Distortion of Round Features  

Page 1 / 2
Eminent Member

Hi Folks!

Last night I did a print of what is a simple cylinder (on the outside anyway) and the result had a very strange lump on either side, 180 degrees apart from each other, but more noticeable on one side.

If it wasn't for the fact that it is on both sides I would have assumed that it is where the printer stepped up to the next layer, but I don't think even that would explain this path profile.

I am used to seeing scars along the line where z is increased on my other (non-Robox) printer, but this looks different. Prettier for sure, but more troublesome from a mechanical point of view.

Print was done on the mid-range quality setting, with ABS.

I was also seeing distortion on small vertical cylinders (8mm diameter), which were up to 0.5mm wider in one direction that the other, which I am now guessing is related.

Any clues? Am I just doing something dumb?

Posted : 11th June 2018 2:35 pm
USA Robox Support Moderator

@ossum Your belt in that direction isn't tight or is failing. Look at the belts and make sure they aren't cracked. A little white on the narrow sides is OK, but white showing on the tooth sides or cracks showing white on the smooth wide face are not OK.

The belts should be tight enough that the bed and head will not move side to side without taking the belt with it. If you can wiggle the bed or head front to back or side to side without the belt moving the belts are the issue. I suspect based on experience that you have one belt that is either failing or woefully under-tensioned.

If you aren't sure how to check the belts, please open a support ticket.

Posted : 11th June 2018 4:55 pm
Eminent Member

Thanks for your quick response, sorry my followup was so slow! I had to wait for a 16hr print to finish and then I did another test print of the same item because I had a nagging worry that I hadn't secured the clip on the PEI sheet (turns out I had, results are the same).

Your explanation makes perfect sense, since there would be two points in a circle where backlash would show up as a bump. I'd expect then the bump to be on either the right/left extremes or front/back extremes though. I've attached a image where you can see they are not at either, so perhaps this is slicing after all...

The belts appear to be tight though, I can't wiggle the bed or head without the belts. One thing I did notice is that that moving the head left to right is very stiff, while the bed is easy to move back and forth (obviously I tested this without the motors being energized).

I can open a support ticket if that is preferable, I figured a forum post is more useful to other users in the future.

Posted : 13th June 2018 10:18 am
USA Robox Support Moderator

@ossum A support ticket will get you specific information for your printer and is always the best way to get help. The support forums are fine for some problems, but are always use at your own risk. A support ticket will provide only CEL approved information.

Based on the orientation of the flats, I suspect your problem is with the X direction. Please check the tension on the belt. Can you easily push the top section of the belt down to the bottom section? When you push the top section of the belt down to the bottom section, does the head move? On my printers even with the head moved all the way to one side it takes a fair amount of force to get the top pushed down to the bottom. My fingers aren't calibrated, but I would guess about .5 kg force.

If the belt is easy to push down or the head doesn't move when you push the top section of the belt down, the belt needs to be tightened.

Another test is to run a square test, then measure the results on each orientation. This will tell you if you have a problem in one axis and in which axis it is. My Robox units print squares with about 0.1 mm difference between the axes at most and most of them are within 0.05 mm.

Posted : 14th June 2018 5:37 pm
Eminent Member

Thank you!

The problem was indeed with the pulley, I have glued it and the cylinders are almost perfectly round. Huzzah!

I will try the square test shortly, I did notice before on a calib cube that I had quite a difference between the axes, so I will be getting back to that now. This is a huge improvement already. I expect I will have much better luck getting bearings to sit snugly on printed shafts now!

Posted : 14th June 2018 7:58 pm
Eminent Member

I have done a square test (xyz calib cube 20mm) and it makes me think I may need to glue the other pulley too. It seemed tight but maybe has enough backlash to introduce the 0.3mm innacuracy in Y.

The white one is most recent, all the others were from before the pulley issue (maybe even from before it worked loose, impossible to say for sure).

Before I learned about the loose pulley problem I was looking into steps/mm as the cause for my undersized cubes. I printed a 100x100mm square and it was only 99.6 in one direction. At the time I couldn't figure it out because the error was not proportional to the increase in size of the square, now it seems to make much more sense.

Posted : 15th June 2018 8:32 am
USA Robox Support Moderator

@ossum I would tighten up that motor in its mount too. I can see the mount screws loose.

I have never seen a motor pulley come loose - they are molded with a D-shaft opening so there isn't a set screw in them. I don't know that what you have done would solve it, based on the photo.

Posted : 15th June 2018 11:32 pm
Eminent Member

I also wondered about those motor screws, only noticed it in the photo unfortunately, I'll get back in there and tighten it up.

The pulley was prevented from spinning completely by the D shape but there was some movement, I could easily wiggle it around a degree or two wihout the motor shaft moving.

I've done more calib cubes, at the same time, one with each nozzle, to compare PETG to remove shrinking from the equation and one with ABS, but the results are clear as mud to me... can't quite figure it out.

Posted : 16th June 2018 8:25 am
I am Pete Admin

Shrinkage and scaling along with backlash and loose bed/belt issues is a recipe for confusion.

I would stick everything back to default (see below) and apply scaling to the actual part you want to print.

This chart is handy (source:

Because the shrinkage is not constant and predictable like in injection moulding it will vary based on design, ambient conditions, print order and print density. I find that X and Y scaling needs to be adjusted but Z is very close to 100%.

Eg for ABS you might pick something in the middle of the range 1.2%

So your scaling becomes X101.2% Y101.2% Z100%

Name of the polymerExplicit name of the polymerMin Value(%)Max Value(%)
ABSAcrylonitrile-Butadiene Styrene0.71.6
ABS FRAcrylonitrile-Butadiene Styrene flame retardant0.30.8
ABS High HeatAcrylonitrile-Butadiene Styrene High Heat0.40.9
ABS High ImpactAcrylonitrile-Butadiene Styrene High Impact0.40.9
ABS/PCAcrylonitrile-Butadiene Styrene/Polycarbonate0.50.7
ABS/PC 20% GFAcrylonitrile-Butadiene Styrene/Polycarbonate 20% glass fiber0.20.3
ABS/PC FRAcrylonitrile-Butadiene Styrene/Polycarbonate flame retardant0.30.6
Amorphous TPI Blend, Ultra-high heat, Chemical Resistant (High Flow)Amorphous TPI Blend, Ultra-high heat, Chemical Resistant (High Flow)0.81
Amorphous TPI Blend, Ultra-high heat, Chemical Resistant (Standard Flow)Amorphous TPI Blend, Ultra-high heat, Chemical Resistant (Standard Flow)0.81
Amorphous TPI, High Heat, High Flow, Transparent, Lead-Free Solderable (High Flow)Amorphous TPI, High Heat, High Flow, Transparent, Lead-Free Solderable (High Flow)11.2
Amorphous TPI, High Heat, High Flow, Transparent, Lead-Free Solderable (Standard Flow)Amorphous TPI, High Heat, High Flow, Transparent, Lead-Free Solderable (Standard Flow)11.2
Amorphous TPI, Moderate Heat, TransparentAmorphous TPI, Moderate Heat, Transparent0.50.7
Amorphous TPI, Moderate Heat, Transparent (Food Contact Approved)Amorphous TPI, Moderate Heat, Transparent (Food Contact Approved)0.50.7
Amorphous TPI, Moderate Heat, Transparent (Mold Release grade)Amorphous TPI, Moderate Heat, Transparent (Mold Release grade)0.50.7
Amorphous TPI, Moderate Heat, Transparent (Powder form)Amorphous TPI, Moderate Heat, Transparent (Powder form)0.50.7
ASAAcrylonitrile Styrene Acrylate0.40.7
ASA/PCAcrylonitrile Styrene Acrylate/Polycarbonate0.30.7
ASA/PC FRAcrylonitrile Styrene Acrylate/Polycarbonate flame retardant0.40.8
ASA/PVCAcrylonitrile Styrene Acrylate/Polyvinyl Chloride0.30.7
CA - Cellulose AcetateCellulose Acetate0.31
CAB - Cellulose Acetate ButyrateCellulose Acetate Butyrate0.20.9
Celllulose Diacetate- Pearlescent FilmsCelllulose Diacetate- Pearlescent Films11.5
Celllulose Diacetate-Matt FilmCelllulose Diacetate-Matt Film11.5
CP - Cellulose ProprionateCellulose Proprionate0.10.9
CPVC - Chlorinated Polyvinyl ChlorideCPVC - Chlorinated Polyvinyl Chloride0.30.7
ETFEEthylene Tetrafluoroethylene34
EVAEthylene Vinyl Acetate0.43.5
FEPFluorinated Ethylene Propylene36
HDPE - High Density PolyethyleneHDPE - High Density Polyethylene1.54
HIPS - High Impact PolystyreneHIPS - High Impact Polystyrene0.20.8
HIPS FR V0High Impact Polystyrene flame retardant V00.30.6
LCPLiquid Crystal Polymer0.10.6
LCP CFLiquid Crystal Polymer carbon fiber0.10.5
LCP GFLiquid Crystal Polymer glass fiber0.10.4
LCP MINERALLiquid Crystal Polymer mineral0.10.5
LDPE - Low Density PolyethyleneLDPE - Low Density Polyethylene24
LLDPE - Linear Low Density PolyethyleneLLDPE - Linear Low Density Polyethylene22.5
MABSTransparent Acrylonitrile Butadiene Styrene0.40.7
PA 11 30% Glass fiber reinforcedPolyamide 11 30% Glass fiber reinforced0.50.5
PA 11 conductivePolyamide 11 conductive0.72
PA 11 flexible