Prints are coming out... stange?

[Azurial 0]

alright, so ive been haveing a great time printing with my ultimaker original but there seems to be an issue where i think the rods are out of alignment. ive only been printing for about a month or so with this machine and while it can produce great prints, lately its been giving me prints that are drifting only on in the front (when facing the printer head on,  z-axis is in the rear) of the printbed and to the left a bit. on smaller prints it doesn't seem to do anything, but the further out of the center of the machine it starts to go haywire. i've included pictures of a recent print.

i've already checked the belts, pullies(even replaced them) and they all seem to look fine. they are bolted in tight and the belts are tensioned well. ive switched over to GT2 betls vs the original ones(open loop) and using the original blocks to keep them closed with a good amount of force. i have no real reliable way of checking the rods, so any help you guys could offer would be awesome. thanks in advance!

Edited April 3, 2016 by Guest

[neotko 1,417]

This for me looks like a dead coupler or too hot + underextrusion. Check the health of your ptfe coupler and check the hotend for leaks.

[gr5 2,173]

Wow - that's a lot of stringing! Do you have retraction checked? What kind of filament is this? Maybe you should lower the temp by 10C. Anyway that's not what you asked about.

I don't think you have a problem with friction/pulleys/belts because whatever "errors" seem to happen on some layers seem to recover on higher-up layers, right? And it's pretty much impossible to have a hardware error like that fix itself. So I'm thinking this is a mixture of your cad program and/or cura settings. I mean the 4 holes are all lined up nicely, right? or are they not supposed to be?

Did you look at the "bad" layers in layer view in cura? I think you will see cura is telling the printer to do that stuff. if so look at the model in xray view and fix anything that you see as red. If nothing is red maybe shows us your gcode file and your stl file (upload them somewhere).

[Azurial 0]

This for me looks like a dead coupler or too hot + underextrusion. Check the health of your ptfe coupler and check the hotend for leaks.

 

thanks. changed out the ptfe before a print and it came out alright but it was a smaller print. as far as the coupler ive switched over to thermistors and a ramps board after my ultiboard blew out. i know they are less accurate, but the setup im running now is just cheaper for me to replace as i have tons of pcbs and access to free ramps.

 

Wow - that's a lot of stringing!  Do you have retraction checked?  What kind of filament is this?  Maybe you should lower the temp by 10C.  Anyway that's not what you asked about.

I don't think you have a problem with friction/pulleys/belts because whatever "errors" seem to happen on some layers seem to recover on higher-up layers, right?  And it's pretty much impossible to have a hardware error like that fix itself.  So I'm thinking this is a mixture of your cad program and/or cura settings.  I mean the 4 holes are all lined up nicely, right?  or are they not supposed to be?

Did you look at the "bad" layers in layer view in cura?  I think you will see cura is telling the printer to do that stuff.  if so look at the model in xray view and fix anything that you see as red.  If nothing is red maybe shows us your gcode file and your stl file (upload them somewhere).

 

retraction is checked, and its 1.75 mm translucent pla from a cheap $20 usd 5 lbs roll. i have better filiment but i use this stuff for test prints and they come out alright.

the layer problem does seem to sort itself out here and there, i kinda thought it could be a skipping issue, but after tensioning the belts the problem only seemed to be reduced and not eliminated. when i touch them they are nice and firm. the file im using came from thingiverse which could explain some compatibility errors, and while i suspected it could be a slicing problem, i never thought to check the layers in cura.

ill load up the exact same print and see if i can replicate the error.

im printing strait from the pc if that helps any?

Edited April 3, 2016 by Guest

[Azurial 0]

UPDATE: Looking at the Layer view the stringing going on is coming directly from the toolpath of Cura. thanks for the heads up!

Well... still have localized drift going on. Any ideas?

Edited April 4, 2016 by Guest

[peggyb 170]

check if the fan underneath the electronics board is running. It can loose steps when they overheat, usually in the first few layers, it makes a strange sound also. It maybe not your issue, but its just an idea..

[gr5 2,173]

But peggy if he loses steps it will be cumulative - it won't suddenly repair itself on a higher layer, right?

Azurial please post gcode file that you used to make the print (not the STL file). I want to look at it please. Post it on some site like dropbox or youmagine or other hosting site and link to it from a new post in this topic.

[Azurial 0]

I'd like to post the GCode, but i print via usb strait to the printer from Cura - im not sure how to post from there. as for the fan i've just done a pretty extreme mod, cutting a hole in the desk its on (no one sits at it) and mounting a 12v fan into it, replacing the small fan i had attached to the RAMPs board. it's set to pull air in and shoot it right at the electronics.

Edited April 5, 2016 by Guest

[avogra 68]

 

that is really extreme, but also very cool Did it help?

[gr5 2,173]

Well I think something about your STL file is messing up cura and giving you a bad gcode file so without seeing that I can't test my theory.

From cura there should be a "folder" icon near the top left. Right click on that and choose "save to disk" or "save gcode" or something like that. Do this for every print. Always. Otherwise a year later when someone asks you what settings you used (e.g. "did you use cool head lift?") you can check. Or if you print something that comes out perfect one time and not the next you can check these settings.

Anyway the next time you have this weird "shifting" post the gcode file and a photo of the print so I can test my theory. I don't think you have a hardware problem because if it is a hardware problem it somehow self corrects after a few layers. That would imply something is very loose - gets in the wrong position and then 20 minutes later it shifts back where it was (e.g. loose extruder).

[Azurial 0]

Well the mod keeps the electronics just under room temp (which in my basement is about 55-60f) by about 2-4 degrees at any given time.

As for the prints it didnt seem to help. I have yet to change any settings so the gcode should remain unchanged shouldnt it?

[Azurial 0]

I think I've found the problem. Even with the fan in place there's a faint squeak coming from the primer. They've been greased and everything. Then I looked at the rods thinking they could be bent. Looking at the hot end rods there are grooves from the bearings digging into them. I've got some replacement rods on order so I'll report back when I get them installed.

[gr5 2,173]

Get replacement bearings also. I think they are only a few dollars each.

[Azurial 0]

New rods finally came in the other day and I installed them with new bearings (thanks for the suggestion!) Still have an issue with layers shifting and definitely not the slicer becasue I have done 3 test cubes since installing the new rods and they have all come out differently. I'm currently taking a look at my belts again just to make sure that nothing's loose there. This machine has maybe 60hours of total print time on it so far so I don't think that any of other components are screwed up yet. I think there may be an issue on the print head as it does seem to give a little when i push it enough so I'll start there. I've also got some extra parts on order to see if the entire assembly is to blame. at this point I could probably build an entirely new printer with the spares ive got laying around lol.

[gr5 2,173]

I have done 3 test cubes since installing the new rods and they have all come out differently.

Photos please. Also do you have an opaque color? Preferably not black or white but somewhere in between? Gray shows up details the best (or almost any color that isn't close to black or white).

[Azurial 0]

 

I have done 3 test cubes since installing the new rods and they have all come out differently.

Photos please.  Also do you have an opaque color?  Preferably not black or white but somewhere in between?  Gray shows up details the best (or almost any color that isn't close to black or white).

 

at the moment. I only have black, but I will get an image as soon as this next calibration cube is done. I ended up having to tighten a belt. Turns out it wasn't holding tension because one of the block clamps snapped in two. I've repaired it and doing a calibration test as I type. I'm guessing that one of the cats wanted to lay down on the build plate (checked and repaired) they got up off of it a bit fast and kocked it just hard enough to put the right amount of pressure on the rod to fracture the clamp. Stranger things have happened, so I'm looking into meshing off either side of the printer and the top to try to prevent that just a bit.

[Azurial 0]

[gr5 2,173]

I watched the first 1:12 and I now think it's slipping in the X axis. This is very common but usually it only slips in one direction but you have it slipping left *and* right. Not impossible.

So it's almost always caused by one of the pulleys not tight enough. There are 6 (SIX!) pulleys on the X axis (not 4). The most likely ones to be slipping are the 2 on the short belt. You should be able to tighten them hopefully without loosening the motor screws if you have the right too. Maybe. I forget. Anyway you have to really tighten the hell out of those set screws. Enough such that the tool twists a bit. Enough that it hurts your fingers if you have an L shaped wrench.

UMO often comes with a second set of set screws that are silver instead of black - the silver ones are pointier and supposedly better. I didn't realize this and I still have my black set screws and they are fine but boy do you have to tighten the hell out of those.

[gr5 2,173]

It's also possible that the X motor is skipping steps. This is much less likely but it can happen. Usually only if X axis has much higher friction than Y. That can happen if the end caps are too tight or if it needs a drop of oil on the long rods. Feel the friction for X versus Y. Are they about the same? If so then it's probably fine. Still consider loosening one screw on one one endcap for each of the 2 rods that spin for the X axis movement.

[Azurial 0]

It's also possible that the X motor is skipping steps.  This is much less likely but it can happen.  Usually only if X axis has much higher friction than Y.  That can happen if the end caps are too tight or if it needs a drop of oil on the long rods.  Feel the friction for X versus Y.  Are they about the same?  If so then it's probably fine.  Still consider loosening one screw on one one endcap for each of the 2 rods that spin for the X axis movement.

 

I've actually torn down the cross rods again. I also initially thought that there was slipping or motor skipping but I checked the motors movement by marking with a sharpie on the pulley. Belts, pulleys all came out fine, but the issue was after I replaced everything is that one of the replacement bearings I used was shot. It was missing an entire row of balls inside one of the tracks.

Thank you guys again for your assistance and helping me nail down the issue. If it werent for trying ALL of your suggestions I couldn't have found the actual problem.