Solved Major shift in print along the Y axis on large prints.

Hello,

So I have an unresolved issue that's been causing me problems with larger prints. It seems like once a print reaches a certain size, it has a problem of shifting on the y axis. I typically haven't been around to see when the shift happens, and when I've done some test prints to try and reproduce it, I haven't got anything. (I'll explain a little better below on exactly what is happening, preceded by the conditions it seems most likely to happen under. Sorry, I don't have any pictures on hand, so I'll try and explain the problem as best as I can)

I've done long 8+ hour prints, but in those cases the prints tend to have a smaller footprint, though they are tall enough to rack the hours up. The prints that seem to have the major layer shifts are ones that take up a good amount of space on the print bed. Though, I've had success too, but it's hit or miss.

So, the layer shift might happen 1-3 times. It seems somewhat inconsistent when it happens, but it's likely to happen at least once early on in the print. We're not talking a small layer shift either. Something like a half an inch up to around an inch. It only seems to happen on the Y axis. The print will look fine before and after the layer shift. It's not a series of shifts. It's a single shift, and the model might print another 50-100 layers before another shift happens again, if it happens again. It's usually on the lower layers where it's most likely to happen. Somewhere within the first 10-20 layers.

Now, I know you're gonna be talking about the Y axis belts, pully, whatever. I did try and search up on the forums previously, but didn't find anything that matched my problem specifically. I checked the pullys, the belt tension, and all that. I ran it back and forth, and when I did test that made use of the full range of the Y axis, there were no issues. The belt seems fine with no noticeable wear that'd cause concern. If I pluck the belt it gives a twang noise, or however you want to describe it. Pretty much it sounds like how people on here commonly describe it, when it comes to the sound the belt should make when properly adjusted and plucked like a guitar string.

Now, I also made sure the area around the printer was free of obstructions, cleared out any loose plastic that was sitting around the bed from previous prints, and made sure the cable that connects to the bed wasn't being obstructed by anything, or that there was anything that it could catch on.

Have I described it well enough? I know you'd like a picture, but since I can't just easily reproduce this, I can't just spit out problem models for you guys to see. There is an old faulty model I gave to a friend, that I've asked him if he could bring it to me, but I'm not sure if/when that'll happen.

So, I've been pretty stumped by this. I've got one idea to go off of though, but I haven't been able to test it. I did notice the last time it happened (This was the middle of the night, and I was tired, so forgive me for not expertly documenting all of this), that it said something about the z-endstop being hit. I noticed it as I was turning the machine off, so I didn't have time to process it (with my sleepy brain) until after my fingers clicked the switch on the back.

What happens if that does happen mid print? Like maybe there was something on the print, and the print head bumped up over it, and caused the z-endstop to trigger? Would it cause an offset like that?? Hmm

I wont be able to get to my 3d printer for a few, so I wanted to reach out and see if I could get a conversation going, so when I do get back to my printer I'll have some ideas on what to look for.

My Robo 3D is the R1+ model, the most awesomeness model.

 

I've carefully checked for both of those issues, and it doesn't seem to be either one. Like I said though, I never saw it when it happened, so I never caught what caused it, and since it's inconsistent as to when it happens, it's not easy to reproduce. It's not practical to babysit it for 5+ hours in the hope that it'll happen.

Besides the Y axis, I was also thinking about the X axis, and the head movement, but it's along the bed movement axis and not the head movement, so while I could see the head being caught and causing a shift, it shouldn't cause a shift along the bed axis. It also doesn't really explain why it happens with larger prints, but not smaller ones. As the movement on the X axis isn't always that much different from other prints, but it's when I make use of the full extent of the bed on the Y axis, that it might cause shifts, but as I've said, I've made test prints specifically that traveled the full distance of the Y axis, and they printed fine. I've been careful to keep the area clear and there's no signs of things snagging.

I'm not saying it can't be one of those things, but I've carefully gone through the most obvious things, the things everyone says first.The thing is though, on issues like those, they tend to show up in all prints. At least from what I've seen on this forum, from other people having shifting issues.

On the z-axis. I don't know that it makes sense for the shift to happen if the z-endstop was hit, it's just I noticed that with the last print I did, where this issue happened. I have the g-code for that print still, so I can print it again and see what happens. I'd like to set a camera up to film it, but I'm not sure how to best handle that, or handle a 4-5 hour recording.

By the way, Hi Mark. I see your name around here a lot; thanks for the work you do in helping people.

 

Shoot, this sucks. I don't want to be afraid of full volume multi-day prints. I want my Robo 3d to keep being the star it was born to be.

 

Ideas are good.

The g-code may not be at issue, but it's the only file I have for a part that caused this particular problem, so my hope would be that the issue might happen again. I'm not exactly sure if it would though! and it's a 5-6 hour print job, so I may be on edge for a bit, ha.

Having problems caused during acceleration crossed my mind. I looked into making adjustments in that regard, but it's hard to say. It's difficult for any of this since any changes I make, I can't really test them, since it hasn't been overly predictable in a testable manner.

Have you ever had the head catch on the part in some manner? Or have a warping part bump the head enough to trigger the z-endstop during a print?

I wonder if it's my cooling setup. Maybe it's cooling the tip of the nozzle too much, causing the nozzle not to be hot enough to push through any plastic that's sticking up higher than the nozzle, so it catches? That seems a stretch.

Maybe someone hacked my computer and subtly changes my g-code just to mess with me??? hmm, I'll file that one away for when I'm much crazier.

 

I more or less took my printer apart last night, checked everything, and reassembled. So far it works, though it appears that it's laying filament down a little closer on one side than the other, on the first layer. Like one side looks awesome, but the other side could have been put down closer. Left to right. - So, I'm sure there's an easy adjustment to be made to fix that. My first print went well, since surgery, so we'll see after the next few days if it gives problems with huge prints. I fixed a few other issues along the way that could have been contributing to my original problem that I posted about.

Ahh, too much cooling, seems silly for PLA. I started getting temp drop errors though when the fans would kick in at full speed. I'm not sure, since I expanded my fan ducts to one on either side, instead of just one. I don't recall that being a problem right away? But now, after a few layers, when the fans crank up, the hot end temperature drops, slowly but surely. I've adjusted it to tone back the power some, and it's been working fine. Do you think it could be a defective heater element? The fans are 40x40x20mm fans (in addition to the heatsink fan on the hotend).

These 3d printers are way cool, but as I work on this I'm reminded how far the industry has to go if it wants to make a bulletproof product that anyone can pick up. I'm not sure any problem I've had can't be fixed, or wasn't a huge deal to fix, but it certainly takes someone that tinkers a lot to troubleshoot them day to day, and even then, sometimes you still need help. Thanks again Mark, for working with me as I try and get this Robo3d back to it's A game.

 

I went ahead and ordered some. This will be nice. I can replace the heater cartridges on all my 3d printers.

 

Hey, here's one of the failed prints.

 

It was orientated along the Y axis, front to back. The shift happened along the Y axis. It's supposed to be, more or less, a glorified box. The only part that needed support was a overhang that was at either end of the box, on the inside. The outside is just a box with holes in it.

 

I've added an image for added clarification. Pardon my lame draw over, but it should be enough. I'm nowhere near my printer, so I can't snap a real picture.

I'll check to make sure the capstans are aligned.

 

I'm with you that it's got to be mechanical. Well, like I said, I went and re hauled it last night and fixed a bunch of minor issues along the way. I'm going to double check the capstans in particular, as you suggested, as that's not something I ever thought to check. Plan is to get a long haul print in tonight and see how it performs.

 

Another option would be to go into your slicer settings and simply slow down the non-printing move speeds, and change the "randomize start points" option to "Optimize start points" if you're using Simplify3D (under the Layer Settings tab).

My not-scientific tests on my original Robo printer was that the shifting would happen whenever the huge glass bed went on a rapid movement between layers that was long enough generate the momentum needed to overpower the stepper motors when it was time to stop.