Is this Z-wobbling? if yes... what to do?

Dear Moai users, here I am again.

Having fixed a few problems by switching to pdms and standard build plate, I went through a few small prints quite happily, and started going bigger.
I printed some sort of lithophane (so… basically, a big flat surface on one face), and got regular horizontal stripes, here is a picture with printing alignment in asura and the results.

I started sanding it before thinking of taking a picture, so please forgive the scratches, but the spacing is about 2.2mm, when printed at 45°. This correspond to about 1.6mm in vertical which -if the leadscrew is standard- is one turn.
I do not use a heater, so that’s not the same problem as what I saw in many similar topics (Temp is 26 to 30°c inside these days, heater seems a bit useless).
Is there a easy fix? is the leadscrew twisted and need replacement (after 1 week and a half, that would be very short lived)? is there something I am missing?

ps: edit to fix the image

Hello @totoreador,
Thank you for the awesome pictures! Before I get to making an analysis I need a little more information from you if you don’t mind.

Are you using the newest firmware?
What type of resin?
software used to generate support:
software used to export gcode:
If using Asura, what is the version and profiles used:
Are you using default Asura support setting?
Support tip diameter size?

Also, can you please post a picture of your setting on the Moai or type them out? That will be helpful as well.

-I am using the latest asura build (2.2.3 after re-release because of merged supports)
-resin is peopoly gray
-supports made in chitubox : standard “medium” settings on automatic (contact diam. 0.55mm), plus a few more light ones (contact diam. 0.3mm) added manually to avoid some “warping”
-slicing/export done in asura (2.2.3, using the unmodified “Moai 130 40µm” profile)

Printer settings are the defaults ones, as in the tutorials (will take the values asap, but it’s printing right now, I don’t think I can display the values without stopping it?)

here they are

NB: didn’t I posted this in hardware? It feels wrong to see this in firmware, doesn’t it?

Probably should be in Hardware haha Ill see if I can move it.

Here is some advice to try out to see if you can get a better print.

  • Dial the laser power down to 57

  • More importantly, your supports are likely too thin at “medium”. Try to use Asura support for now as we have beefed up the supports in there.

  • Also, Try to print it flat on its back and decrease the pm motor speed / initial pm moto speed to 10

I hope this helps and let me know how this goes!

I didn’t placed my print flat because it’s too large for the bed (about 15cm), but anyway… wouldn’t placing things flat be bad? (I am still a complete noob here, but from what I understood reading here and there creating big adhesion surfaces should be avoided ?)

For the rest, I will try asap, to check without wasting too much resin, I made a small piece 10x20x20mm, with 200um walls.

I will compare different supports, from chitubox and asura:
btw, if the supports are indeed beefy now, asura is still “not very efficient” at placing them on automatic:

1 Like

So… I did a few tests and couldn’t really get anything conclusive:

I first printed said models, and got this:

As my troubles were with a print made at an angle, I tried to print the same boxes at 45°, with half of the surface supported to see if it was support related, and got this:

Again, not much to see… So I started a third test, with a new model, with thicker walls and three orientations (flat, vertical, 45°):

With all this, I cannot reproduce this weird “thicker layer” effect (as I had to sand it, I can confirm it was “larger” layers). Only thing that was changed is the laser power (from 59 to 57 as advised), but I fail to see how it could change a “layer” effect.

Any advice about what to test in order to understand what was going on?

At first, the “Z wobble” was starting to make sense, but now with these new photos I am starting to think it is the orientation of the model and the model type. Try to orient it differently on the build plate and add supports along the back on the cubes you made. Try printing a cylinder and see how that turns out. I am convinced it is not a Z wobble just slicing and model selection.

let me know how it turns out. :slight_smile:

The layers getting thicker and thinner may be a power problem. If the power dips, the laser will weaken as well, leading to thinner layers. Some people have reported this happening with the heater module as the heating coil turns off and on… Also some have reported having a faulty power brick.

Before we go further, let’s refocus on what layer lines and many types

  • generally, if it is a mechanical problem like something is not well secured, the layer line is random. not at a consistent pattern

  • layer lines can also be cause by excessive suction force by bad orientation and hollows. To take your example here:

if the bottom of the object is flat, then during printing, there is a vacuum forming inside (like a suction cup) and that will create a lot of pull force. It may be strong enough to cause print object to shift a bit each layer and thus leading to layer lines. The lines are normally random by quite frequent. Solution to this is to create drain holes to release the vaccume

the other potential cause of layer lines is object moving during printing. here you have this object at an angle but the top is not quite well support.
This could also lead to layer lines at the top portion where it can dangle a bit (shift) and thus leading to layer shifts. You can investigate where the shifts start to occur and look at how the support is set up to determine if this is the cause. Chituboix support is optimized for lcd printers and a generally too thin for laser SLA printer so we advised you not to use their default settings. Asura default is strong.

Another that can potential cause by heater but does not happen very frequently is in a cold environment where heat is lost quickly. The Peopoly heater will quickly heats up the chamber (expanding the part) and then the heat is lost quickly (the part contracts), this can lead to layer lines in some can, especially if the print has a large surface area. It has nothing to do with power fluctuation because heater has its own power supplier. Not sure if you have any proof but we love to take a look at the claim. @ShadowDrakken The solution is to actually reduce the output from the heater by blocking part of the out to slow down how fast the heater increase the chamber temperature. One can also cover the top of the Moai to reduce heat loss at very cold environment. This has been reported by some users but we also sold a lot of heaters so the percentage is not high and we believe the reduction in heat output and heat loss can fix that.

Lastly, is that z-axis could shift during due to bend lead screw or unsecured parts. You can check that by
our trusty checkup guide: Moai Checkup guide to make sure key parts of Moai is well secured.

and also some point about

  1. printing flat. Every bottom up resin printers is about minimzing peel force, so generally playing large (like 40% of the surface area) on any layer is not a good idea. It is not just Moai and Moai with peel mechanism can handle solid flat surface print better than most. When you print, avoid too much flat surface area but also don’t be afraid to print one side flat on the plate if that gives the most efficient orientation

  2. the key system setting that can affect your peel force is pm motor speed. I see you have it at 15. set it to 10 will reduce peel force further more. Give it a try and see how that comes out

Hi everyone,
thanks a lot for your feedbacks, I will try a “cylinder” print soon (had a few small but tricky prints to run, which left me with the conclusion that it’s better to have “big supports annoying to clean” rather than “small supports but no print”), and let you know how it turns out.

Just for clarity sake, this was on purpose: the part I printed that showed “whatcouldhavebeenwobbling” was quite well supported I think (see first pictures), so I wanted to see if this was support related, thus part of the print being left unsupported.

Thanks for the figure, that may prove helpful. In the same sort of idea: is there a specific ratio to keep in mind for the surface of a layer that will be printed in regard to the total surface of the supports under (or rather, over) the print?
Concrete example: if I am printing a layer with surface of say 4cm² (or 400mm², as it would happen for a traditional 2cm calibration cube, not tilting it), using default supports (with a tip diameter of 1,2 in asura) I get about 1.1mm² of section per support… how many would be required for the print to actually work (=for the print to stick to those X*1.1mm² surfaces rather than to the FEP or PDMS)?

@peopoly If I have proof? You mean like the thread here on the forum where you acknowledged the problem and directed affected users to contact you for warranty replacement?

So, here is a partial result:

No layering/wobbling effect on the tilted cylinder (though it is somewhat deformed).

Regarding my second question about supports/surface ratio: II got no print out of the vertical one (supports bases on the build plate, a circle on the vat).

on the tilted cylinder, the deformation is due to not having internal supports, you would need it for every circular structure
the vertical cylinder didn’t print due to separation between the support and the model, you might want to check using the z slice view, increase the support tip and manually add supports if needed.
Still, it’s not recommended to orient like that since SLA machines are not good when it comes to printing strait up