Vertical Banding Problem and Stepper Waveforms
-
With the gates curvilinear profile (GT, GT2, GT3, etc), it's acceptable to run the teeth side against a smooth idler if it has a diameter equal to or greater than the effective diameter of a 40 tooth pulley of the same tooth pitch.
With 2 mm pitch belt, this equates to a smooth pulley with about 26 mm diameter.
(Source: Pg 174 https://www.gates.com/~/media/files/gates/industrial/power-transmission/manuals/powergripdrivedesignmanual_17195_2014.pdf?la=en))
*not actually a robot
-
I checked the D-Bot build guide and F623ZZ flanged bearings are the recommended idlers and I'm sure that's what I have. They have a diameter of 10mm which explains the artifacts. I'll pass it on so hopefully the Bill of Materials can be updated. Thanks Bot!
-
Hello, I have the same Issue on my P3Steel with a Duet WiFi, so I thought I'd share my experiences. For me, they're more noticeable along the Y-Axis (maybe a problem with the bed?) than along the X-Axis, but it's still there on both.
I too have an e3d Titan Extruder (with a 0.9° Stepper) and played around with the extrusion, but it didn't change much.
I tried various acceleration, jerk and speed settings. Also I don't seem to be getting those ripples on curved surfaces.
I also tried printing without any fan running on the Printer (Had to unclog my Hotend after that…) but it didn't help.
Another thing I tried was putting astrosyn dampeners on all motors, but that also didn't help.
Then I tried something "just for fun" and replaced the duet WiFi with an Ultimachine RAMBo 1.3, running the latest Marlin Firmware with the same Motion settings that I settled on in the Duet Firmware. Now it's still happening but its much less noticeable. On the Duet print I could easily feel the ripples when running over the surface with my fingernail, and on the RAMBo print I can't.
Very odd. Are you sure you were using the same motor currents on both controllers? What motors are you using?
-
Hello, I have the same Issue on my P3Steel with a Duet WiFi, so I thought I'd share my experiences. For me, they're more noticeable along the Y-Axis (maybe a problem with the bed?) than along the X-Axis, but it's still there on both.
I too have an e3d Titan Extruder (with a 0.9° Stepper) and played around with the extrusion, but it didn't change much.
I tried various acceleration, jerk and speed settings. Also I don't seem to be getting those ripples on curved surfaces.
I also tried printing without any fan running on the Printer (Had to unclog my Hotend after that…) but it didn't help.
Another thing I tried was putting astrosyn dampeners on all motors, but that also didn't help.
Then I tried something "just for fun" and replaced the duet WiFi with an Ultimachine RAMBo 1.3, running the latest Marlin Firmware with the same Motion settings that I settled on in the Duet Firmware. Now it's still happening but its much less noticeable. On the Duet print I could easily feel the ripples when running over the surface with my fingernail, and on the RAMBo print I can't.
Very odd. Are you sure you were using the same motor currents on both controllers? What motors are you using?
Yes, 800mA. I was using These Nema 17s from Robotdigg:https://www.robotdigg.com/m/product/1137/42HS40-1704-58-oz.in-stepper-motor#
I generally had good experiences with robotdigg Motors and I've used them in a different Duet equipped Machine before, and didn't have a Problem. Maybe it was just a coincidence that it came out better on the RAMBo, I didn't really do any further Testing on it because I couldn't directly compare everything in the Firmware since I only know enough about Firmware to configure it for a Printer and not much more.
I can easily install the RAMBo again with different Settings if that helps.
While searching around I found this thread in the reprap forum:http://forums.reprap.org/read.php?246,702962,723107 it's in German but they're having the Same problem, and their Solution was to have the left End of the X-Rods fixed and the other End slightly loose in the right End (Like fixed/floating bearings), so my Right x-end might be too tight. I'll try printing a new one with a looser fit, and I'm going to drill out one side of the y-rod holders in the frame.
-
I checked the D-Bot build guide and F623ZZ flanged bearings are the recommended idlers and I'm sure that's what I have. They have a diameter of 10mm which explains the artifacts. I'll pass it on so hopefully the Bill of Materials can be updated. Thanks Bot!
I certainly twist my belts when going over smooth idlers which are small diameter like 16t/20t pulley equivalents, I do this on my kossel XL. My corexy uses 608 bearings as idler pulleys which are 22mm outside diameter so I don't see these artefacts. I notice a lot of corexy designs use small pulleys, sometimes for necessity/packaging reasons, but often I suspect for aesthetic reasons. You could always consider replacing them with toothed idlers, especially if twisting belts isn't possible as it would limit axis travel - they're not quite as readily available but you can buy them.
-
I also twisted my belt on my CoreXY, as I could find toothed idlers for 3GT/9mm at reasonable price.
BTW, I don't see how it reduces axis travel… It is only twisted between motors and rear idlers, where length does not change.
-
Depends on the design, if I twisted belts on my modified smartrapcore I would not be able to move the y carriages close enough to the pulleys to print on all of the bed without frcing the twisted section into far too small length, which I am sure would have an effect on positional accuracy. I think it depends how much build volume you try to squeeze out of a given size corexy.
-
Do you have a picture of the your modified smartrapcore belts path?
-
But as I said twisting the belts here doesn't help me as I don't get the repeating ripple artefact with my big pulleys.
-
Oh all interesting. I just logged in to post that I feel like I found a major contributing factor to the problem. Perhaps Denos has found the other factor.
In trying to rule out vibration of the linear shafts themselves (see photos above for my gantry arrangement) I again ran a print with a part in the center and in each of the four corners. Interesting result: The amplitude is reduced in the corners, but the period of the ripple didn't change. If my linear shafts were vibrating I'd expect higher frequency ripple closer to the edges of the bed. If it's mechanical vibration at all, it would seem to be something local to the toolhead.
Last night I found that if I push my finger against the lower part of the heatsink of the Titan extruder during a print, the ripples pretty much go away. I suspect the heatsink is vibrating, which may explain why the ripple frequency is independent of position on the build plate. It may also explain the reduced amplitude by the edges, but that would require too many words to write right now.
I ordered the Titan Aero upgrade kit and am redesigning things to suit. I kind of wanted to do that anyway, so if it wasn't the entirety of the problem I'll be OK with that. Will report back when that's working.
But clearly something is causing the heatsink to vibrate to begin with. I have a standard (non-Aero) e3d Titan on a work printer and that shows no ripple whatsoever.
Denos, youre test with the idlers is most excellent. I have toothed idlers for my GT2 belts, BUT, the quality control is bad: the holes for the bearings are clearly too small for the press fit, and they grind noticeably when rotated. I installed them anyway because I really didn't think it would cause trouble, but perhaps that's the source of the extraneous vibration that's driving the vibration in the heatsink. Eureka?
The frustrating thing is, I can't find a reliable (i.e. not super cheap Chinese) source of toothed GT2 idlers. Anyone know? Even good smooth ones seems hard to find. Toothed is preferable for me as my gantry style doesn't lend itself to half-twisting the belts.
EDIT:
The toothed idlers I used are from RepRap Champion (not as cheap as I recalled), so, not good.
~Justine
-
If the toothed Chinese idlers do not appeal, then consider using large diameter bearings like the F608ZZ bearings I use. Running the belts on the toothed side is not a problem, certainly not for print quality.
-
An update:
After changing the idlers and switching to the Titan Aero, the ripples are exactly the same, though possibly a bit more crisp.
I'm going back to blaming some kind of variation in the extrusion speed. Could the large (.8mm) nozzle have something to do with it?
I'm in the process of making a second toolhead with a .5mm nozzle.
This weekend going to redo all the motor wires using shielded twisted pairs too.
-
What hotend are you using, with the 0.8mm nozzle? You're right, and it can be related, if the filament does not have time to melt. Do you see any temperature variations?
-
@justine-haupt were you able to resolve the vertical banding? Did you find the reason for the vertical banding? I have a similiar issue.
-
I have been having this problem for a while as well. My theory is that the motors create a vibration that is transferred to the frames and to the z axis, this takes place at a micron level and the vibrations cannot be noticed by the eye.
Most Industrial machines reduce vibrations by increasing the weight of the frames by adding concrete in them or using a strong concrete foundation. Even the older stratasys machines were built like tanks and they used nema stepper motors as well.
Vibration meters can be used to measure the vibration of the frames while printing and then stiffeners or weights can be used to increase the rigidity of the machine. This should hopefully reduce the effect of vertical banding.
Not sure if this is even the correct method.
Thanks,
Hussain. -
This is quite interesting, Would like to understand more about this.
The attached link is not working though.
Could you guide me to any other resource that explains this calculation.Thanks
Hussain. -
@hussainsail2002 said in Vertical Banding Problem and Stepper Waveforms:
This is quite interesting, Would like to understand more about this.
The attached link is not working though.
Could you guide me to any other resource that explains this calculation.Thanks
Hussain.Yes, the link seems broken but I found a new link. It was to a PDF version of the "Gates GT3 Design Manual" in case the link gets broken again.
https://www.gates.com/resources/resources-library/catalogs/power-grip-gt3-drive-design-manual
