Welcome! Log In Create A New Profile

Advanced

Layer shift MeCreator2

Posted by fumo 
Layer shift MeCreator2
March 06, 2019 01:52PM
Hi all,
since some weeks I experience layer shifts in y-direction shift when printing objects higher than approx. 10mm on my MeCreator2. After some googling I found that probably the A4988 overheat and go into thermal shutdown. So I replaced them (x and y) with DRV8825, set Vref to .8V and left the enclosure bottom open to ensure appropriate cooling. Unfortunately, the problem persists. Any helpful suggestions are welcome.
Attachments:
open | download - IMG_20190306_193603__01_1551897788111.jpg (560.3 KB)
open | download - IMG_20190306_193500__01_1551897690634.jpg (512.4 KB)
Re: Layer shift MeCreator2
March 06, 2019 02:29PM
I would say that it is possible the Y-Axis is binding, could be in need of lubrication.
Have heard that some use grease though in my opinion grease is to thick. I use a sewing machine style oil, garage door silicone spray or a good Gun oil like Ballistol works well...


Electronics Technician / Computer Programmer
Re: Layer shift MeCreator2
March 06, 2019 03:09PM
It looks like you have some over extrusion issues too.
Re: Layer shift MeCreator2
March 06, 2019 03:59PM
Thanks for the suggestions. I'll try tomorrow and report smiling smiley
Re: Layer shift MeCreator2
March 07, 2019 01:37PM
Ok then, I slightly reduced the extrusion multiplier, cleaned and lubricated the x and y rails, and they run much smoother now. But....

after cleaning and lubricating I watched a print of a small cylinder (12mm diameter, 15mm high). At about 11 mm height I noticed an audible "clack", and the y-axis was offset by some 12mm confused smiley. So I'm still after the reason for that error!!!
Re: Layer shift MeCreator2
March 07, 2019 03:17PM
Without being able to see your printer all anyone can do is make some best guesses. Is it possible that your nozzel hit the print? If your putting down too much plastic it sticks up and the nozzel hits it.That could account for the noise. Check the z travel to make sure its moving up the correct heighth for every layer. Enabling Z-hop sometimes helps but its not a fix. Sorry if I'm telling things you already know.
Re: Layer shift MeCreator2
March 07, 2019 03:27PM
Quote
fumo
Ok then, I slightly reduced the extrusion multiplier, cleaned and lubricated the x and y rails, and they run much smoother now. But....

after cleaning and lubricating I watched a print of a small cylinder (12mm diameter, 15mm high). At about 11 mm height I noticed an audible "clack", and the y-axis was offset by some 12mm confused smiley. So I'm still after the reason for that error!!!

That isn't an error, that is an event!
You know exactly what your problem is, you just don't know what caused it. That clack means that you had a mechanical collision of some sort. On my printers it is usually a fan shroud or nozzle that impacts a piece of plastic that is sticking up. The piece of plastic is usually something that curled up and hardened at an elevation higher than the rest of the print. If that plastic is large then when the nozzle hits it only one of two things can happen: the nozzle melts through it, or the nozzle smacks it and dislodges the print or stalls the end effector or carriage. It sounds like the latter occurred in your case. This makes me think of these things:
Your part fan isn't doing its job and the plastic is not cooling fast enough.
Your Y axis stepper current isn't high enough and your stepper is therefore too weak to handle the collision.
Your Y axis belt has something lodged in it that is stalling the stepper (this is a lower probability).
Your stepper is jamming or the pulley on the stepper has a loose set screw.
The belt pulley on the other end of the gantry jammed and caused a belt skip.

having fun yet?


Kits: He3D K200 Kossel, Folgertech Kossel 2020 upgraded E3Dv6, Anet A8 upgraded E3Dv6, Tevo Tarantula enhanced parts and dual-head
Scratch: Large bed Cartesian, Linear slide Delta, Maker-Beam XL Micro Delta
Re: Layer shift MeCreator2
March 10, 2019 08:18AM
Hi there,
a short update:

The problem is still there. I fully watched a print. Everything went fine until a height of about 22mm, then the slight "clack" and I had my layer shift in y-directon again. It always happens exactly (!) in y-direction, so I ruled out a nozzle collision, because that would happen in variuos directions, depending on the location, height and momentary print direction. I could prove that there is positively no filament that builds up, the nozzle could move freely all the time, no collision with the printed part happened.

So my next task is to check the belt and pulley.

Hermann
Attachments:
open | download - 20190310_132156.jpg (343 KB)
Re: Layer shift MeCreator2
March 11, 2019 08:10AM
Assuming this is a direct extruder.
Might be worth a good look at the extruder mounting. If the hot end throat is close to the gear maybe when the bed moves in Y the nozle catches, tilts the extruder and the throat hits the gear making it jamb and miss.
Re: Layer shift MeCreator2
March 12, 2019 09:46AM
Hi guys,
I'm afraid I found the reason for my layer shifts angry smiley

A detailed examination showed that both ends of the x gantry the LM8UU-holders have a crack!

Just tell me if you share my opinion that this might have caused the y-layer-shifts. If anyone knows if (and where) I can get the STL's of these components, that information would be greatly appreciated. I could't find them in the internet yet.

Hermann
Attachments:
open | download - MeCreator2_xCarriageLeft.png (207 KB)
open | download - MeCreator2_xCarriageRight.png (213.3 KB)
Re: Layer shift MeCreator2
March 14, 2019 03:59PM
Latest (not so good) news:

I fixed (i.e. glued) the cracks in the x-ends, did a complete alignment of rods, belts, etc. and made sure everything is running as desired. The calibration cube came out fine (it finished before a layer shift occured).


Unfortunately, the layer shifts are still there. So I sat beside the printer watching what happens. This is what I found:

At first, printing is going on as usual. After some time (maybe 3 minutes, maybe 10 minutes) the printer seems to hold on for about a second, then resumes printing with shifted y layers. To rule out any network errors, I printed from SD-card (no network, no USB connection). But the issue is still there.

My last idea (for the moment) is to increase the stepper current.

Comments are welcome!

Hermann
Re: Layer shift MeCreator2
March 18, 2019 03:42PM
Try to eliminate as many factors as you can. Try a dry run to rule out filament catching. Just leave out the filament, or if the slicer supports a dry run, try that.

Lower the maximum velocity, acceleration and jerk settings on the printer by setting it on the display or using gcode.

Monitor the temperature of the stepsticks, monitor the voltage. Look for spikes, dips, ripple etc.


--
Kind regards
Imqqmi

NFAN CoreXY printer:
[reprap.org]
Re: Layer shift MeCreator2
March 18, 2019 04:52PM
Quote
fumo
Latest (not so good) news:

I fixed (i.e. glued) the cracks in the x-ends, did a complete alignment of rods, belts, etc. and made sure everything is running as desired. The calibration cube came out fine (it finished before a layer shift occured).


Unfortunately, the layer shifts are still there. So I sat beside the printer watching what happens. This is what I found:

At first, printing is going on as usual. After some time (maybe 3 minutes, maybe 10 minutes) the printer seems to hold on for about a second, then resumes printing with shifted y layers. To rule out any network errors, I printed from SD-card (no network, no USB connection). But the issue is still there.

My last idea (for the moment) is to increase the stepper current.

Comments are welcome!

Hermann

If the problem always occurs at a particular layer, move your Z to that layer and scan back-and-forth on the Y axis to see if you can get your hit. You do not have an extrusion issue, since the layer shifts over. That is a gantry movement error. This may or may not discover anything.

Having the print pause, and then start up several mm over means that your printer did not get the intervening commands, it stopped getting codes, waited and started getting codes, but the head was now in the wrong place. Because you say that this occurs when connected to USB or when running from the SD card, I am beginning to suspect buffer memory corruption, which would be the main controller. It could also be that your stepper driver wasn't getting instructions, which again means a main controller communications error from the processor to the stepper driver. Because this happens after a certain period of time, that says "thermal" issue to me. Put a fan in the enclosure blowing across the stepper boards/chips. I had freezes on two different printers that were corrected by just putting a 25mm fan near the driver chip "row" blowing over the chips. The freezes went away.

DLC


Kits: He3D K200 Kossel, Folgertech Kossel 2020 upgraded E3Dv6, Anet A8 upgraded E3Dv6, Tevo Tarantula enhanced parts and dual-head
Scratch: Large bed Cartesian, Linear slide Delta, Maker-Beam XL Micro Delta
Re: Layer shift MeCreator2
March 18, 2019 05:17PM
Thanks to all of you who gave good hints. I'll share my latest experiences:

Modifying the stepper current was a good idea, but I got heavy improvement after lowering (!) the stepper current. I set Vref on the x and y-8825 piggyback boards to approx. 0.8V, and the first prints were very promising and made me hope the issue was solved.

It is not! I printed a bigger part, and after two and a half hour or so the printer stopped totally. Examination showed a VERY hot x-stepper and a blown fuse on the GT2560 control board. Since I have no spare fuses (Littlefuse nano, 10A), I ordered some. Now all further action will be delayed until arrival of that fuses ☹.

Probably I have to set a much lower stepper current. Btw, does anyone know the "right" current for the steppers in the MeCreator2?
Re: Layer shift MeCreator2
March 18, 2019 08:30PM
It could be a short from something else too, maybe you need to replace the board or at least check everything that's connected. Use an amp clamp to measure the current on the input, then add all the sensors and other devices back to the board until you see a big jump in current usage.

Was the fuse also powering the heated bed? If not, it should never have happened as stepper motors really don't draw that much current. My corexy with nema23 steppers draw around 1-2A when running At 12V and lower at 24V. I have no idea what steppers are in your machine. Maybe you can look at the sticker of the stepper motor what type it is? Generally you use a current well below what the stepper motor is specified to handle.
I usually create some gcode that runs the axis at top speed back and forth. If no skipping occurs then I call it good. The noise it makes is also a factor. I also run the motors at very slow speed to see when they run the smoothest. Especially with A4988 steppers it makes a big difference. The TMC2660 not as much, it only produces more noise at higher currents (and heat).


--
Kind regards
Imqqmi

NFAN CoreXY printer:
[reprap.org]
Re: Layer shift MeCreator2
March 19, 2019 03:23AM
I totally agree that the fuse never should have blown. It does not power the heated bed, which has a separate fuse, motors are NEMA17, and the whole system is running at 24V.

Lets find out what happened...
Sorry, only registered users may post in this forum.

Click here to login