Tronxy x5s Scewed layer/layer shifting
December 12, 2018 08:04AM
So i got my X5S up and running with the MKS SBASE 1.3 instead of the standard electronic. Beside running quiter because of the microstepping and 32bit board it also allowed me to run the dual axis with the extruder driver #2 (instead of using a shared stepper driver on the old electronic). This allows me to home the two axis individually, so i do not have to worry about misallignment of z1 and z2.

But my prints comes out with layer shifting or something that looks like it (see pictures). I do not think that it is skipping, because it seems to find back to the original position. As you can see the problem is repeatable, the shift seems to come around the same layers every time. I have tried to fiddle around with the z acme screws and bed to see if this was the culprit, but it seems rock solid. I have tried to tension the Core XY belts, but the problem seems to persist. Any idea what is causing this and how to remedy it?

It should be noted that the layers, when measured individually are within 0.1mm of the 40mm target.
Attachments:
open | download - Skewed Layer 1.png (698.2 KB)
open | download - Skewed Layer 2.png (274.9 KB)
Re: Tronxy x5s Scewed layer/layer shifting
December 12, 2018 09:58AM
Is the Z axis tilting back and forth as it moves? That would squish layers out on one side and make then thicker (but narrower) on the other side. The Z axis generally looks rough in the prints...


Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]
Re: Tronxy x5s Scewed layer/layer shifting
December 14, 2018 02:27PM
Okay so i figured out what was wrong after a lot of print and tweaking.

I tried to print two identical test samples at the same time, but placing them as far apart as possible. They came out perfectly identical, meaning that the bed could not be wobling (it also feels rock solid). The primary suspect turned to the stepper driver. As you cannot tune the current for the MKS SBASE i had to turn down acceleration- and jerkfactor in Marlin. This did help the quality of the print, there was no layers that were shifting rapidly to one side anymore, but the print did lean to one side, showing me that i was still losing steps. I am kind of new to 3d printers, but have experience with programming and electronics and i am a mechanical engineer by profession. If this was not a coreXY setup i would immediately have seen that one axis was drifting, but because the drifting was so minor and pretty consistent (in equaling itself out.... almost) it was hard to identify... be aware newcomers to corexy.

I have already dialed the accel and jerk down so much that i think the only solution is to buy a board with more powerful stepper driver.... i think i am going for the Duet. Or manage to hook up my external Leadshine DM422 stepperdriver (have tried but can't figure out what i am doing wrong). I will let you know what i did and if it helped.
Re: Tronxy x5s Scewed layer/layer shifting
December 14, 2018 03:00PM
On MKS SBASE 1.3 you can adjust the 8825 current at least with Smoothieware or RRF (just played with RRF these last couple of days) and in fact very easily in the config files ! If you are running Marlin on it, I am surprised it can't ????
With RRF I had in addition to play with pulse timings for the 8825 drivers as I was getting Pisa tower effect due to missed steps. So could be that too.

Edited 1 time(s). Last edit at 12/14/2018 03:03PM by MKSA.


"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)
Re: Tronxy x5s Scewed layer/layer shifting
December 14, 2018 10:36PM
There has to be a way to adjust the driver current.

The DRV8825s are known to skip usteps at low speeds. They are a terrible choice for a 3D printer drivers. They can be fixed by adding diodes across the motor windings. See: [cabristor.blogspot.com]


Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]
Re: Tronxy x5s Scewed layer/layer shifting
December 15, 2018 02:22AM
Quote
MKSA
With RRF I had in addition to play with pulse timings for the 8825 drivers as I was getting Pisa tower effect due to missed steps. So could be that too.

The DRV8825 need very slow step pulse timings, AFAIR it's around 2us, compared to 400ns for A4988 drivers and even lower for Trinamic drivers. Using any 32-bit firmware, you will probably need to slow down the step pulse generation when using them, to avoid missed steps.



Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].
Re: Tronxy x5s Scewed layer/layer shifting
December 15, 2018 10:50AM
Hi guys

thank you so much for the help. That are some good points for fixing it. I should have researched further on the MKS SBASE before diving into it.

There should be a way to tune them through Marlin, but i can't seem to find the parameters and i did try to google but could not find an answer. Any guys here know where to find them in the configuration.h file?
Re: Tronxy x5s Scewed layer/layer shifting
December 15, 2018 11:21AM
Okay after digging the internet for a couple of hours i found out that this is apparently a bug in the Marlin software and the cure is supposedly to set the pulse width to 4-5. I will try this an report back. More reading here, but suffice to say that this is a Marlin 2 / MKS Sbase issue and nothing with the Tronxy printer.

[github.com]
Re: Tronxy x5s Scewed layer/layer shifting
December 17, 2018 09:08AM
Quote
Palmhoej
Okay after digging the internet for a couple of hours i found out that this is apparently a bug in the Marlin software and the cure is supposedly to set the pulse width to 4-5. I will try this an report back. More reading here, but suffice to say that this is a Marlin 2 / MKS Sbase issue and nothing with the Tronxy printer.

[github.com]

Looks like you are confusing "Marlin bugs" with parameters that you don't set up properly.
Marlin developers offer you the possibility to set pulse timings, current, accel, jerk ... to match the drivers/ steppers specifications.
The 8825 in addition have their own issues as written above.

Yes there are bugs and missed steps can also be due to bugs ! To cover so many different processors, boards, displays, com, machine kinematics is not small feat. Besides the guys do that fro free !
Then how about asking in the Marlin fora ?

Edited 2 time(s). Last edit at 12/17/2018 09:52AM by MKSA.


"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)
Re: Tronxy x5s Scewed layer/layer shifting
December 17, 2018 02:51PM
Not square X and Y may cause additional tension on Y-axis, which will lead to vertical skewing along Y. Print calibration cube and measure diagonals in the XY plane. If it is a case you need to play with each belt tension to square XY.
Re: Tronxy x5s Scewed layer/layer shifting
December 17, 2018 05:13PM
MKSA i am not blaming anyone for this bug and i agree that it is amazing what this open software has accomplished. It was not to smite anyone, but if you see the link i posted, it is reported and accepted as a bug in Marlin regarding 32bit controllers if i read it correct. And yes the picky stepper drivers does not further help this issue, but i would still call it a bug in using Marlin with MKS SBASE as i wrote. Sorry if it was formulated in an ungrateful way. Just want to make sure that others having the same issue read this, so they do not make the mistake as i did and begin to fingle with the machine, when the bug is with the software/electronic smiling smiley
Re: Tronxy x5s Scewed layer/layer shifting
December 18, 2018 06:56PM
I have opened a bug report on Marlin 2.0 Github:

[github.com]
Sorry, only registered users may post in this forum.

Click here to login