Welcome to the SeeMeCNC Community

Alright guys, I'm new here and I'm running into an odd issue that I can't seem to find a resolution for. Let me know what you guys think. First let me say that overall my prints are beautiful for only having had this puppy running about a week...nice layers, nice dimensions, the whole nine. So, what's my problem? The first several layers of my print are leaning, then when the print gets to a certain point, the rest of the part prints out beautifully. It really has me puzzled as I can't in any way imagine this being a mechanical issue..my belts are tensioned nicely, everything is square, my bed is good and level, etc. I also keep coming back to ruling out interference as an issue as the problem is only for so many layers and this is not effecting the entire print. Anyone have any thoughts/advice? I'll post some pictures as an example. Also, I'll post a picture of my top layers...they're not so pretty. I thought I'd take this chance to see if anyone can offer advice on how to tweak things for better looking top layers.

Here's an example of the odd layer problem I'm having with the first several layers:

[img]https://dl.dropboxusercontent.com/u/564 ... 9%20PM.jpg[/img]

And here is an example of one of my top layers:

[img]https://dl.dropboxusercontent.com/u/564 ... 1%20PM.jpg[/img]

Any help would be greatly appreciated, guys! Thanks!

Graham

For the top layer try overriding the default extrusion width. For example, I have a 0.4mm nozzle with 0.44mm default width. For top layers I use 0.34mm width and every little nook and cranny gets filled oh-so nice. Apply similar logic if you have a larger nozzle and experiment.

For the lower layer issue, I'm assuming you're gcode doesn't look like that? Hmmm...hopefully some more experienced users will have an idea.

I had a lean on my machine when I first got it. I went through everything and I ended up solving it by upping the current on my steppers from 185 to 205.

Pretty maddening though... Went through towers squaring,belts tensioning, cheapsk ate adjustment, z height, electronic noise, pulley tightening...

Try printing some calibration prints, and see if you can isolate it a bit more

Alright Cope, as per request, here is a few shots of a calibration cube. Ignoring the fact that I haven't really played with bridging yet, the only real problem I see is this goofy lean in the first several layers. I went ahead and included a picture I managed to get on the first layer of the print after it was down. In the picture of the first layer you can already see the filament not quite touching one set of perimeters and starting to go over the opposite edge perimeters which is exactly what is continuing for the first several layers. You can also see that the layer is noticeably thinner in one corner. At first I thought this (thinner in one corner) might just be a leveling error or something of the sort, but I've re-leveled and have had the same results. Anyhow, any thoughts?

Here is the print job:

[img]https://dl.dropboxusercontent.com/u/564 ... 6%20AM.jpg[/img]

Note: it does not lean back the other direction later in the print, I just didn't hold the camera level for the picture.

Here is the picture of the first layer:

[img]https://dl.dropboxusercontent.com/u/564 ... 4%20AM.jpg[/img]

Any help is greatly appreciated!

Graham

Disclaimer: I really have no idea

But here's my *guess* anyway -- is something limiting one of the carriages from being able to get all the way down (e.g. is there a missing tooth or too in a belt when the carriage is almost to the "bottom"? Such that one of the axis isn't actually going to where it needs to?

One thing to try would be to print the cube somewhere other than center. Does it lean differently?

Yes, mine was very similar, though my lean went a bit farther up the print...

What are your print settings and what slicer are you using?

Alright guys, I did go ahead and print another print, closer to the axis I would have suspected of missing teeth (making the lower teeth not get used) and had identical results. I also took this opportunity to check out my belts a bit closer, but am not missing any teeth that I could find. Cope, I'm running pretty much straight settings as recommended from the manual by Gene, the only thing that is different is that my top layers are now a bit tighter per suggestions by Edward.

Cope, I followed through your thread about "the lean" pretty closely before posting here as to not have duplicate data, but figured I had a different problem since the lean I'm encountering seems to travel only about a fourth of the way up my cube and yours seemed to travel through nearly the whole print. (perhaps I was wrong) This problem does seem to be carrying through every stl I slice (stl's I know many have successfully sliced and printed). I'm currently running whatever slic3r comes with the most recent repetier host (9.0 I believe). Do you think I should consider playing with my motor current to see if it has the potential of solving the issue?

It can't hurt to bump it up a little to see if it changes. I went to 205 at first. That fixed it but seemed higher than necessary based on other people in the forum, so I experimented in intervals of 5 until I landed on 190. 185 gave me a slight lean, 190 was perfect.

At first, I did have the lean going all the way through the print, but after trying all the various things being suggested, I did get to the point where it went through the first 30-50 layers, and then printed straight.

Have you tried printing slower/faster to see if that changes anything? If I printed really slowly (5-10mm/s), the lean wasn't there except for the first 4-5 layers.

I'm fairly certain lean was been caused by one or two of my delta arms having slightly more resistance than the others. I think that bumping the current up overcame that and fixed the issue. Of course, ensuring that all the towers and arms are identical is the better fix, but it's pretty tough with the stock arms. I sanded, re-sanded - checked and double checked everything, and without a way to measure anything, it seemed like the cheapskates, towers, and arms were all as they should have been, but I was still getting the lean.

Upgrading to the TL carbon arms eliminated 1 variable for me. That may be something to consider as well.

This may be a dumb question, but where do I alter the current? Is that in the firmware I assume? If so, what lines am I looking for?

Nope, not a dumb question. In config.h you're looking for the line

#define MOTOR_CURRENT {190,190,190,190,0} // Values 0-255 (RAMBO 135 = ~0.75A, 185 = ~1A)

The first 3 values are your tower steppers, the 3rd is your extruder. The 0 would be if you had a second extruder. You don't need to increase your extruder current - I just like things to be consistent.

cope413 wrote:The first 3 values are your tower steppers, the 3rd is your extruder.

You mean the second third is the extruder right?... I'm sorry I couldn't help it, I'm do similar things all the time

Haha, yep. I said what I meant and meant what I said....

Anyway, in case that wasn't clear for grabredemeyer...

The first 3 values are your tower steppers, the 4th value is your extruder, and the 5th (which should be a 0) would be for a second extruder.

No worries guys, I figured it out . Sadly, upping my current didn't make any difference for me. The newest firmware from the GitHub actually has 195 as the values for all the steppers right off the bat by the way. I upped them to 200, with no results. I'm afraid to go too much higher for fear of overheating my RAMBo. Any other thoughts, gentlemen?

One way you could determine if it's added friction on one of your sets of arms would be to take all the arms off at the cheapskates and rotate the effector 120 or 240 degrees, and then print something. If the lean is still there but in a different direction, you would know it's the arms. If it's still there in the same direction, then you can likely eliminate the arms from the equation.

I know it's maddening. Took me about a week to get it figured out...

Another thing to check that may not have been covered, but are you sure your build plate is firmly in place and not moving at all? If it were sliding ever so slightly it could do that. And it could make sense too because the closer (steeper angle) to the glass the hot end is, the more lateral force it is exerting on the plate.

I'm reaching, but there are just so many things that it could be...

It's funny that the calibration cube leans until the layer shape changes, and then it's alright. Can you show us some other parts that lean that much?

Also, you can put a little fan to blow over your RAMBo if you're concerned about overheating. I got this squirrel cage fan because it has Molex connectors, so I can just plug it in to the rat's nest coming out of the power supply. Squirrel cage fans are pretty quiet and deliver a good air flow, so it's no trouble to have it running all the time.

If you prefer a normal (box shaped) fan, there are mounting holes for 30 and 40mm fans drilled into the RAMBo mounting plate surrounding the big hole in the middle. Someone told me the holes will take an M2 screw.

Problem solved, guys! I feel a little dumb for missing this for so long...the axis I had suspicions of being the culprit indeed was. Instead of teeth being missing on the belt, I found that my gear that mounts right on the motor shaft on that axis was slipping ever so slightly. A quick tighten with a hex fixed the problem right up...back to beautiful prints...now back to tweaking settings in pursuit of the perfect print. haha. Thank you guys for all the help and thoughts it took to get to this point!

Graham

The effort you spent figuring this out will pay off, because now you know that much more about your printer. Good job.