Small round holes/parts not rounds

[jet 0]

For anything smaller than about 15mm any round shape i try to print is very noticeably not round, whats going on?

[owen 19]

For anything smaller than about 15mm any round shape i try to print is very noticeably not round, whats going on?

Hi

maybe loose belts

[jet 0]

the belts are pretty tight....but ill try making them tighter, it seems to consistent to be the belts, all the holes seem to be kinda of squashed, but in exactly the same way on every layer, so that the walls are smooth....its weird, ive had a fair bit of experience with cnc routers, and never seen anything like this

[italianmaxx 0]

hi,

I am still waiting for my UM but I have a good experience with cnc.

Verify if the non-roundness is the same if you print on different place of the base, verify if there is non-roundness also in multiply of the measure you told (15-30-45), verify again if the calibration (mm/step) is correct at different scale objects, verify if the mechanical movement is fluid. at the end you could think about swapping the x and the Y motors to see if the squashing is changing axis, try to plot different gcode ( I mean use a pen and not the PLA/ABS) and use simple gcode for different shapes of different sizes. (a simple mach 3 demo could be useful to obtain the gcode)

last but not least: good luck!

Come on delivery boy...I'm waiting and I'm starting to be nervous... :lol:

[Daid 306]

the belts are pretty tight....but ill try making them tighter, it seems to consistent to be the belts, all the holes seem to be kinda of squashed, but in exactly the same way on every layer, so that the walls are smooth....its weird, ive had a fair bit of experience with cnc routers, and never seen anything like this

Some photos might help in this case.

[mm_build 1]

the belts are pretty tight....but ill try making them tighter, it seems to consistent to be the belts, all the holes seem to be kinda of squashed, but in exactly the same way on every layer, so that the walls are smooth....its weird, ive had a fair bit of experience with cnc routers, and never seen anything like this

Verify that the small belts connected directly to the steppers are tight, if they do not already have washers they really should.

Also, does it print squashed holes both vertically (through the z layers), and also horizontally (x and y laying flat)?

[jet 0]

sorry i havnt had a chance to take photos yet.

the z layer holes are perfect

im thinking at the moment it must be somthing to do with the way skeinpypy generates the path.....ill take a photo soon

[jet 0]

ok so i printed the quadcopter arm to test the strength of pla parts, after breaking it, this is what the cross section looks like. Its like this extremely consistently, all the way through the part, the black arrow shows where the wall is so distorted it doesnt even meet up with infill, and the blue shape exaggerates the distortion of the inside wall of the tube, while the red shape exaggerates the distortion of the outside wall of the tube.

Ive double checked the belt tensions, and there all fine, I find it extremely weird that the inside and outside walls are not distorted in the same direction, and that its distorted in exactly the same way on every layer

[mr_seeker 0]

I heard that problem before from someone, so I asked for the g-code and tried the same print. Didnt had any problems with it. Same firmware, so I am thinking it has something to do with the steppers not getting enough power or some other problem.

[Daid 306]

Ah, the inner circle escapes problem. Printing slower really helps in this case. Also, a proper feedrate, you might have a bit of under-extrusion. If I look at those walls they are not connecting to the infill very well.

[glasswalker 0]

I used to consistantly run into circles out of round on my reprap (Sells Mendel)

The circles would come out in a slightly oval shaped pattern on 45 degrees from the X or Y axis. And sometimes infill wouldn't line up with the skin due to the offset.

It was caused 100% by backlash.

So loose belts aren't too far fetched. (my issue was EXACTLY consistant every time, it was out by about 1mm total from round, so larger circles you didn't notice it, really tiny holes looked badly squashed)

[almi 0]

Hi all,

I can emphasize the negative influence of loose belts on round circles.

As you can see on the photo (at least I hope can see it) there is a straight curb in the Z-Axis.

This position is where the Y-Motor starts turning back its direction. If the Y-Motor belt is loose, it will not do the right circle movement here at that point (going too much in X direction).

I have another question. Does anybody know where the left line in the picture comes from?

It is not related to circle movements, I have seen this on rectangles as well.

It is the point where a Z-Movement for the next layer takes place.

Could it be that while moving up the x,y position steps are also changed?

Thanks for any help and Cheers

almi

:

[Daid 306]

It's the ZBlob!

It happens because the printer head moves up and then there is still some plastic flowing out of it. A lower printing temperature reduces this ZBlob. But I don't know of a way to fully prevent it, except for "Joris", but that's not an option for your current object.

[joergen 2]

It's the ZBlob!

It happens because the printer head moves up and then there is still some plastic flowing out of it. A lower printing temperature reduces this ZBlob. But I don't know of a way to fully prevent it, except for "Joris", but that's not an option for your current object.

yes, the dreaded zblob!™

netfabb 4.9 introduced the layer start randomizer, which will prevent getting the blobs all in a row, but depending on your temp setting, will distribute the blobs all over the surface, which, when seen as a feature, adds more grip to objects when held in your hand. or you could try lowering the temp.

[mr_seeker 0]

It's the ZBlob!

It happens because the printer head moves up and then there is still some plastic flowing out of it. A lower printing temperature reduces this ZBlob. But I don't know of a way to fully prevent it, except for "Joris", but that's not an option for your current object.

Retraction!

[devoney 0]

I have another question. Does anybody know where the left line in the picture comes from?

Would be the Z-axis moving up I guess. Netfabb has a function to randomly 'seam'. It moves up on a different position every time. So the ZBlobl does not become a blob but a more a speckle randomly divided on your object.