Gaps / Calibration Issues?

[matt-brailsford 0]

Hi Guys,

I've just finished putting my UMO+ together and have run a few test prints through and seem to have some issues with gapping / incorrect widths. I've been through the visual trouble shooting guide here which seems to suggest it'll be down to belt tension. Having checked all around, I can't see any real issues. The smaller timing belts are as tight as can be, and the longer ones aren't very tight, but I believe that's fine.

I've attached a couple of pics to highlight the problem. As you can see, there are places where the infill doesn't touch the outline, and on the B shape in the circle you can see that some of the horizontal pieces are much thinner than others so I'm getting inconsistent thicknesses.

Anyone got an ideas on what I can tweak to get these printing right?

Many thanks

Matt

 

[nallath 1,124]

These are usually caused by backlash. Check the tension of your belts & tighten your pulleys.

 

[iefbr14 0]

Check how it renders in cura's layer view.

i had really strange slicing issue with cura lately, namely the 5mm Qbert stairs calibration model was rendered with rightmost stairs 3mm wide instead of 5, and i finally realized it was sliced this way by cura !

as a rule of thumb, always check the layer view before printing : maybe your printer prints just what it has been told to.

 

[gr5 2,265]

You can check the gcode more exactly with this simple web application - just drag and drop your gcode file onto it:

http://gcode.ws/

Backlash can be caused by either too loose or too tight belts. What you say? If belts are too tight then friction is too high and when an axis comes to a stop the head hasn't moved all the way and the belt is tighter at that moment in one direction than the other. If belts are too loose obviously also it might not drag the head all the way.

Often the end caps over the rods cause too much friction. You should be able to push the head easily with your smallest finger of each hand touching the blocks on either side and pushing (with steppers powered off).

But still - I'm guessing you could tighten your long belts a tiny bit more. I recommend clothes pin springs placed on the opposite position of the belt from the blocks.

Or it might not be backlash at all - it might be underextrusion. Test by printing at half the speed - if that fixes the gap then it was underextrusion.

Or it could be a slicing error.

 

[IRobertI 521]

Also check that your rods are squared properly.

Oh, and make sure that the motor pulleys/belts aren't rubbing against the side of the machine causing extra friction. The pulleys shouldn't be a problem on an UMO (as they poke out through the side) but the belt might be rubbing.

 

[matt-brailsford 0]

Thanks guys, I'm just going through your recommendations and trialing them to see if any solve the issue. I've printed and added some belt tensioners which seems to have made it a bit better.

Will try your other suggestions too.

@gr5 I'm wondering if underextreusion is playing a part. On the first layer, it often prints the first line with blobs, which from looking around sounds like under-extrusion so I might try your suggestion next. I had to replace the standard small black gear of the feeder with a 3D printed part as the original snapped during installation so wasn't sure if this could have affected the amount of material being fed through.

 

[matt-brailsford 0]

Ok, it looks like I'm getting closer. I tentioned the long belts and realized I had the fan on backwards so switched that around and things are coming out somewhat better. I do still seem to have a few issues.

If you take a look at these pics, you'll see there is a bit of layer shifting around the eyes / arms, however i don't believe this to be loose pulleys as once it gets past those eye/arm details, it shifts back to where it should be, so it seems when a new detail comes into print, it's struggling to come back in line. Could this be over-tensioned belts? Given that it does this on the front and the side though does that mean all belts could be over tensioned?

Also, on the logo on his chest, there is a bit of stringing, but I'm wondering if this is just to be expected with the nature of the design (ie, it's bridging to mid air so there is nothing to hold it in place). That said, the bridging in his mouth also seems to have drooped a bit which I believe should have been able to print ok.

PS this guy was printed upside down on his head facing the back of the machine should orientation make any difference to your advice.

Cheers guys

Matt

 

[IRobertI 521]

Looks like your images are corrupted. If you want to use the forum to host your pictures (which is preferable so they stay even after you've deleted them from your dropbox) you can see how here.

 

[matt-brailsford 0]

@IRobertl Oops, thanks for the heads up. I've uploaded them to a gallery now and updated my last post (I was struggling to figure out how to upload media so thanks for the link)

 

[gr5 2,265]

1) If the bottom layer seems over or underextruded it's more dependent on your leveling than anything else. By default Cura prints a .3mm bottom layer so you don't have to level quite as perfectly. But I recommend you move the bed up or down a little with the screws so you get perfectly flattened skirt. Do this live while printing. Do only small adjustments at a time - maybe 1/2 turn at the most. 1/4 turn more typical.

2) That's not stringing - that's typical overhang issues.

3) The "shifting" is very exact at the eyes. Look at the pattern - each shift is the same height as each eye. It's caused by backlash which again could be too much friction or overly loose belts. You can tell that the nozzle is moving right to left in this orientation of the photo with the eyes and doesn't return completely to the proper "wall" position before extruding the wall. Did you check the friction of X versus Y? Did you try loosening the end caps?

 

[matt-brailsford 0]

@gr5 thanks for the feedback.

1) I'll have a play with the leveling and see if I can't get it spot on. It was definitely leveled to be "just" touching the nossle, but I'll see if tweaking helps any.

2) I thought so, but thought it worth checking.

3) Yea, it's very exact as to when it happens hence why I doubted it being a slipping pulley. I'm pretty sure it can't be down to slack belts now as they are all pretty tight. The head unit for me has always been pretty stiff to move around, however earlier I did spray some 3-in-1 oil into the bearings behind the end caps which did seem to loosen it up a bit. I've tried oiling the rods themselves which doesn't seem to have made much difference so I'm thinking the only two things it could now be is like you say, too much friction on the end caps (is it worth printing some of these?) or possibly too tight Y bands?

 

[gr5 2,265]

it's trivial (takes seconds) to loosen all the end caps and see if that makes a difference.

Belts too tight can also increase friction. My belts are rather loose. Just tight enough to make a pitch. Anyway, use light oil on the rods. Any light oil will do. Any brand. Preferably without many additives.

 

[chrisr 18]

.. and not WD40

 

[matt-brailsford 0]

By jove, I think we've got it!

In the end it came down to 3 things

1) Too much tension on the X/Y long bands - removed all the tensioners I added

2) Not enough tension on the X/Y short bands - pulled these tighter

3) Too much friction at the end caps - loosened up and added some 3-in-1 oil

Thanks everyone for the help, I'm so much happier with the results now and have a much better understanding of the tensions needed in the bands (hard to portray via text instructions so hard for newcomers to 100% know).

Matt