Unwanted thin film appearing with V1176

GVDP · 2024-04-11 13:45:45 UTC

This one took some days to trace the origin…

Since last week we started getting a very thin film that extended from the base layers towards the edges of the buildplate on our Anycubic Mono X.
trying to find where the sudden cause could be , i did some online research on the possible causes of such a strange phenomena.
possible causes :
-frozen liquid resin damage
-too high UV intensity
-not level of the buildplate
-over exposure+bleeding
-defective USB
-software bug
-failing LCD screen

I’ve tackled all individual possible causes one by one, and in the end i was left with the software bug as last remaining.(apart from replacing the LCD screen)

-replaced resin completly : no difference
-reduced UV intensity: slighlty better, but still a thin film
-releveled buildplate : no difference
-reduced exposure times : failed print, so obviously i’m at the edge there
-swapped usb stick : still thin film
-exported slice file from v1145 instead of v1176, and behold : NO FILM !!!

I did a comparative between exports from both v1145/v1176, and both files have the same size…

I got the idea of a “software bug” because since last week, we did start using v1176 in production, about the same time the issues started. Just to big of a coincidence there? We’re running 2 licenses of which only 1 was upgraded to v1176.

To find the possible cause I eventually did use v1145 on the other system to print a few samples and found no issues…

To put it further to the test, I reprinted on the other pc with v1176 and got the dreaded thin film again…

Must say I’m baffled on what can cause such a behavior. I hope you have an idea to what the cause can be…

We can still continue to use v1145, so that we do not have to do vat-cleaning after each print

A wild idea, but it almost seems that, for some reason , the black is not 100% black anymore and lets a small amount of UV light slip through, resulting in a thin film.

I’ll upload a picture of that thin film later…

Anything you need to start digging into this?

formware · 2024-04-12 08:52:01 UTC

Hi Gvdp,

Are you able to send me 2 files (over email), 1 with and 1 without the problem?

I will inspect them for you.

Furthermore I’ve investigated a parameter that I renamed in my code (the value didn’t change from my code); but the parameter was named ‘machinebackground’ and it seems to have some very strange randomness in it when i export the same file over and over from anycubic workshop software.
You can see my question (and the strangeness) here:

Please understand that the code generating the file formats for these low end machines is public knowledge, largely thanks to tools like UVTools. We write our own code; but based on this public knowledge.
For these low end machine; you don’t get any help from manufacturers. They rather don’t have anybody control their printers than their own software

The other thing I did change on feb8 is add 2 parameters, that were 0 before in the data.
MaxAntiAliasLevel = 16
Properties=1

when I remember correctly this should fix a bug that 2 stage movement is not activated. But if you don’t use this it won’t be activated anyway. Do you use this?

kind regards
Elco

GVDP · 2024-04-12 09:31:08 UTC

Elco,
thnx for the repsonse.
I did upload a picture in my old post, if you want to see what i’m talking about.

However, I noticed some1 cleaned up our print folder and removed the files… sigh…
I will have to recreate the files. I can re-export a sample file, but then I’m not certain those files show the problem during print…

As for the 2stage movement, atm I did not use it, but I have been looking at it…

I do understand the application field for such an old, low end 3Dprinter type is small and does not warrant an indept study. But I’m puzzled by the resulting behavior of the update… it used to work fine before…

atm i’m in friday morning rush mode and cant spend too much time.
i’ll look into it further this afternoon, if possible…

formware · 2024-04-12 09:52:08 UTC

Yeah sure no rush.

Well. It’s basically ‘we don’t know what happens on the inside of the machine’ and we have ‘no one to ask’.

As far as I know; there have been no other reports of this behaviour; and this machine is used quite a lot.

What puzzles me is why that 1 parameter with the name ‘machinebackground’ changes so randomly. I have no idea what it does or why it changes seemingly randomly. But just to be clear; it’s value didn’t change in our software update. (just me naming it differently in my code only)
There is also no documentation on this file format… so no one to ask ;S

The parameters that did change were the max antialias setting and the propertiesCount, but they are the same coming from the anycubic software.
So that would be another question; does a similar print file from anycubic case the same problem?

Have you used UVTools before?
It allows you to inspect the parameters in the binary headers of the files very easily.
It can be a good way for you to check them quickly; copy them over to 2 notepad++ files and just visually compare what is different. (that’s how I usually handle this)

Elco