Print layers drifting towards rear

[DDLLC]

I have a Mars that is relatively new to me. As each layer prints, it is offset toward the rear of the machine on each layer. I believe this is related to the drive screw being tilted relative to the base. I've loosened the two small hex-head bolts and attempted to realign screw but it seems to shift back once re-tightened. The first picture shows two parts printed: the one closest to the granite block is after "realignment" while the other was printed prior to thinking this was a problem. They are essentially the same.

I have pictures of a right angle along the drive screw which is a bit exaggerated since the right angle needs to be pressed down at the lower corner since it tilts into the recess. The second one uses the right angle with the ruler to bridge the vat recess. It will rock if held against the drive screw versus the base. I've used my height gage to estimate how much the difference is about 6.5-7.0 inches out from the screw. The display is in inches so if the right angle is held against the screw it raises about 0.009 inches out at the 6.5-7.0 inch distance.

It is difficult to measure the total amount of displacement but it appears to shift 0.9 mm in 138 layers or 0.0065 mm/layer.  This is a very small amount so I'm not sure how successfully I can align the system to remove it. If somebody has a method they've used, I'd be interested in learning it.

Additionally there is a thickness difference of about 0.5 mm from the thin edge to the thick edge on these samples. In the past I printed a similar block slightly shorter but the same width and it didn't seem to show the same amount of drift or thickness difference. The stl file was from a different program and the FEP film was replaced. I reset the zero before every print, so I don't think it is a misalignment between the print platten and base.

Is there a known issue in Chitubox that can introduce such a step in each print layer?

I've run an exposure test and the defaults for the Elegoo seem fine for their resin (grey).

I plan to make some structures that are taller to get some more accurate measurements but won't be able to print them for a couple of days. I should mention that the test ROOKs seemed to print fine. Unfortunately the base of the Rook was slightly bowed after curing so it rocks on the granite flat and I can't really measure it to see if there is a vertical drift in it.

[cliffyk]

The X- and Y-axis position of the platen assembly's travel is restricted by the trolley's travel within the vertical guide tower rails in complete indifference to the lead screw's orientation if the trolley's guide wheels are properly adjusted. The lead screw only affects vertical position. If indeed the platform is shifting rearward as it rises look to the trolley's mounting in the tower rails and/or alignment of the tower itself.

i would be more inclined (no pun intended) to suspect the platen's alignment with the LCD/FEP--the so-called "leveling" adjustment.

What resin/exposure settings are you using? THis is a long shot, but might be relevant--IDK...

[DDLLC]

The X- and Y-axis position of the platen assembly's travel is restricted by the trolley's travel within the vertical guide tower rails in complete indifference to the lead screw's orientation if the trolley's guide wheels are properly adjusted. The lead screw only affects vertical position. If indeed the platform is shifting rearward as it rises look to the trolley's mounting in the tower rails and/or alignment of the tower itself.

i would be more inclined (no pun intended) to suspect the platen's alignment with the LCD/FEP--the so-called "leveling" adjustment.

What resin/exposure settings are you using? THis is a long shot, but might be relevant--IDK...

The elegoo resin exposures settings are 60 seconds for the first 5 layers and off the top of my head I think the other layers were 8 seconds. I did print an exposure test and it looked rather okay. The overall thickness of the test piece is small so hard to measure the various features to see if there is any vertical shifting. I think I see a slant on some ID's but can't measure it.

I do align the platen before every print using a piece of paper (the same one for now) on the LCD screen as in the directions. I haven't tried to zero it to the FEP itself as I think I've seen some people do, but its hard to imagine that would make a difference. Since I did change the FEP once, I will inspect the film holder is fully seated into the VAT part. If one side is slightly proud of the VAT frame, that could be a culprit and will be less time consuming to correct than a tear down.

You mention the trolley guide wheels should be properly adjusted. Is there something specific I should be looking for when I disassemble the rail? It seems to me that a wheel misaligned might cause more of a wobble instead of a constant shift.

Did you have to realign the vertical guide tower rails at some point and if so, about how much shimming was required?

Thank you for your inputs

Greg

[cliffyk]

I just posted in another thread that I found Elegoo's suggestion of using a "sheet of paper" to be rather nebulous as "sheet of paper" is not a standard unit of measure. A sheet of plain ol' everyday 20 lb. bond copy/printer paper is 0.097 mm (97 microns) thick, however the OEM FEP is 0.15 mm (150 microns). I have been using a sheet of 30 lb vellum I found to be just 0.15 mm thick. The difference, 53 microns, is a bit more than one default setting "layer".

Set in this manner, in printing it would seem to me that the platen would fall to the mechanical "zero" limit--which would in reality be quite firmly up against and in fact compressing the FEP, then rise what it believed to be 50 microns above the FEP for the first layer--actually just 3 microns above the FEP for a 3 micron first layer.

There is a YouTube video re: adjusting the trolley (mistakenly titled as "Z-axis rail calibration")¹ that explains disassembly/assembly well, though i disagree with the producer's opinion that the trolley should fall under it's own weight as that would provide no pre-load to the assembly. I would state that "it should move freely without binding, but not fall under it's own weight". All precision moving assemblies require some preload to accommodate any distortions introduced under normal dynamic loads.

The vertical beam on my plain ol' Mars (serial # "M20xxx") is as perfectly perpendicular the the base as any mensuration devices I have can indicate. I have never changed that.

--------------------------------------------------
¹ - It is of course not "calibration" as calibration would require comparison of some characteristic of the Z-axis to that same chatacteristic of some reference standard. OTH Vertical alignment of the Z-Axis beam would likely be "calibration" as it would require use of some reference standard.

[DDLLC]

I just posted in another thread that I found Elegoo's suggestion of using a "sheet of paper" to be rather nebulous as "sheet of paper" is not a standard unit of measure. A sheet of plain ol' everyday 20 lb. bond copy/printer paper is 0.097 mm (97 microns) thick, however the OEM FEP is 0.15 mm (150 microns). I have been using a sheet of 30 lb vellum I found to be just 0.15 mm thick. The difference, 53 microns, is a bit more than one default setting "layer".

Set in this manner, in printing it would seem to me that the platen would fall to the mechanical "zero" limit--which would in reality be quite firmly up against and in fact compressing the FEP, then rise what it believed to be 50 microns above the FEP for the first layer--actually just 3 microns above the FEP for a 3 micron first layer.

There is a YouTube video re: adjusting the trolley (mistakenly titled as "Z-axis rail calibration")¹ that explains disassembly/assembly well, though i disagree with the producer's opinion that the trolley should fall under it's own weight as that would provide no pre-load to the assembly. I would state that "it should move freely without binding, but not fall under it's own weight". All precision moving assemblies require some preload to accommodate any distortions introduced under normal dynamic loads.

The vertical beam on my plain ol' Mars (serial # "M20xxx") is as perfectly perpendicular the the base as any mensuration devices I have can indicate. I have never changed that.

--------------------------------------------------
¹ - It is of course not "calibration" as calibration would require comparison of some characteristic of the Z-axis to that same chatacteristic of some reference standard. OTH Vertical alignment of the Z-Axis beam would likely be "calibration" as it would require use of some reference standard.


Cliff,
I think I watched that calibration video yesterday. I agree that there should be a bit of resistance between the wheels and the column, but would also say that before he adjusted it, the trolley did seem to have excessive stiffness. Stepper motors typically don't respond well to excessive force and anything above the amount to remove back lash or slop just leads to excessive wear.

The paper comment is well taken. The first few zeroings I did I just used the "manual's" cover sheet. Probably the most value one can get from the manual! Subsequent ones were printing paper (20#) which is decidedly thinner. I'll source something beefer. I'm not convince either issue is the cause of the problem I've seen but it never hurts to eliminate possibilities

Greg

[cliffyk]

The first few zeroings I did I just used the "manual's" cover sheet. Probably the most value one can get from the manual!
Greg

Well stated!!!

I'm not convince either issue is the cause of the problem I've seen but it never hurts to eliminate possibilities

I agree 110%

[DDLLC]

I did have a chance to test some different papers for zeroing (170 micron and 100 micron) but did not see anything that indicated a difference. All the samples had a slight printing shift. I printed both a simple rectangle (92 x 65 x 5 mm) and then a set of nested hollow rectangles just to save on resin. I did have some tests that showed an extreme tilt in the build plate and the parts on two occasions (see pictures). I zeroed on each print and was careful to tighten the set screws. in the one picture you can see the plate tilt quite clearly. I tried to move the build plate by hand and it was nearly impossible. I'm doubtful that the stepper motor could have exerted as much force as I did (I think it would have just lost steps). I didn't have a chance to test any further until yesterday. The slants were much less but it is still there.

The last picture is of the one of the set screws contrasted with an unused spare. I think the "flattening" of the screw end that tightens against the split cylindrical sleeve is obvious. Is there some mechanical reason why the ends of the set screws are concave?

Hopefully next week I will find time to tear apart the Z post and check the bearing tightness. I'm not clear as how that would cause the problems I've seen, but might as well eliminate it as a possibility. I did check the z post perpendicularity to the base plate. Using a typical hardware store (not a machinist) square, the post seems to be slanted back enough to leave a 0.010 inch gap at a height of 4.75 inches above the base plate. That's a pretty small angle and since I'm seeing deviations far greater than that at a design print height of 5mm (sorry for mixed units), I'm pretty sure that's not the source of my problem.

Greg

[cliffyk]

The concave end of the screw is just the way standard machine screws are produced. For this application I ground mine the a ball end.

For any critical use I cannot highly enough recommend junking the OEM mast and replacing it with the Jackson Products' completely redesigned component.

[DDLLC]

The concave end of the screw is just the way standard machine screws are produced. For this application I ground mine the a ball end.

For any critical use I cannot highly enough recommend junking the OEM mast and replacing it with the Jackson Products' completely redesigned component.

I agree that the Jackson Product's mast looks much better. I've built some large scale CNC routers over the years and I think his mast is well designed. I'll put it on my wish list but first want to find the source of the current problem I'm experiencing.

Filing the set screws is also on my list and hopefully it will compress the sleeve tighter.

[cliffyk]

As we are clamping a sphere, the perimeter of the inverse conical nose of the OEM set screws will want to do one or two, or both, of two bad things; a) they will dig into the sphere and make irreparable divots that will become the new favourite seating/setting positon for the sphere and screws, or b) they will cause the sphere to rotate along the screw nose's path as the screw is tightened screwing up (no pun intended) up any positioning of the platen just accomplished.

I also put a dab of high pressure grease on the screw's new ball nose to minimise any tendency of he screw's nose to rotate the sphere.