Jump to content

Recommended Posts

I've given in and bought a Photon Mono since they are on offer, as well as a wash/cure station. Looking forward to giving this a go, been designing stuff for printing long enough, only right that I get my hands dirty with the business end too.

 

I just need to clear some space first...

 

623860004_IMG_20210205_1843461.jpg.ece53e546bbe863aaea913c7c55b6422.jpg

Edited by Quarryscapes
  • Like 3
Link to post
Share on other sites
2 hours ago, Quarryscapes said:

I've given in and bought a Photon Mono since they are on offer, as well as a wash/cure station. Looking forward to giving this a go, been designing stuff for printing long enough, only right that I get my hands dirty with the business end too.

 

I just need to clear some space first...

 

623860004_IMG_20210205_1843461.jpg.ece53e546bbe863aaea913c7c55b6422.jpg

A Unimat 3 - I used to gaze longingly  at the one in the window of Fantastic Toy and Hobby Shop in Sydney when we'd go into the city during school holidays back when I was about 12... Ended up getting a second hand SL with a heap of attachments after saving my pocket money for about 3 years.

 

I've still got it, seems like its from another world beside the CNC router, the laser cutter and the 3D printers, but I think I got more joy out of using it, and it never ever 'just stopped' working unlike the others!

 

  • Like 1
Link to post
Share on other sites

I mentioned earlier having to resort to using the supplied metal scraper to get a flat-on-the-bed moulding off and the possible damage to the build plate it caused. Following recommendations on the various support groups I have purchased some scrapers fitted with plastic razor blades like this one:

IMG_3694.JPG.d93e5ffe6a1ca144ecf2abe5c39c4a20.JPG

 

They work brilliantly and do not cause any damage, so a good investment.

 

There have been reports of Anti Aliasing not working on the Mono X. In spite of updating the firmware to 3.4.7 as suggested I could not find any evidence that it was working. As it was something I used with my OG Photon it was proving frustrating.

 

First off I needed a simple test piece that I could try the various settings on:

208603558_testpiece.jpg.33ab580699d0a7b02ed090c50e83d3ee.jpg

 

The test piece (attached) is a simple hollow sphere 3cm in diameter that is printed flat on the build plate. There is a vent hole to avoid any suction issues and act as a handy visual reference as to the sphere's orientation on the build plate. Being spherical the impact of any stepping of the print can be examined and the impact at different positions assessed allowing for a good educated guess at the best orientation of any future component.

 

Various prints were done at 0.040 layer thickness ranging from no anti-aliasing enabled to 2, 4 and 8. The last test was with blur set to 8:

IMG_3693.JPG.d3227366408fd29623256518627e05e4.JPG

 

There is no physical difference in the quality of the prints, so either anti-aliasing is not working in firmware 3.4.7 or not working in Chitubox 1.8.0 and 1.8.1.

 

I am currently trying a print using lychee and I guess I should also do one with the Photon Workshop.

 

The only physical difference, unsurprisingly, was printing the test piece with a lower layer height, in this case 0.020 (0.040 on left) :

IMG_3697.JPG.125c34c254c7a1df02b51f3a30880a14.JPG

 

Anti_Alias.stl

  • Like 1
Link to post
Share on other sites

I have now finished printing samples using both Lychee and Photon Workshop all with anti-aliasing settings. As with the earlier examples there are no dicernable differences which leads me to conclude that firmware version 3.4.7 does NOT restore the missing anti-aliasing functionality despite assurances to the contrary.

  • Informative/Useful 3
Link to post
Share on other sites

Well I couldn't be bothered to make space on the bench, and after taking a whiff of the resin I opted to set up camp in the bathroom where there is an extraction fan, and wipe-able surfaces should the worst happen.

 

Whilst I was out doing battle with IT and my Volvo's version thereof, I let it print the test cube thing. (After levelling the plate of course). It printed OK so now I just need to get in some more consumables so I can get down to some serious experimentation. 

Link to post
Share on other sites

First Railway models attempted. These are a couple of versions of parts to backdate the MM1 models 2 Plank to Cambrian Railways condition. The separate springs and axleboxes were printed on the build platform, the back was chamfered to mitigate the elephants foot effect but it still popped up on the underside of the springs. 

 

Layer height was 0.035mm, and the complete sets were angled at 55.something degrees to optimize for the front face. I'd drafted the spring collar a few degrees which has caused a little un-necessary stepping, which I will remove for the next attempt.

 

I'm very happy with both versions, and was quite lucky - Prior to these I had printed the Exposure reference test, and part of the 0.8 second one had stuck to the FEP without me realising. Luckily these were placed to miss the lump of cured resin, which wasn't found until after I'd printed these and emptied the VAT. 

BFI-XT2B6860.jpg

BFI-XT2B6863.jpg

BFI-XT2B6873.jpg

BFI-XT2B6881.jpg

BFI-XT2B6886.jpg

BFI-XT2B6890.jpg

BFI-XT2B6893.jpg

  • Like 12
  • Craftsmanship/clever 6
Link to post
Share on other sites
On 06/02/2021 at 08:58, MikeTrice said:

I have now finished printing samples using both Lychee and Photon Workshop all with anti-aliasing settings. As with the earlier examples there are no dicernable differences which leads me to conclude that firmware version 3.4.7 does NOT restore the missing anti-aliasing functionality despite assurances to the contrary.

 

Interesting Mike, the bloke in the video below went wild over it. I'm about to set off a print, so I'll give your stl a go on my printer and see how I get on. I'll be using the latest version of Lychee Pro, and I'm on 3.4.7. That said, I've done the supports of the other models on the plate in Chitubox to see how they might work for me. I'll do the full set of settings when I have a print.

 

 

Edited by JCL
  • Informative/Useful 1
Link to post
Share on other sites

He is the only person I have heard of that claims it works, many others find it does not. Irrespective of Anti Aliasing I believe there was a print quality improvement when 3.4.7 was released which Eilte Geek is mistaking for AA working. What he does not do is compare a non AA print under 3.4.7 with an AA enabled print under 3.4.7.

  • Informative/Useful 3
Link to post
Share on other sites

That's true, he doesn't. I've got one printing at the moment (last night's failed), so I'll let you know what V2 AA set to 4 looks like on my machine.

 

Link to post
Share on other sites

balls.jpg.551afd3a65137ae752071357fba5bf9f.jpgHi Mike

 

It looks like our balls are very similar (as it were). This was printed using:

  • layer height 0.05mm
  • Antialias V2, level 4
  • Sliced in Lychee Pro 3.2.1

I took the photo in a raking light, but in reality, at least to me, the lines are generally barely visible.

 

Depending on the weather tomorrow, I'll see if the layer lines show up under a coat of primer.

Edited by JCL
Link to post
Share on other sites

Hi all, I've a Chitubox question.

 

I normally use Lychee, but I thought I'd give Chitubox a go as it has been getting some good comments.

 

I have two STL models (diesel and a building), and I made sure both were good by running them through 3D Builder.

In Chitubox, I placed the models on the plate and added supports

Then I exported the result to an STL file.

When I opened this file in 3D Builder it needed repairing again. Is there something I'm doing wrong, or is this a problem with Chitubox?

Edited by JCL
Link to post
Share on other sites
5 hours ago, JCL said:

Hi Mike

 

It looks like our balls are very similar (as it were).

 

I would agree. The printing is very fine and I think you will find with no AA the same result. As you say hard to see with the naked eye. What is interesting, and useful with the test piece, is seeing how the stepping changes around the circumference which might help determine printing angles in relation to "Z"

  • Like 1
  • Agree 1
Link to post
Share on other sites

Hi Mike,

 

Just popped over from the Mono X FB group, and a poster on there has said he gets the same lines we do when he uses Lychee, but not on Photon Workshop. I'll reslice in Workshop and I'll pop over to the frigid shed and reprint the ball.

  • Thanks 1
  • Informative/Useful 1
Link to post
Share on other sites

Here's the ball sliced in Photon Workshop. 

 

1973224668_photonball.jpg.b7b30e80b4b56c967186c40f529d3dbf.jpg

 

Interestingly, I don't see the those other circles on the side of the ball on the Photon Workshop print. Other than that they look very similar. In my Photon Workshop, I don't get the Antialias V2 option, so I just set it to 4. In real life, I don't see  the lines except under particular lighting conditions, and the ball just looks a little matt.

 

And to remind you, here's the one done in Lychee.

 

balls.jpg.b6c12de1eb481ca2286de88462b4192d.jpg

 

  • Informative/Useful 1
Link to post
Share on other sites

Well I was all ready to do some prints on the Photon this morning but suddenly had an epiphany moment. I was expecting far too much of anti-aliasing.

 

I will try to explain but it will largely be words, any attempt at photography is not proving helpful.

 

So lets start with how anti-aliasing works. Without AA a voxel, which is a rectangular prism defined in size as being 1 pixel of the screen * the layer height, is the smallest printable unit. When printing a sphere, as in my test piece, the angle of the sides varies around the sphere with the equator being as near as possible to 90 degrees to the horizontal and the poles being near 0 degrees. When printing, as that angle moves from 90 degrees to 0, the stepping becomes more pronounced.

 

Enter anti-aliasing. When enabled the slicer software generates images that enable printing of halftones (sort of). When a halftone is encountered by the printer it exposes that voxel for a shorter period of time which resulting a in bubble forming adjacent to an already printed voxel effectively pushing the voxel out in the horizontal direction. Vertical height is controlled by the distance between the build plate and the last layer of the print so AA can only impact horizontally. The AA setting determines how many halftones are created and applied so given an AA setting of 8 means the bubble can expand from 0 to 7/8ths of a voxel.

 

Still with me?

 

So basically anti-aliasing can only adjust the width of a voxel to a limited degree. So what does ths mean to our sphere? Well the top layering that is very evident is more of less unaffected by AA as a fraction of a voxel difference has minimal visible effect when the step is several voxels in width. It is only when we look nearer to the sphere's equator can some form of difference be seen and only then under magnification. So it appears AA is working but I was setting my expectations way too high. As @jcl observed some of the artifacts on the sides of the sphere printed smoother which fits in with my new found theory. Thinking this through further AA will be most noticable with the angle being between roughly (it really depends on layer height) 45 and 90 degrees to the horizontal which is why when I was comparing the stepping on the top of the sphere I could not see any difference.

 

Sphere on left has been printed at 0.04mm layer height with no anti-aliasing the one on the right with an anti-aliasing value of 4 using chitubox 1.8.1.

IMG_3725.JPG.4461cde1cfa094220707d5fcec36729f.JPG

 

Depending on where you get visible stepping it might be that the only option is to print at a lower layer height with the consequence that print times will increase. Alternatively find a position on the test sphere with minimum layering and angle your print to correspond.

Edited by MikeTrice
  • Like 3
  • Agree 1
  • Informative/Useful 1
  • Interesting/Thought-provoking 1
Link to post
Share on other sites

Absolutely. And I’ve just noticed, the column holding up the ball on the right is also a lot smoother as you’d expect

 

There is always going to be a trade off between speed and quality, and where that point is, depends on personal tastes and circumstances. It was an interesting exercise to do, but given the size of the pixels, at some point you get to a hard barrier.

 

Although science and engineering has created these machines, using them is definitely an art.

  • Agree 1
Link to post
Share on other sites

One more thing, we’ve also shown that the artifacts also depend on the slicer. My two ball prints had the same (available) print settings, but achieved slightly different results.

Edited by JCL
Link to post
Share on other sites

In lithographic printing it was found that 1,250 lines per inch showed stepping on curves, but 2,500 didn't show any stepping to the naked eye. Under a scope you could still see the steps though. So in theory:rolleyes: a layer height of 0.01mm should be invisible to the naked eye, of course that doesn't always work. It was strange how a tiny speck of white (or clear on the film) would show up but not a speck of black on white. We used to use neg' film so that's why I know, it wasn't until you checked the finished plate that it showed up:angry:.

 

I would imagine that the stepping on 3d prints will show more on a 'flatter' curve ie the long edge of an oval rather than the short edge, as it will be a greater proportional jump in height. 

  • Informative/Useful 1
Link to post
Share on other sites
  • 2 weeks later...

Sorry to hijack this thread again, I posted over on a Mars thread but haven't got a response. My current problem is that although the build plate is properly set to zero, when I start to print, the build plate descends, begins the print process, but then on subsequent layers does not descend into the vat but stops short and then rises again. The result is of course that the print only forms on the FEP. Any ideas? I guess it's a setting somewhere that has got inadvertently changed? Many thanks.

Link to post
Share on other sites
11 hours ago, Tricky said:

Sorry to hijack this thread again, I posted over on a Mars thread but haven't got a response. My current problem is that although the build plate is properly set to zero, when I start to print, the build plate descends, begins the print process, but then on subsequent layers does not descend into the vat but stops short and then rises again. The result is of course that the print only forms on the FEP. Any ideas? I guess it's a setting somewhere that has got inadvertently changed? Many thanks.

A few observations if I may?

 

I assume the printer has worked correctly previously. If so what has changed (I am thinking something like a new version of the slicer, updated firmware, replaced FEP etc)? What slicer are you using? Do you have something that has printed correctly in the past, such as the tower test piece, that you can print without reslicing? Is there a later version of the firmware available or later release for the slicer program?

Link to post
Share on other sites

That’s certainly an odd problem. No settings that would explicitly cause that. To add to Mike’s excellent suggestions it could also be a mechanical problem - failing stepper motor on lead screw or something. Id not expect that to be quite as repetitive as you say though. 

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.