Form1+ laser flare issues illustrated - pics and video

@RocusHalbasch - No worries… we’ve all been there.

@Monger_Designs to the contrary I think it is the responsibility of support to admit and acknowledge my printer is not behaving as intended. Which it clearly is not. If I have a hard drive go bad on my computer the support guy can admit my drive is not behaving as intended. This does not require the president of the company. If it turns out there is a major flaw in the line of hard drives it may not be the support guys place to admit that, but it is still wrong for him to not acknowledge that my given drive is not behaving as intended. In this case @Stephen is skirting admitting that my printer is not behaving as intended, and is vaguely suggesting the problem is user error instead. If @Stephen is aware my printer is not behaving as intended and not acknowledging so and instead insinuating the problem is user error that is in my opinion borderline amoral. To a less knowledgeable user this could divert them off the path of getting the problem with their printer fixed which could possibly lead to their warranty expiring without having yet gotten it fixed.

As for them checking every other part of the printer first, that strategy seems ridiculous and foolish. Here is a quicker more direct way. Find prints where there are clear artifacts that seem to be of the type we users are attributing to flare. Find a printer that when printing these parts shows these artifacts. Replace the laser with one that does not have any flare, or develop and install one single iris, since this doesn’t have to be the final solution and is just a proof of concept it does not need to be fancy just functional enough to work for a few prints and remove most of the flare. Now reprint the objects if the artifacts are gone or significantly better it’s the laser. If I had a group of engineers on hand, who worked 40 hours a week for me, and a giant pile of Form1+'s that I didn’t have to worry about voiding the warranty on I’m pretty sure I could have whether or not it’s the laser cleared up in a pretty short period of time.

As a consumer I agree with all you said. Ultimately though, taking responsibility can take many forms. They are not refusing to fix your printer are they?

I’m not sure what to make of the following statement:

“I’d like to help you however we can; this said, I’m a little unclear on what assistance that I can offer. Would you like me to orient your test print to improve the surface quality?”

It is not a refusal to fix my printer (though I have had them pretty much refuse to fix one of my previous ones in the past), it’s just a statement suggesting that all he can do for me is rotate my part, and that if I don’t want that he doesn’t know why we are talking. Like I said completely noncommittal, and evasive. They aren’t refusing, they are just not even talking about it as a possibility, and seem to be asking for me to go away.

There is probably a good chance that the particular lasers used happen to have these flaws? If so what to do they do, replace it with another, then another and so on? If I could find the specs on it I am sure there are other manufacturers out there.
How difficult is it to swap the laser module out? It must sit inside some sort of frame work and I don’t thing a small self contained module like that would need collimating because if that was possible then there probably wouldn’t be a flair.

Curious if inside the barrel of the module if it were just a hair wider would there even be a flair. Seems like as Rocus suggested it gets worse as intensity is increased which means it might just be off just slightly.

I’m pretty sure most if not all of the new lasers have some degree of flare. I also don’t think that is such a big deal. I also think they all probably have a perfectly good 300um spot in the middle of all that flare and halo. I’m pretty sure the current lasers are adequate. I think the problem is that there is no iris to block all the extra stray light they are throwing. So the idea there is simple a little ways in front of the laser make a wall with a very small pinhole. Line the pinhole up so the 300um bright point is roughly in the middle. End result. No flare.

As for how easy it is to switch out the lasers, the answer is very easy. Loosen a screw remove the old one, slide in the new one, then tighten the screw back. The only problem is then you need to recalibrate the printer.

2 Likes

Bad news - it turns out that the F1+ beam is not collimated (btw I suspect the original F1 laser was) - the beam is actually focused to a spot only at a certain distance from the laser - at approx 30cm - ie the distance from the laser outlet to the PDMS surface.

That is - the beam profile changes shape as distance from the laser increases - it starts large, shrinks down (but grows flare) and then starts increasing again.

The beam profile as it exits the laser is actually quite large - almost 1mm square (it’s literally square) - with no flare that can be trimmed away with an aperture,

It’s not until the you look at the beam profile about 15cm from the laser that the flare begins to take form.

If you somehow managed to set up the laser (move it back from galvo mirrors) so that the beam travelled enough distance for an aperture to effectively trim the flare - then the dot would be out of focus on the PDMS surface, since the lens inside the laser has a focal length tuned to the distance from the laser to the PDMS surface - beyond that distance the dot starts getting bigger…

I’m still looking at whether it might be possible to home-brew a laser spot conditioning system using an extra-mirror+aperture+extra-lens combo, simply because I’ve put so much time into this already - and if I manage it (feeling much less confident about it now though) that might help to focus @Formlabs attention on this - since they have yet to respond in any concrete way.

3 Likes

@KevinHolmes if that is true how did @JoshK do it with just an iris?

I have no idea - although it’s clear that flare profiles are different, so it’s possible he could have fluked something - however having looked at the laser profile and how it changes over distance - and looking at the lens (which appears to be a rectangular planar convex) visible inside the laser - I have to say I think the straight iris/aperture route is a non-starter

Nope, no fluke. When I removed the choke to send the printer back I did another laser-spot-test and it was back to previous. You are right Kevin that it changes with distance because it looks fairly tight close to the laser, then larger further out. But the flare does exist up close, you just need to split hairs to build the perfect “spacial filter” as I guess they are called.

Then the design problem they (formlabs) are probably having in designing a retrofit is probably one of alignment?

@RocusHalbasch I printed your helical filled cylinder thing. I only have clear resin version 2. My form1+ did mess it up but not quite in the way you showed. I will post photos later and add dimensions.

What I meant by fluke was not the picture of your laser spot test - or that your choke (iris/aperture) was not doing anything, but that what you had achieved was a fluke, ie, you got lucky, even though it was presumably still a lot of work…

As I said, although flares do follow a pattern, there is considerable variation; some people don’t have the “carrot” opposite the “rabbit ears” for instance - so perhaps the particular qualities of your laser flare worked with your design and manual positioning.

I think it’s relatively easy to trim the spot with an aperture to remove the flare, but the difficulty is in not removing power from the actual spot at the same time - since all the light is passing through the same small window close the laser.

I also think it’s next to impossible to judge “split hairs” simply by eye-balling the resulting spot shining on the wall, I can only think you got lucky - both with your laser flare profile, and in fitting your choke. I cant see it being a repeatable procedure.

Also - how much print testing did you do?

1 Like

Yep, during my second attempt I actually removed too much light and the resin didn’t cure enough. So I enlarged the hole. I did lots of prints with the choke, too many to remember. And they were all successful. Before installing it all my prints failed. They were large prints. I should add, even though the prints were all successes, I still had to filter the resin after each print. The success came from not getting any curing too far from the model to allow it to grow. But the hazy nature of the spot was still screwing with viscosity of the resin.

Here is the design concept, I based it on “laminar water jet” nozzle designs. It seemed like a clever and obvious choice to prevent new reflections from causing issues.

5 Likes

This is the device they threatened your warranty over as well. I don’t see the harm it could have done. If formlabs had allowed you to use it for the sake of testing we might have been able to better demonstrate what was and what was not the result of laser flare, with before and after shots.

@RocusHalbasch,
I think the final chapter of this issue is that the Form1+ has a design flaw in that they did not condition the laser spot, like in a laser printer. I wasn’t aware laser printers had that thing in there until I read it here, but if it is feasible on millions of laser printers FormLabs can do it too. I think support is being rude to you because your printer is performing as designed, even though the design is slightly flawed. But since your printer can’t print the beautiful models shown on the FormLabs advertisements, I feel/felt like you. I think FormLabs should re-print and re-photograph all their advertising models with YOUR printer since it is supposedly normal quality.

You have the same choices I did, live with a less-than advertised printer, or ditch it for a decent price while you still have the opportunity. Once the Form2 comes out and people start saying how much better it prints, yours will have little value.

2 Likes

This is where I disagree. It is not printing as designed. The design involved a laser which had a flare too weak to negatively effect prints. This is what formlabs claimed. However the printer I got has a flare that is strong enough to negatively effect prints. That is a quality issue or a miscalculation, not the design. Had they designed it knowing that it would cause these problems and said we think this is acceptable, then I would understand more.

Holy crap, i never thought the beam might not be collimated.

That’s crazy! Having an uncollimated beam in a machine where the laser path length changes depending on the angle/distance from center is insane. That would pretty much limit printing within spec (i.e. 300 um feature size) to the center of the build area. Everything away from center would have progressively worse detail.

Is there any chance it might just not be collimated well?

@Ante_Vukorepa I wanted to show a video of how the profile changes over a short distance - but my phone camera is just not up to the job.

I might try and dig out my old sony compact camera - maybe I’ll be able to get some lower exposure shots with that - although I’ve got a long print running at the moment.

But here’s some basic math for you - approx 1mm sized rectangular spot at laser outlet - converging to a 0.3mm sized dot at PDMS surface - and then diverging to a much larger dot (not measured - but maybe approaching 1cm?) at a meter and beyond.

So I’d guess it’s somewhere between what you mean when you say “not collimated well” - and really divergent. In that it’s good enough to maintain a laser spot size that only varies a bit from 0.3mm over the surface of the PDMS.

Basically I think the lens they have in the laser - has a focal length of somewhere around 30cm - so the loss of focus over the build surface would not be huge.

However - if you pull the laser back from its housing far enough to catch the flare with an aperture - you’re definitely going to have a defocused laser spot

It would be nice to be able to ask formlabs for details on the laser, however I wouldn’t know where to ask them, and I doubt they would answer.