Major bug, x-axis leaning to the right (positive)

[Binni1969]

Hi,

I ran into massive trouble.
Some printed parts are leaning to the right on my leapfrog creatr, it seems that every layer is minimal shifted in x-axis +direction!
So my parts have a ~88° angle in z-axis not a rectangular.
The glass bed is absolutely leveled and all the screws of the pulleys are thight.
After a few tests I have to say the creatr is innocent!
I even tried generating the gcode with simplify3D and sending it to the creatr with repetier-host: parts leaning to the right!
Generating gcode with cura and sending with simplify3D: working perfect, so it seems there is a major problem in simplify3D! Generating with cura or slicer will lead to no problems.
It happens only at more complex parts or when I print more than one object on the creatr.
Is this a known bug?

Best regards,
Wolfgang

[JoeJ]

There's a pretty easy way to test this. Open the gcode file in the S3D previewer. Does the part look perfectly vertical there or is it also slanted by a few degrees? If the part looks fine in the previewer, that means the gcode is correct. In that case, this is likely due to the stepper motors skipping steps and causing misalignment issues. That's a pretty common problem. Sometimes it's due to incorrect belt tension, worm screws that are too loose, or even trying to run the printer at too high of speeds.

But the best way to choose where to start looking is to see if the gcode file is correct to begin with.

[andre]

That is most likely not a software bug, but a hardware issue. I had the same thing occur on my prints, later found out my endstop cable was getting interference from the motor wiring. I noticed this after when initiating a home, the X or Y axis would home a bit earlier than normal without hitting the switch. I rerouted the cables so that they were not near each other. No issues afterwards.

At the time I suspected S3D, but after printing the same part with Makerware and noticing the same angle, I knew this was hardware. There were times where I went back and forth S3D to Makerware, and MW was printing them upright. By shear chance, MW printed the part angled, and that was the tip off.

[Binni1969]

Hi,

preview looks fine. But if it was a hardware problem then why do I get perfect prints in Cura and Slic3r with the same stl-file? And the results are repeatable.
When printing easier objects like calibration cubes, etc. everything works fine.
In another forum a user told about the same problems with Simplify3D.

[JoeJ]

And I've had this same problem before with many other applications. I think what andre and I are trying to say is that the problem is that the motors are skipping steps, not that the software is creating slanted gcode. Like andre said, there are many possible mechanical issues that could cause this. It's also possible that you are trying to print at too high of a speed in the S3D software which only makes things worse.

[Festus440]

I've had this happen on my Creatr too. Not just in the X axis, but in both the X and Y axis at the same time.
However it's very intermittent. I don't think it's S3D. I also don't think it's due to speed causing it to miss steps (Creatr has two steppers on the Y axis).

I've posted an old photo of an example. It was a fairly simple part and I was only printing one.

Wolfgang, if you have a file that does this consistently you should post it, or maybe even a factory file of it.

[Binni1969]

Hi,

thanks for the Support! I tried to print slow, but this had no effect. With Cura and Slicer I printed this stl (rips of an airplane wing) even faster than my fastest Trial on simplify3D without any negative effect.
I will upload the gcode when I am back home, not sure if I saved the factory file.

[CatDog]

If you could post your gcode we could take a look at it and see if there is something in the code. I'm very curious about this since you say that you sliced in Cura and print in S3D works fine, slice in S3D and run in Cura host results in a slanted part. That is a good test (along with your other permutations on this) and points to something in the generated gcode or an interpretation (including accelerations, speeds, etc) of it. Recently, a friend had an odd issue with another slicer that was inserting code that caused his hot end to careen off in the -Y direction after printing 37 layers. It was very consistent and only occurred on this one part using this slicer. I figured out the issue and it had to do with his source geometry (STL file, Scott). An odd situation with triangles (6 "almost" equilateral triangles were causing the slicer to act irrationally since it interpreted them as planar when in fact they were not - a rounding error.

~CatDog>

[andre]

If its a simple part, send us the S3D gcode. We can convert it over to x3g and try to print it on our printers, see what happens.

[CatDog]

No need to spend time printing it, let's look at it first. You never know what you might find tucked away in there.

Another Q&D test is to run it through a gcode analyzer. I like http://gcode.ws

~CatDog>