.

But I'm seeing the following on an M503:

12:51:54.093: echo:Steps per unit:
12:51:54.097: echo: M92 X100.00 Y100.00 Z400.00 E415.00
12:51:54.097: echo:UStep resolution:
12:51:54.102: echo: M350 X8 Y8 Z16 E32

From the Configure_prusa.h see the following:

//Motors used should be 1 amp or lower current rating to avoid overheating TMC2130 drivers in Stealthchop.
//My recommended 0.9 degree motors for X, Y, or direct drive E are Moons MS17HA2P4100 or OMC 17HM15-0904S
#define X_AXIS_MOTOR_09 //kuo exper X axis
#define Y_AXIS_MOTOR_09 //kuo exper Y axis
//#define Z_AXIS_MOTOR_09 //kuo exper Z axis
//#define E_AXIS_MOTOR_09 //kuo exper EXTRUDER

//Uncomment ONLY ONE or NONE of below for geared extruders
//Don't forget to also send gcode to set e-steps
//Reversion back from geared extruder requires sending M92 E280 & M500 to printer
//#define BMG_EXTRUDER //Kuo Uncomment for BMG 3:1 extruder. This also sets BMG height for you. MUST also send M92 E415 & M500 to set esteps
//#define EXTRUDER_GEARRATIO_30 //Kuo Uncomment for extruder with gear ratio 3.0. MUST also send M92 E420 & M500 to set esteps
//#define EXTRUDER_GEARRATIO_3375 //Kuo Uncomment for extruder with gear ratio 3.375 like 54:16 BNBSX. MUST also send M92 E473 & M500 to set esteps
#define EXTRUDER_GEARRATIO_35 //Kuo Uncomment for extruder with gear ratio 3.5 like 56:16 Bunny and Bear Short Ears or Skelestruder. MUST also send M92 E490 & M500 to set esteps

---

#ifndef EXTRUDER_GEARED
#ifndef E_AXIS_MOTOR_09
#define TMC2130_USTEPS_E 32 // Kuo1.8 motor, non-geared
#else
#define TMC2130_USTEPS_E 16 // Kuo 0.9 motor, non-geared
#endif
#else
#ifndef E_AXIS_MOTOR_09
#define TMC2130_USTEPS_E 16 // Kuo 1.8 motor, geared extruder
#else
#define TMC2130_USTEPS_E 8 // Kuo 9.9 motor, geared extruder
#endif
#endif

Maybe my case is different though because I'm not using a 0.9 extruder motor, but a 1.8.  So should I be at 32 microsteps or 16 running a BNBSX and 1.8 stepper?  Maybe the if statement isn't reaching a case that matches my setup or the primary if logic is reversed?  Shouldn't the first block of the main IF be for those running a geared extruder setup and the else be for non-geared because you set the EXTRUDER_GEARED variable to 1 if a geared extruder is selected?

This post was modified 5 years ago 2 times by AnatomicFlack
Posted : 03/07/2019 5:06 pm
CybrSage
(@cybrsage)
Honorable Member
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Posted by: AnatomicFlack

Hmmm... I just updated to your July 1st firmware and I'm getting the same result.  I made sure to copy over the new configure from the variant folder.



Maybe my case is different though because I'm not using a 0.9 extruder motor, but a 1.8.  So should I be at 32 microsteps or 16 running a BNBSX and 1.8 stepper?  Maybe the if statement isn't reaching a case that matches my setup or the primary if logic is reversed?  Shouldn't the first block of the main IF be for those running a geared extruder setup and the else be for non-geared because you set the EXTRUDER_GEARED variable to 1 if a geared extruder is selected?

The BMG extruder uses a 1.8 degree stepper motor.  Normally, it would be at 32, but he is changing it (he posted this back on the 1st):

"Ironically, Bunnies will soon change the 0.9 firmware to always reduce extruder microstepping to 16 if a geared extrude is selected. This is to avoid EINSY speed limits during fast extruder moves like retractions and MMU2S maneuvers.

When that firmware update is made, e-steps will need to be halved."

Posted : 03/07/2019 5:40 pm
AnatomicFlack
(@anatomicflack)
Eminent Member
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Posted by: CybrSage
Posted by: AnatomicFlack

Hmmm... I just updated to your July 1st firmware and I'm getting the same result.  I made sure to copy over the new configure from the variant folder.



Maybe my case is different though because I'm not using a 0.9 extruder motor, but a 1.8.  So should I be at 32 microsteps or 16 running a BNBSX and 1.8 stepper?  Maybe the if statement isn't reaching a case that matches my setup or the primary if logic is reversed?  Shouldn't the first block of the main IF be for those running a geared extruder setup and the else be for non-geared because you set the EXTRUDER_GEARED variable to 1 if a geared extruder is selected?

The BMG extruder uses a 1.8 degree stepper motor.  Normally, it would be at 32, but he is changing it (he posted this back on the 1st):

"Ironically, Bunnies will soon change the 0.9 firmware to always reduce extruder microstepping to 16 if a geared extrude is selected. This is to avoid EINSY speed limits during fast extruder moves like retractions and MMU2S maneuvers.

When that firmware update is made, e-steps will need to be halved."

I read all that and I though I understood what was going on, I think I'm confusing myself as to what we should be using for microsteps and esteps.  

So with the 7/1 version of the firmware we should both be at 16 microsteps and a halved value of our original esteps?  

So for me running the BNBSX on an 1.8 motor I should see E16 (not E32) and I should set my esteps to E490 (I just realized I set the wrong value for mine, I since corrected it, but I'm still running half the proper extruder speed). However even trying commenting out the logic tree for the extruder microstepping and just passing (#define TMC2130_USTEPS_E 16 // Kuo 1.8 motor, geared extruder) it's still showing a setting of E32 when I run M503.   

Posted : 03/07/2019 5:50 pm
Bunny Science
(@bunny-science)
Noble Member
Topic starter answered:
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)

The design intent is....

non-geared, 1.8 degree motor --> micr0step 32 (the Prusa default)

non-geared, 0.9 degree motor --> micr0step 16 (half of Prusa default so e-steps are no different from stock)

geared, 1.8 degree motor --> micr0step 16 (half normal should work with 1/2 usual geared esteps)

geared, 0.9 degree motor --> micr0step 8 (quarter of normal to keep esteps same as with 1.8 degree motor,  1/2 usual geared esteps)

 

This post was modified 5 years ago by Bunny Science
Posted : 03/07/2019 5:51 pm
AnatomicFlack
(@anatomicflack)
Eminent Member
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Posted by: guy.k2

The design intent is....

non-geared, 1.8 degree motor --> micr0step 32 (the Prusa default)

non-geared, 0.9 degree motor --> micr0step 16 (half of Prusa default so e-steps are no different from stock)

geared, 1.8 degree motor --> micr0step 16 (half normal should work with 1/2 usual geared esteps)

geared, 0.9 degree motor --> micr0step 8 (quarter of normal to keep esteps same as with 1.8 degree motor,  1/2 usual geared esteps)

 

So should this result in a value of 16? 

//#ifndef EXTRUDER_GEARED
// #ifndef E_AXIS_MOTOR_09
// #define TMC2130_USTEPS_E 32 // Kuo1.8 motor, non-geared
// #else
// #define TMC2130_USTEPS_E 16 // Kuo 0.9 motor, non-geared
// #endif
//#else
// #ifndef E_AXIS_MOTOR_09
// #define TMC2130_USTEPS_E 16 // Kuo 1.8 motor, geared extruder
// #else
// #define TMC2130_USTEPS_E 8 // Kuo 9.9 motor, geared extruder
// #endif
//#endif

#define TMC2130_USTEPS_E 16 // Kuo 1.8 motor, geared extruder

I'm still getting E32 from an M503.

13:56:37.118: echo: M92 X100.00 Y100.00 Z400.00 E490.00
13:56:37.122: echo:UStep resolution:
13:56:37.122: echo: M350 X8 Y8 Z16 E32

 

Posted : 03/07/2019 5:57 pm
AnatomicFlack
(@anatomicflack)
Eminent Member
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)

So one more additional factory reset did the trick.  I already did one when I first flashed the 7/1 firmware with the original logic tree for the extruder steps.  So it must have somehow still set it to 32 in the firmware.  

14:02:41.190: echo: M92 X100.00 Y100.00 Z400.00 E490.00
14:02:41.190: echo:UStep resolution:
14:02:41.194: echo: M350 X8 Y8 Z16 E16
Posted : 03/07/2019 6:04 pm
Bunny Science
(@bunny-science)
Noble Member
Topic starter answered:
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)

WTF is going on with needing all these factory resets to set microstepping? 

Posted : 03/07/2019 6:10 pm
Dave Avery
(@dave-avery)
Honorable Member
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)

i had weird problems after the latest 0.9 firmware flash ( y axis failed self test - "loose pulley?" ) a data clear and self test and xyz cal all ran flawlessly

Posted : 03/07/2019 6:17 pm
AnatomicFlack
(@anatomicflack)
Eminent Member
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Posted by: guy.k2

WTF is going on with needing all these factory resets to set microstepping? 

😉

Does seem odd though without a good reason.  

This post was modified 5 years ago by AnatomicFlack
Posted : 03/07/2019 6:21 pm
AnatomicFlack
(@anatomicflack)
Eminent Member
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)

I've been getting them for a while even on the original version.  I believe it's related to a mismatch size when it's trying to compare the bed axis.  If you watch the console when it's running the self-test you can see the mismatch value.  Basically, the logic says if the bed is a certain amount larger than I expect it to be, then the motor pulley must be loose.  I think if the value is much larger it may throw the loose belt message.  

So a reset and new XYZ calibration gathers new values for the axis dimensions and offsets. 

This post was modified 5 years ago by AnatomicFlack
Posted : 03/07/2019 6:24 pm
JPTa
 JPTa
(@jpta)
Active Member
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Posted by: AnatomicFlack
Posted by: JTa

 

I have been testing with LA 1.5 and the 0.9 degree forks and merged them manually for the reason I'm not familiar with Git (CVS and SVN I know better :D). You can find my merge of forks here: https://github.com/JPTa/Prusa-Firmware

Can you fill in some details, I'm really interested in trying this out.  I guess the big questions are; Are you running the merge?  And have you encountered any issues?  Also is the OLED support something those of us with an LCD need to make edits around?

I'm running that fork of mine on my stock MK3 and on my separately sourced Bear which has 0.9 motors on xye (and 9mm belt on Y, skelestruder, OLED, Volcano with 0.8 nozzle, PT100, Sunon fans..etc.) and with it only issue has been the quite common problem with X homing which I got fixed by tuning the belt tightness. I had that same homing issue with "stock" Guy's 0.9 fork so it was not something the LA1.5 patching caused.

There hasn't been any problems with stock MK3. Note: I also merged Yuri's LCD/planner stall fix yesterday. And I also started another branch with Yuri's double-edge stepping implementation which I have not yet tried on printer, just did a quick compile test.

OLED fixes are not gonna cause any problems as long as you don't enable support for OLED. 🙂

 

Posted : 04/07/2019 4:31 am
dobbewan
(@dobbewan)
Trusted Member
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)

With your help and Guys help I have this now running perfect with LA 1.5 on my bondtech LDO 0.9 pancake! I manually did some of my merge with a older version of Guys firmware. No more clicking, no more violent retractions! Almost perfect! I will post some pictures shortly!

Posted : 04/07/2019 2:55 pm
AnatomicFlack
(@anatomicflack)
Eminent Member
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Posted by: JTa
Posted by: AnatomicFlack
Posted by: JTa

 

I have been testing with LA 1.5 and the 0.9 degree forks and merged them manually for the reason I'm not familiar with Git (CVS and SVN I know better :D). You can find my merge of forks here: https://github.com/JPTa/Prusa-Firmware

Can you fill in some details, I'm really interested in trying this out.  I guess the big questions are; Are you running the merge?  And have you encountered any issues?  Also is the OLED support something those of us with an LCD need to make edits around?

I'm running that fork of mine on my stock MK3 and on my separately sourced Bear which has 0.9 motors on xye (and 9mm belt on Y, skelestruder, OLED, Volcano with 0.8 nozzle, PT100, Sunon fans..etc.) and with it only issue has been the quite common problem with X homing which I got fixed by tuning the belt tightness. I had that same homing issue with "stock" Guy's 0.9 fork so it was not something the LA1.5 patching caused.

There hasn't been any problems with stock MK3. Note: I also merged Yuri's LCD/planner stall fix yesterday. And I also started another branch with Yuri's double-edge stepping implementation which I have not yet tried on printer, just did a quick compile test.

OLED fixes are not gonna cause any problems as long as you don't enable support for OLED. 🙂

 

Cool... I saw the Yuri stuff yesterday when Chris posted about it.  

I may have to give yours a spin.  

Any helpful tips for adjusting the configure (for the extra variables yours adds)?

I'm running 0.9 XY but 1.8 on E with a BNB extruder.  

Posted : 04/07/2019 2:58 pm
Bunny Science
(@bunny-science)
Noble Member
Topic starter answered:
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)

Dobbewan, that sounds very promising. Could I have you do a retraction test print for us? The attached 3mf is a simple rectangular prism. 

Print setting should be set up to slice as two perimeters, no infill, no top layers. Basically, it must print hollow with two perimeters, definitely NOT in vase mode. This force rectangle to it will have a seam at which the printer will do a retraction and restart for both inner and outer layers. I noticed early in my 0.9 pancake testing that examination of the seam, especially the inside layer would reliably show gaps at the restart with 0.9 pancakes. From the outside, a failing restart looks like an odd distortion at the seam. Under transillumination, a failing restart looks thin at the seam.

If your LA 1.5 firmware and 0.9 pancake can print this with perfect restarts at the inside layer, it would be objective evidence that the LA 1.5 is achieving the result intended.

Thanks for forging forward for us all. 

 

 

Attachment removed
Posted : 04/07/2019 5:27 pm
dobbewan liked
Page 22 / 53
Share: