Bad Layer Shift on Y axis!

I have had my Mini since December, and it’s been relatively trouble-free. Today, however, I experienced some isolated but significant “Y axis layer shifting”.

The model was http://www.thingiverse.com/thing:1340784. I was using 0.30mm layer height, and scaled to 80% so the model was about 125mm tall.

  • Everything printed perfectly up to about 85mm (i.e., up to around layer ~283).

  • Over the course of the next ~4mm of Z (so layers ~290 to ~303), there are several layer shifts evident where the Y position increased abnormally. It wasn’t a single shift; rather it appears there were about 5 or 6 separate +Y shifts that occurred during the course of those 12-14 layers.

  • The rest of the model (from ~89mm to 125mm, so ~120 more layers) printed without any further Y shifts. Of course the Y shifts that occurred between Z=85 and Z=89 mean the upper portion of the model is offset by about +20mm Y.

So what could cause several +Y shifts in this isolated area of a print?

Here’s what I’ve checked/considered so far:

  1. The Y-axis pulley is in the correct position, grub screws are tight, motor screws are tight.
  2. The Y-axis belt looks perfect, no wear or missing teeth. Idler moves smoothly.
  3. The Y-axis belt tension appears normal, same as when delivered and similar to the X-axis belt tension.
  4. The electronics cooling fan is running, vents are clear with good flow out the back vents, room temp is 68F (20C).
  5. The bed moves normally, no binding or odd noises, no problems on any other prints (and no problems on most of this print, before and after the narrow Z range where the Y shifting occurred).
  6. There was no outside interference during the print (no pets, kids, etc. in the house)
  7. The Y-axis stepper does get pretty hot, like ~65C, but from what I’ve read that is pretty normal.
  8. Firmware is stock, as delivered from Lulzbot.
  9. Slicing and print control was via Simplify3d. I saved the Gcode and loaded it into a visualizer, and it looks fine so I am confident that the Y shifts are NOT in the Gcode (i.e., not a slicer problem).

What else could cause sporadic Y axis shifts like this?


Seems like you have checked everything that I would have.

Could it be possible you had a brown out situation midway through the print? Are you hooked up to a back up battery supply? Don’t know if that could cause it to lose its Y Home position and think that it is in a different place.

I don’t think there were any power problems during the print. The printer is not on a UPS, but there are several other UPS units in the house and none of them “screamed” as they do when there is low voltage or momentary outage.

Since the “mechanical” stuff (pulley, belt, bearings, etc) all seem fine, and because the problem was intermittent, my first thought was missed steps from an overheated stepper driver. But the electronics fan is working fine, good air flow and a cool environment, so that doesn’t seem logical either.

Maybe something got in the way on that specific height, a cable for example? I Don’t know the mini, but on the big TAZ I had something simmilar, where the cables of the X axis stepper can catch the heat bed cable connector. I had to relocate a zip tie to avoid that.
Check every cable (I can see at last 3 bundles on your picture) how far they can be moved…

Edit: Do you have choosen a propper scale factor? The Mini has a Print Area of 152mm x 152mm x 158mm.

I was just browsing the 3Dprinting subreddit when I found someone that experience a similar outcome.

https://www.reddit.com/r/3Dprinting/comments/49r1h0/help_thats_not_good_y_axis_slipping/

I’m stumped as well. If the mechanical seem fine and secured and the board is getting ample cooling, the other thing I can think of is something electrical is malfunctioning. Connections loose, bad motor, bad board would be guesses. The NEMA motors operating temperatures seem normal from my understanding as well.

Next would probably be to see if the layer shift is repeatable. Maybe its time dependent (something heating up) or maybe it will always shift at those specific layers.

Maybe check for cracks in the parts? Though I can’t imagine there be one without propagating into a loose belt or a loose motor mount.

Sorry I could not be of more help.

Agree, and that’s my plan. Going to print again today using Cura, and see what happens.
But I am concerned about my sanity, i.e., doing the same thing over and expecting different results. :smiley:

Nothing can get in the way, all cables are secured well away from the bed. As mentioned in original post, the object is ~125mm high so well within the Mini’s print area, and the problem occurred between ~85mm and 89mm. No problems with even larger objects in the past.

If you have an extra heat-sink you could try and put on the the Y Stepper driver. Maybe it is getting hot.

How would I locate that driver? Haven’t looked at the board, are they marked?

Yes, on a Full size RAMBo they are, on the Mini it is the second one in from the right side edge of the board(U6) looking at the component side.

Hmmmm…

I opened the case to take a look. I located the three drivers, labeled U5, U6, U7. I noticed U7 (which I assume is for Z axis?) looked different than the other two. Upon closer inspection, U7 it had adhesive residue on it. And caught up in the wires – a 6mm aluminum heat sink that had fallen off the chip at U7!

So apparently the board came with U7 heat sinked from Ultimaker, or Aleph adds a heat sink to U7. (I’m the original owner, I didn’t put it there, and the printer shipped directly from Aleph Objects).

So I guess I will clean up the old adhesive and install another heat sink, although I’m concerned about it falling off again and perhaps shorting out something on the board. (Lucky that didn’t happen this time.)

And assuming U7 is for Z, this likely has nothing to do with the Y layer offset problem I’m trying to solve, right?

Do you have any pets or children about? Curiosity killed the print?

Les

Nope. (See #6 in my original post. )

Oh, sorry, I tend to skim over text too much and miss details like that.

Earthquake? Act of God? Power spike? Overheat trigger on Stepper driver?

Les

No earthquakes or power anomalies during that time. I don’t know how to rule out the Hand of God, but I suspect any God would have more important things to do than mess with my prints. :slight_smile:

Overheat of stepper driver seems plausible, but… The fan is functioning and it blows directly on the mini-Rambo board. I’ve seen one report of someone losing Z steps on a mini due to stepper driver heat, but that was because his fan was disconnected. Other than that, this doesn’t seem to be a common problem with the Mini.

The search continues.

I had a Mini and I should remember but I don’t know if you can print from an SD card? This looks like a long print and a fun one to make but from a USB port might be kind of iffy. I do think I will print one for my grandson!

So the Mini does not have a SD card slot and when I sliced this with S3D it came out to a 17 hr 46 min print and that is usually an estimate that is off by a lot, my guess it would be nearly 24 hours. That’s full size on a Makergear M2 using PETG. I can do that off my SD card but it would be a long haul on direct USB.
My grandson would love it!!

Confirmed – Aleph Objects does install a small 6.5mm aluminum heat sink on the Z axis stepper driver of the mini-Rambo. It is attached only with a self-adhesive pad. See step #5, here:
https://ohai.lulzbot.com/project/86510011-df83-408c-a220-1dcf8c5a7262/mini/

When I opened the electronics cover of my Mini, I found that aluminum heat sink loose in the enclosure. Seems like that could have caused some real damage if it had touched/shorted components on the board. Luckily, it was caught up on some wires and not touching the PCB. At least that’s where it was when I opened the cover.

The heat sink itself looks 100% identical to these PCCooler brand VRam heat sinks (I had a pack of these in my desk drawer):
http://www.amazon.com/Aluminum-Chips-Cooling-Heatsinks-cooler/dp/B00E36SEB4/ref=sr_1_1?ie=UTF8&qid=1457650788&sr=8-1&keywords=rhs-01+heat+sink

I cleaned off the remaining adhesive from the Z stepper driver and heat sink, grabbed two more heat sinks, and mounted one on each of the X, Y, and Z stepper driver chips using Arctic Silver Ceramic 2-part adhesive. That should hold better (and transfer heat better) than the self-adhesive pads.

While the printer was on the bench, I also added a couple 40mm heat sinks to the Y stepper motor, since it tends to get really hot.

So now there are heat sinks on XYZ stepper drivers, and the Y stepper motor. I hope to try the same print tomorrow with that combination.

Nope, no SD card port on the Mini. I’ve thought of going the RPi/Octoprint route, but I had an extra Win7 laptop so just dedicated it to the printer. Seems to work fine – I have done 24+ hour prints without any USB connection issues. Probably helps that the laptop isn’t doing anything else, just driving the printer.

That must have been small layer height? I used S3D, scaled to 80% (to fit mini dimensions) and a .30mm layer height – I think the print ran just over 8 hours actual time. S3D estimated about 6.5 hours.

No doubt! It’s a really nice design, three tracks so lots of action. No support needed with all the “arches”. It was a great looking machine all the way up to where my printer took a powder. :slight_smile:

I’ll be giving it another try tomorrow.

S3D is set where I get good results, never do Qiuck or Fast mode, its on meduim. Print with PETG 99% of the time. Yes I think this will good fun . I think layer height is .2000 , my nozzle size is .35, filament 1.75.

No, but the model could be hammering on the Y driver chip just enough that it starts missing steps. Plus if you have the heat-sinks it really cannot hurt to have them on there to help cool those things off. :slight_smile: As long as they do not fall off!