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

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

Page 41 / 49
  RSS
guy.k2
(@guy-k2)
Noble Member

You're right. When we do mods, even if they improve or fix functionality, we can exceed the manufacturers reasonable responsibility.

There are so few stock parts left on my printer that I get to ask myself for warranty coverage.

A stock of spare parts is my "warranty."

 

This post was modified 11 months ago by guy.k2
Posted : 10/09/2019 8:59 pm
david.a66 liked
abulafia
(@abulafia)
New Member

Hello all,

After reading this long-ish thread, I have to admit that I don't understand some of the basic premises. Let me explain.

I would like to replace 1.8 steppers with 0.9 on all axes, including Z, and also replace the leadscrew with a 2mm lead (8mmOD, 1 start, 2mm pitch), but not to half the microsteps or in the case of Z, 1/8th.

So, can anybody explain why & how is Einsy limited to the number of steps/microsteps, as such that it is required to half the microsteping? I really don't understand where this limitation comes from. I use this setup on my Marlin on a full Rambo 1.3 w/o any problems.

Sorry if I asked a stupid question...

Best regards,

Bogdan

 

 

Posted : 14/09/2019 2:03 pm
guy.k2
(@guy-k2)
Noble Member

why & how is Einsy limited to the number of steps/microsteps, as such that it is required to half the microsteping?

During high speed carriage and extruder moves, the EINSY cannot keep pace with the microstepping if it isn't halved. You can try it without the halving. It will initially appear OK, but on complex prints, examination will reveal portions where curves are printed not as curves but tangents across where the curve would have been smooth. Planner falls behind and skips some intermediate positions. Extruder is also halved because it can fall behind fast MMU2S filament moves.

Posted : 14/09/2019 9:55 pm
abulafia
(@abulafia)
New Member

Thank u guy-k2!!

If this is the case, can one just reduce the   

axis speed/accel/jerk to a max=1/2 of the original?

Posted : 14/09/2019 10:42 pm
guy.k2
(@guy-k2)
Noble Member

can one just reduce the   

axis speed/accel/jerk to a max=1/2 of the original?

That approach could solve the X Y part of the issue, but doesn't help with the extruder and is less transparent to the end user.

I don't like going about it that way because it makes the user adjust all their existing gcodes and alter the slicer. By doing the fix (silently) in firmare microstepping, I save the user from modifying their PrusaSlicer settings or needing to reslice existing models. Also takes care of the extruder and MMU2 move issues.

Posted : 15/09/2019 1:06 am
christoph.e5
(@christoph-e5)
Trusted Member

I‘ve upgraded to OMC 0.9° for X and Y axis now and I have installed your 0.9 Firmware branch @guy-k2. I am a little confused regarding linearity correction for the steppers. Is this already done via your firmware or do I still set it via printer menu? 

Posted : 29/09/2019 9:33 am
guy.k2
(@guy-k2)
Noble Member

Linearity correction needs to be set in the printer menu. It is not set by my firmware, but is a user controlled setting. 

Posted : 29/09/2019 10:31 am
christoph.e5
(@christoph-e5)
Trusted Member

Thanks for your work and the support @guy-k2. It really is a substantial contribution to the community and much appreciated!

This post was modified 10 months ago by christoph.e5
Posted : 29/09/2019 2:05 pm
davystruyf
(@davystruyf)
Active Member

Came out of a upgrade frenzy. bondtech's MK3S upgrade kit and I got LDO's power 0.9 motors on X and Y. Had issues with it just crashing during home with just a rapid humming noise. essentially homing is broken.

somebody directed me here. got into your github and setup your firmware for the 0.9 motors. followed everything. flashed, factory reset and the setup. 

M92 X200.00 Y200.00 Z400.00 E415.00
M350 X8 Y8 Z16 E32

Is correct no?

Homing is still broken AF. when X homes left during the setup wizard it just hums rapidly and then when it homes right it just starts crashing back and forth on the right side and obviously calibration error. 

At this point I am just regretting getting talked into getting those LDO steppers. 🤣 

Posted : 30/09/2019 1:29 am
guy.k2
(@guy-k2)
Noble Member

Homing is still broken AF. when X homes left during the setup wizard it just hums rapidly and then when it homes right it just starts crashing back and forth on the right side and obviously calibration error.

Probably belt tension when I hear that despite using my firmware. Start with belt slack and gradually re-test and increase tension. Aim for 6 lbs tension on belt. Should be at 2-3 mm deflection of lower belt when you feel belt grow palpably tighter.

 

Posted : 30/09/2019 2:04 am
davystruyf
(@davystruyf)
Active Member
Posted by: @guy-k2

Homing is still broken AF. when X homes left during the setup wizard it just hums rapidly and then when it homes right it just starts crashing back and forth on the right side and obviously calibration error.

Probably belt tension when I hear that despite using my firmware. Start with belt slack and gradually re-test and increase tension. Aim for 6 lbs tension on belt. Should be at 2-3 mm deflection of lower belt when you feel belt grow palpably tighter.

 

I doubt it is my tension on the belt. Both belts are tight. No slack. I think they are even tighter than when i first build it. Also if they arent tight enough shouldn't the drive gear be skipping because of the lack of proper tension?

On the left it just hits the side. And hums. No grinding or skipping. Just sits there and acts like its Buddhist monk

But eh I will try to tension it up even more 

42STH40-1684MAC are the steppers by the way. Something somebody on the prusa community forum facebook group.(where i am posting videos) somebody mentioned that apparently LDO stepper motors are wire differently and need the two middle wires swapped. Dont know how much of that is true though.

 

Posted : 30/09/2019 2:15 am
christoph.e5
(@christoph-e5)
Trusted Member

No your belts are probably too tight! Start with a slack/loose/not tight belt and gradually increase tension. You could loosen (just a bit to let them spin freely on the motor shaft) the grub screws on the x and y pulleys and check if the bed and extruder are movable by hand with slight resistance. 

This post was modified 10 months ago by christoph.e5
Posted : 30/09/2019 8:43 am
christoph.e5
(@christoph-e5)
Trusted Member

But don‘t forget to retighten the grub screws after checking! I know a guy who missed a screw and had fun for hours...

Posted : 30/09/2019 8:49 am
guy.k2
(@guy-k2)
Noble Member

You ARE a bit out on your own with LDO motors. If you can't get those to work, switch to one of the ones I recommend (Moon's or OMC) in the GitHub firmware page.

M350 X8 Y8 Z16 E32 doesn't look right for a geared extruder and my firmware.

E should be 16 in my current firmware. Please do the microstepping and e-steps in specified order

Instructions on my Github spells it out.

 send M350 and M92 commands to force correct micro-stepping and e-step rates. M350 must be first
//because M350 command will sometimes alter existing M92 setting
//geared extruder, 1.8 degree motor
M350 E16
M92 E415
M500

 

This post was modified 10 months ago by guy.k2
Posted : 30/09/2019 3:51 pm
chocki
(@chocki)
Prominent Member

Just in case you experience any issues with motors ramming the axis into the end stops, you may have to drop these value by 1.

(Last two lines of the code in my case, were 4 now 3)

For some reason my Y axis just did not play nicely and kept giving me issues and when a self test was carried out, it really used to ram hard into the ends and would rattle and give length errors, I had the Y axis belt well tight, too tight, but it did not always work. Had a different motor mount etc, went back to standard, but still the same, but not as bad as before. Now with the settings at 3, the printer works perfectly, no issues even printing at 200% speed.

I'm using the OMC motors, so maybe I just happen to have one that is a bit strange, even though they have the same part numbers.

(@Guy - Note, PN Etched into motor not sticker.)

#define TMC2130_SG_HOMING 1 // stallguard homing
#ifndef X_AXIS_MOTOR_09 //Kuo
#define TMC2130_SG_THRS_X 3 // std stallguard sensitivity for X axis
#define TMC2130_SG_THRS_X_HOME 3 // std homing stallguard threshold for X axis
#else
#define TMC2130_SG_THRS_X 4 // Kuo adjust here if different needed for 0.9 degree motors
#define TMC2130_SG_THRS_X_HOME 4
#endif

#ifndef Y_AXIS_MOTOR_09 //Kuo
#define TMC2130_SG_THRS_Y 3 // std stallguard sensitivity for Y axis
#define TMC2130_SG_THRS_Y_HOME 3 // std homing stallguard threshold for Y axis
#else
#define TMC2130_SG_THRS_Y 3 // Kuo adjust here if different needed for 0.9 degree motors 
#define TMC2130_SG_THRS_Y_HOME 3 //Both were 4 now 3 ARC
#endif

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 : 30/09/2019 4:29 pm
davystruyf
(@davystruyf)
Active Member
Posted by: @guy-k2

You ARE a bit out on your own with LDO motors. If you can't get those to work, switch to one of the ones I recommend (Moon's or OMC) in the GitHub firmware page.

M350 X8 Y8 Z16 E32 doesn't look right for a geared extruder and my firmware.

E should be 16 in my current firmware. Please do the microstepping and e-steps in specified order

Instructions on my Github spells it out.

 send M350 and M92 commands to force correct micro-stepping and e-step rates. M350 must be first
//because M350 command will sometimes alter existing M92 setting
//geared extruder, 1.8 degree motor
M350 E16
M92 E415
M500

 

Well that sucks haha. I guess I will just go back to the stock motors. I definitely aren't going to buy even more motors. I already got three stock ones and having two LDO motors and a prusa stock extruder motor. 😓 

tried messing with the stallguard sensitivity. no change. it all just stays the same. tried loosening the belt nothing. every tightness has the same effect.

Uploaded a video of what is going on so maybe can have a pointer on what I could mess with to get this fixed up.

Posted : 30/09/2019 9:05 pm
guy.k2
(@guy-k2)
Noble Member

That looks like the motors are completely failing to produce a usable stallguard signal.

It's totally up to you to decide whether to get rid of VFA's in your prints with 0.9 degree motors. I simply would not be happy with the stock motor print artifacts. Yes, you have some sunk cost in the motors that won't work. A pair of OMC motors, a connector kit will put you back under $100 for a pretty big print quality improvement. Only you can decide that is unsurmountable a cost. Otherwise....

OMC Stepper Motors $20 x 2 STEPPERONLINE 0.9° OMC 17HM15-0904S Voltage: 12-24V Resistance 6.0ohms Hold Torque 0.36 Nm Current 0.9A Weight 280 gm Inductance 12.0mH

JST-PH Connector Kit $15 2.0mm JST-PH Connector Kit, with JST-PH 5/6/7 Pin Housing 6 pin connector to adapt OMC motor wires to harness. Optionally, skip these connectors and directly solder motor wires to harness I prefer to add standardized connector because I test multiple motors.

Driver Heatsinks $9 Driver Heatsinks for TMC2130 (12 pcs) Use these or similar self-stick, driver heatsinks for stock EINSY enclosure.

Stepper Motor Cables $7 YOTINO Bipolar Stepper Motor Cables, 4 x 100cm Long NEMA 17 Extended Connector Cable (XH2.54 4Pin-6Pin)

Actually follow the directions in my GitHub for proven configurations and your success odds are good. Striking out on your own with the LDO's puts you in uncharted waters. Stallguard needs appropriate motors to work. As you are discovering, it simply does not work with all motors. 

This post was modified 10 months ago by guy.k2
Posted : 01/10/2019 10:27 am
davystruyf
(@davystruyf)
Active Member

Well I have been tinkering abit last night and I got it to work mostly. Had to set the homing sensitivity to 1 which makes it home perfectly fine. the steps readouts are stable. XYZ calibration came out without a issue. Just had a single layershift on the cabin on a benchy but I wasn't there during it so I don't know if its a belt issue or the stepper motor. 

But atleast there is SOME progress. sensitivity at 2 completely fails but 1 works so I got something to work with. Feeling abit more confident now. 

Got bad advice anyway. these stepper motors are 1.64A while apparently the einsy board can only handle 1a per driver. ah man this sucks. 

Posted : 01/10/2019 2:08 pm
CybrSage
(@cybrsage)
Honorable Member

I wanted to make you aware of changes Bondtech made to their firmware.  There are several changes that are different from yours.  You can find them at the link below.

https://github.com/BondtechAB/Bondtech_Prusa_i3/tree/master/Firmware/3.8.0

Posted : 03/10/2019 9:42 pm
guy.k2
(@guy-k2)
Noble Member

Looked at their changes and only the FinalRetract is effectively different.

All the microsteps and e-steps are already as such in my firmware

Their LOAD_FILAMENT_1, and UNLOAD_FILAMENT_1 are defines that 3.8.0 firmware completely ignores. No matter what values Bondtech sets for those, nothing different happens. The actual moves are hard coded in stock Prusa firmware. Filament motion is unaffected.

On the other hand, my firmware respects those values and changing them would actually alter filament movement. Not changing unless someone reports they can't load / unload their single material MK3S or MK3.

FILAMENTCHANGE_FIRSTFEED I have changed, but it really isn't necessary because I already did the extra 10 mm in 2nd feed step. Just to match up, I moved the extra 10 mm to the firstfeed rather than the 2nd feed. Effectively, just a cosmetics change.

The only thing to meaningfully update in my firmware is FILAMENTCHANGE_FINALRETRACT -95

Such has now been done in my 0.9 branch.

Posted : 03/10/2019 11:00 pm
Page 41 / 49
Share:

Please Login or Register