RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
I will give the full reset a try, cannot hurt. I also contacted Bondtech, maybe the motor is bad.I will see if I can find, or build firmware using a bondtech upgrade firmware as the base. I will look over guy's firmware and add the bondtech parts to it.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
I just don't see anything in the firmware. Just weird.
Can we see a pict inside your extruder (with idler door open)?
Is there any way to assemble a BMG with 1.5:1 gearing?
I would never claim to be an expert in firmware, but I am looking over your logic path and found this, which is different from all the other extruder logic paths. All the others have a #ifdef line. I also do not see where the Z axis stepping is set. EDIT: I have also attached the firmware folder from Kodiac's Bondtech upgrade firmware, found here https://github.com/BondtechAB/Bondtech_Prusa_i3
Parts in question from your firmware I mentioned:
#ifndef EXTRUDER_DEFS_SET //Kuo for e-axis msteps
#define DEFAULT_AXIS_STEPS_PER_UNIT {100,100,3200/8,280} //default steps/unit e-axis
#define TMC2130_UNLOAD_CURRENT_R 12 //lower current for M600 to protect filament sensor with stock extruder
#define EXTRUDER_DEFS_SET 1
#endif
Here is the portion for the BMG
#ifndef EXTRUDER_DEFS_SET //Kuo for e-axis msteps
#ifdef BMG_EXTRUDER
#define DEFAULT_AXIS_STEPS_PER_UNIT {100,100,3200/8,830} //BMG approx 3:1 geared extruder
#define TMC2130_UNLOAD_CURRENT_R 20 //BMG unload current for M600
#define EXTRUDER_DEFS_SET 1
#endif
#endif
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Too late to edit this into my last post, but I already copied the variant file into the firmware folder and renamed it. So when you see it you will not be confused. Oh, and the codiac github where I snatched it from is controlled by Chris Warkocki.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
I wish I could edit long afterwards...but since I cannot, here is the picture inside the extruder housing. With it zoomed in so much, I see I need to do some cleaning!
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Also, I do not see where the USteps_E is defined for the BMG Extruder. I added this:
#ifndef BMG_EXTRUDER
#define TMC2130_USTEPS_E 32
#else
#define TMC2130_USTEPS_E 32
#endif
1/32 microstepping gives us 1660 steps for a 0.9 degree motor and 830 for those with a 1.8 degree motor. If the microstepping was set to 1/16, that would explain why I need to set the steps to 415 to get the proper line thickness. I created new firmware with those lines and installed it, ran the estep gcode, power cycled after 10 seconds, and printed the cube again...but it gave me the same double thickness as before. 🙁
I also just heard back from Bondtech Support and they said the machine is definitely running 1/16th instead of the default 1/32...but I obviously have no idea how to make that change.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Its been awhile since I have had a chance to add my X 0.9 moons back to my machine and test.... Since then, I have added it back and successfully had the X axis homing perfectly every time. I printed some towers, benchys, and x/y/z cubes just to test and fine tune.
My issue now is with really sloppy retractions on the LDO 0.9 on the bondtech extruder. Even though the 0.9 is more consistent extrusion, it has way worse retractions then the moons 1.8 I had on it prior. Has anybody tried linear advanced 1.5 with the 0.9's yet? I have heard nothing but good things about it and it has totally changed the way my stock Mk3 is printing so far.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Also, I do not see where the USteps_E is defined for the BMG Extruder.
It's later in the TMC2130 section of configuration_Prusa.h.
I have simplified the logic there in case the compiler is choking on nested if's. If E_AXIS_MOTOR_09 remains commented out (undefined),
#define TMC2130_USTEPS_E 32 should be compiled.
Attached is another test firmware for you. Decompress to get the hex. After flashing, perform usual setting of esteps to 830 and printer off/on before test printing
Also be sure to remove any e-step setting codes in test objects you created. That would really confuse things.
The extrusion test STL's I made for you do not have any e-step setting gcodes.
The e-steps setting section of the test firmware is as below...
/*------------------------------------
TMC2130 default settings
*------------------------------------*/
#define TMC2130_FCLK 12000000 // fclk = 12MHz
// Kuo independently define x y and e microsteps for stepper type
#ifndef X_AXIS_MOTOR_09
#define TMC2130_USTEPS_X 16
#else
#define TMC2130_USTEPS_X 8 // Kuo reduce X microsteps to 8 because EINSY cannot keep up with 16 on 0.9 degree motor
#endif
#ifndef Y_AXIS_MOTOR_09
#define TMC2130_USTEPS_Y 16
#else
#define TMC2130_USTEPS_Y 8 // Kuo reduce Y microsteps to 8 because EINSY cannot keep up with 16 on 0.9 degree motor
#endif
#ifndef E_AXIS_MOTOR_09
#define TMC2130_USTEPS_E 32
#else
#define TMC2130_USTEPS_E 16 // Kuo reduce Y microsteps to e-axis
#endif
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
My issue now is with really sloppy retractions on the LDO 0.9 on the bondtech extruder. Even though the 0.9 is more consistent extrusion, it has way worse retractions then the moons 1.8 I had on it prior.
Unless it is a full size 0.9 (not a pancake), it is not going to have enough torque to get good resumes after retractions.
Which LDO 0.9 were you testing?
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Cybersage, after all is done and installed
What does M503 report in the Pronterface terminal window?
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
My issue now is with really sloppy retractions on the LDO 0.9 on the bondtech extruder. Even though the 0.9 is more consistent extrusion, it has way worse retractions then the moons 1.8 I had on it prior.
Unless it is a full size 0.9 (not a pancake), it is not going to have enough torque to get good resumes after retractions.
Which LDO 0.9 were you testing?
It is a LDO 0.9 pancake for the bondtech extruder. I have heard that the Linear advanced 1.5 clears up those retraction issues with that setup (modified firmware fork available) and that the clicking from retractions is non-existent. I have had some really impressive results with the LA 1.5 firmware on my stock machine with 1.8 steppers. I didn't know if this is something that somebody has tried with the 0.9 on the X,Y and the extruder yet? Of course, this would probably require merging firmware or modifying it even more...
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Took a peek. It's well beyond what can be done with automated merging. I would first go back to 1.8's on X and Y and test the LA 15 firmware with your 0.9 LDO pancake and see if you can get clean retractions. The X and Y 0.9 degree motor changes are independent of the extruder stuff. Once you know LA 15 firmware works with the 0.9 pancacke, you can manually patch in changes for X and Y 0.9 degree motors.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Cybersage, after all is done and installed
What does M503 report in the Pronterface terminal window?
I ran the 503 code both before and after running the esteps gcode, just to see what it would say...I stripped out the relevant portions. It looks like running the gcode is not needed since you already define the 830 in the firmware.
Before I ran the esteps gcode:
>>> M503
SENDING:M503
echo:Steps per unit:
echo: M92 X100.00 Y100.00 Z400.00 E830.00
echo:UStep resolution:
echo: M350 X8 Y8 Z16 E16
After I ran the esteps gcode:
>>> M503
SENDING:M503
echo:Steps per unit:
echo: M92 X100.00 Y100.00 Z400.00 E830.00
echo:UStep resolution:
echo: M350 X8 Y8 Z16 E16
I am running the calibration block now, will post results in about 20 mins.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
The cube thickness is 0.8mm, give or take a tiny amount.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Is M503 (set usteps) still supported in stock firmware? I recently took the plunge on a stepperonline pancake for my skelestruder; contemplating plonking down some cash on a 17HM15-0904S or two for X/Y ... but I'd much prefer to not have to tweak/customize firmware as it makes future upgrades more of a hassle.
Is that OMC motor still also the best bang for buck people have found? I seem to recall it had been as good as the Moons when linearity correction was enabled.
M350 should be, according to this page, but it's quite outdated.
https://github.com/prusa3d/Prusa-Firmware/wiki/Supported-G-codes
Please forgive me if this has been addressed before - I've been keeping a half-hearted eye on this thread but the search function is somewhat lackluster and I don' fancy re-reading 20 pages here + umpteen more in the predecessor threads... :-/
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
AnatomicFlack, those look really very nice. There is a little bit of residual VFA, but your results are already about the best we can get out of the 0.9 motors we are using.
The pulley belt artifact is VERY faint, broader vertical bands with 2mm horizontal spacing - which is consistent with our 2 mm spacing of belt and pulley teeth. Usually needs a flat surface about 5 cm or longer before you reliably find it.
Ok, good to know, I'm really close. I thought I was missing something obvious.
It'll be interesting to see how/if the Misumi MR2 pullies and belts alter the VFAs when they arrive.
Printing in any filament besides ones meant to highlight all the surface flaws and VFAs looks astounding right now. I'm printing up a new set of Y-axis parts so I can get back my physical bed stops and use the E3D idler. ( https://www.thingiverse.com/thing:3381517 ) While searching I also found this Y motor mount with stop and shaft bearing support. https://www.thingiverse.com/thing:3562828
I'm currently using the Bear bearing holders, but I think I'm going to swap to these as they should be a LOT more rigid in holding the bed without having compromise solid bed attachment and securing the bearing with one single pair of screws ( https://www.thingiverse.com/thing:3342147 ). As a bonus for me, I can use them with my SKF bearings that have no notches in the bearing housing, so I modified it in Fusion to accept a smooth bearing. The only real issue is that it raises the bed up 3.3mm from stock, which is going to negatively affect the angle of the belt. Going to have to play with belt holders and make sure I'm still able to hit the stops.
Out of curiosity, what Y axis parts are people here using on their bear, or stock frame? It always seems like the Y axis options are overwhelming. And seems that when trying to eliminate VFAs it's a solid first step to pick the right parts.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Out of curiosity, what Y axis parts are people here using on their bear, or stock frame? It always seems like the Y axis options are overwhelming. And seems that when trying to eliminate VFAs it's a solid first step to pick the right parts.
I am using aluminum rod holders on my bear that have set screws instead of zip ties.I modified a version of this https://www.thingiverse.com/thing:3502543 and added some small RC car bearings with a 5mm brass tube for the Gates pully. Similar concept as the "risky business" mod but with smaller bearings. It is nice, smooth and perfectly centered. For the motor mount I went with https://www.thingiverse.com/thing:3562828 It has a built in endstop and a support bearing.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Cybersage, after all is done and installed
What does M503 report in the Pronterface terminal window?
This is also just to report in that I am using your firmware just with X, Y, and Bondtech defines and it works fine at 830 e-steps. This issue is really bizarre.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
I simplified the BNB firmware to remove all the X Y and E decision branches and replaced them with static settings for X and Y of 0.9 and a BMG on E. The compile check was good so I uploaded it to the printer. It reports 32 microstep resolution and I am now running the cube again.
I forgot to do a factory reset, if this does not work, I will do it and try the unchanged (well, other than the required changes) firmware again to see the results. Will know in a few minutes.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Cybersage, after all is done and installed
What does M503 report in the Pronterface terminal window?
This is also just to report in that I am using your firmware just with X, Y, and Bondtech defines and it works fine at 830 e-steps. This issue is really bizarre.
Do you still have the firmware folder, with the changes made (selecting the 0.9 motors, language settings, etc)? If you do, zip it up for me and I will do some compares against mine when I create it per the instructions.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
echo: M350 X8 Y8 Z16 E16
That is very strange indeed. Your microsteps for E are somehow stuck at 16 instead of 32 despite my firmware setting it to 32.
When the printer does not obey the firmware commands, we're left with only two things.
1. There is something in the printed gcode setting e-steps. NO. This is happening even with a gcode file I sliced which definitely does not have such an entry.
2. Data in EEPROM is corrupt and interfering with settings. Repeat a full factory reset WITH deletion of all data.
I don't have an answer beyond the above.