A response to Sam_Jacoby

[quote=“VinceErb, post:21, topic:3398”]
dont want to take this thread off the tracks
[/quote] @VinceErb yeah - fair point - would you consider starting a new thread exploring your current printer issues? adding your experience with support after they respond to your new ticket? and perhaps a laser spot test and test piece prints? (as @Steve_Johnstone has noted previously it’s good to get shots of the prints still on the platform unwashed as well as after removal and wash)

Vince, 5 refurbish? How did they get out of QC to you with errors like that?

Nice visual on the laser flair Kevin. I wonder if the new lasers where the lens is glued in if either the lenses have a small spot of glue or the glue melts the lens along one edge (guessing the bottom edge).

interesting question - once my current print is finished I’ll take it out and see if I can get a good view and a pic with my USB microscope.

That’s 5 machines - 3 refurbishments and one upgrade (or new purchase). Two of the replacements were the original F1 which was extremely prone to laser failure, and also galvo failure - so I think only the most recent replacement could have been a DOA QA failure - and who knows, they’re still using that terrible packaging I think - so it could have been shipping damage…

@Sam_Jacoby I understand that there are limitations to the device. The thing is to me right now it looks like something, most likely laser flare, is reducing the area where I can print vertical edges cleanly to a less than 1/3 of the volume of the machine. That is kind of a big restriction to not stated in the spec. At my end we were in the middle of measuring all kind of dimensions on the good and bad prints to see how Z resolution and build location effected things like vertical wall thickness. Almost nothing I have printed worked at resolutions below 0.1mm and what did come out was covered in goo and often off dimension. Until I saw this thread I figured it was just me and something I was doing in orientation. I accept that quality can change with orientation but if it prints at 0.1mm it should also print at 0.025mm. If I am missing something here technically please explain.

As I pointed out @CraigBroady did open it up to suggestions. You acknowledged @KevinHolmes good work in diagnosing the issue with the laser but now casts doubt or at least disregard @RocusHalbasch’s work diagnosing his identical problem. You presumably can see from this thread that all evidence is pointing to laser flare. If I had been in your position I would have marched down to the person who suggested reorienting the part and explained that they were missing the point. What gives? In all our technical talk we got only one post that looked like it was from someone in engineering (@CraigBroady) and it was a polite brush off with no work shown.

I posted a photo of how this is solved in laser printers. I also have a pretty good idea what your base parts cost is. I do not think that you can seriously tell me this is a fix too expensive to institute. At the very least tell the crowd here why it is technically infeasible and over budget. After all the work we the users put into that thread on fixing your product we want, no we deserve, to see some serious analysis posted. Show me the MATH. It is the least you can do. I think we put substantial effort into proving our point, where is formlabs proof of theirs?

I was trying to bring us back to focus on what is causing it. Yes I called out out to do it but talking to us is part of your job as the forum admin. It seemed like you took it and my technical criticism personally which is not constructive. You never did offer up an alternative, instead I got a multiparagraph take down and no direct chance to respond since the thread was then unceremoniously closed. I get closing the thread but it looked like you were grabbing the last word while trying to shut the door. When you close a thread that popular the onus is on you to launch a few new threads for the conversation to move over too. I see we now have a 13 day lifespan to threads. I don’t see how having more threads about the same problems looks better than having one big one but hey it is your forum.

You say you value what we were saying on that thread but it is hard to see that as users when the ratio of users posts to FormLabs ones on that thread is as skewed as it was. I guesstimate something like 10 to 290. I do not want to dismiss your idea of an R&D area but if our interactions continue to look like that thread I do not see the point. Formlabs had the ability to engage more in the conversation and in the process steer it. In my view that opportunity was wasted.

  1. Post some real analysis. I wish I could get everyone to chant “Show me the math”
  2. Be more transparent
  3. Be more involved (no more of this +300 posts vs barely a peep)
1 Like

@EvanFoss — I really appreciate your thoughts. This isn’t a brush-off — but you should definitely open up a support ticket if you haven’t already. While various resolutions can affect your surface finish, you shouldn’t be having such dramatic print failures simply in changing Our team will want to walk through the basics with you, of course, so please be patient. As many here can attest to, as frustrating as it may sometimes be, our support team is second-to-none, and will work hard to make sure your machine is operating well.

I’m sorry I overlooked that! Could you please start a thread about that? I can do so if you’d like. I looked through the previous threads, but I couldn’t find the post you’re referring to — Discourse’s internal search leaves something to be desired. At any rate, it was certainly not my intention to gloss over your contributions, which have been super useful.

Points well taken.

As for your specific technical questions, I apologize for not being more detailed in my responses. We’ve had multiple internal discussions about some of the issues you guys have raised. I’ll see if @CraigBroady can provide more details, as he’s been heavily involved.

I think a lot of the frustration, here, has to do with the timescales involved between the research guys are doing, and how long it takes us to learn about, solve, design, and distribute a solution. You guys have done fantastic work, but at the end of the day, we’re developing a proprietary product, and we have to keep much of what we do under-wraps.

As those backers who’ve been with us since the very beginning can attest to, making a product is slooooooow.

For the record, my job isn’t really as forum admin. I step in here, because I am a real curmudgeon, and love passing on your guys concerns (which I do all the time). I am pretty new at this sort’ve thing, I must admit, so am definitely open to suggestions on how you guys think we can have a better run, happier community! Let’s put em’ over here.

It does appear that the auto-close isn’t what we’re looking for, exactly — what I was looking to do was to prevent people from re-opening very old threads. This frequently happens with very new visitors, who’ll post their first post on an old troubleshooting thread.

1 Like

@Sam_Jacoby I understand the need to keep things proprietary but it is very clear to me that Formlabs needs to work on being more discriminating about what is really secret sauce. There were a number of people on the old thread who had ever reason to be jaded about the form 1+ lasers but got a confidence boost in your product after I told them the back story on the form1 failures. Giving users things like the autopsy reports on their returned printers should be standard operating procedure. There is so much to making any modern product that a lot of information is just not that secret. (By the way doing testing for the 1st 1K/Hrs or a thermally accelerated version of that is standard fair for anyone using diode lasers in things more expensive than laser pointers)

I posted the laser printer scanning assembly where as I said this is a solved problem. The printer industry has not only solved it, they solved it in a commodity product where the cost of the solution has to be much lower per unit than anything in the Form 1+. The printing industry is also not the only group to do this. No secret sauce there.

The secret sauce is not in how the printer’s basic operation works or the design of the circuit boards. All of that some competitor could just duplicate. The secret sauce is in how you squeeze the finest tolerance on things like the laser of the cheapest components. Too be really good at that is about more than design it is about process control and statistics on things like the quality of the lasers delivered. That optimization for cost is where your secrets are for the hardware. I understand if this policy is not changed but understand that the value of an R&D area on the forum is wasted if the dialog is purely one way. With no meaningful feedback and info the users might as well be doing their own open source printer.

I am not asking for anyone there to make a new product. Just fix a design oversight in the original. If we are wrong about this spatial filter and the collimator then so be it. The way things look right now it if I send my printer back a guaranteed solution is not really available to my case. As far as I can tell I will get a changed laser which might or might not have a better profile. I want to help put a fix in and then know the outcome.

Sorry I just assumed from all the posts on the forum that you were the admin. Thank you very much for doing it by default. Your management owes you big time for dealing with us.

PS: I will reprise my laser printer post as a new thread. I don’t know what you are expecting to happen but anything is worth a throw at this point.

2 Likes

@EvanFoss and all, a little bit about what we’re doing here in engineering — apologies for not being super-present here. We’ve got a lot going on, and there are more of you than there are of us! First off, a spatial filter or collimator might indeed do the trick to improve your printer’s optics and produce better prints (in fact, we already use a collimator for the Form 1+ laser), but a lot of design and validation testing is required before issuing an engineering change order. We have several engineers at Formlabs who are investigating potential print issues caused by laser aberrations. The previous forum thread has been instrumental in helping us design methodologies for our tests. Thank you.

Our first step is to reproduce the issues that have been described. It’s actually been really tricky to do so because the laser aberrations are quite rare. No doubt they exist, as we’ve seen in the photographs in the forum thread, but it only shows up noticeably in a small fraction of Form 1+ lasers. We’re working on ways to improve detection. (As a side note, I had similar issues with my Form 1+ several months back. The laser spot looked speckle and most of my prints were failures. The few successful ones has significant rashing. It turned out the first galvo mirror had a smudge on it, and once that was cleaned my printer was printing great. So I didn’t have the same issue described here, but it took a while for me to figure that out).

The most important thing for us is to ensure that any change we make is clearly beneficial. We want to significantly test any change before releasing it to the wild to ensure that there are no delayed negative consequences, and if there are, how we can reduce them to make sure that they impact fewer customers than are currently affected. In the case of these laser aberrations, already just a small fraction of customers are seeing the issue, so we are working extra carefully.

This process of system improvement through rigorous design and testing doesn’t just apply to the laser aberrations discussed here, but to everything we do at Formlabs. We want our products to work great and require minimal troubleshooting, so we work incredibly hard on the backend to make that a reality. It doesn’t always work, but when it doesn’t we work tirelessly to correct our issues. And I’m not talking about putting a bandaid on a problem as a temporary fix or a way to hide it. That’s the last thing we want to do. I’m talking about real, significant improvements.

So hang tight. Your voice is definitely being heard. Sometimes we’re quiet here at Formlabs HQ, but that’s because we only want to communicate information that we know is accurate from theory and practice, that has been proven substantially in-house.

2 Likes

@KenCitron looks like I was a bit previous in suggesting the lens is glued in place - it looks like it’s fixed in place by brass plates.

@CraigBroady - how can you say the beam is collimated? I don’t believe it’s a simple thing to do with a single lens? moreover how do you explain my photos showing beam profile convergence and then divergence?
Form1+ laser flare issues illustrated - pics and video

A collimated beam is parallel light - although obviously not perfectly to infinity in the real world, but ignorant as I am optics, I don’t imagine that light that converges to a point 30 cm from the laser barrel and then diverges can be called collimated.

here’s some pics showing the lens inside the F1+ laser.

@CraigBroady could you please respond, and let us know in what fashion a collimating lens is currently being used in the Form1+, and why @KevinHolmes got the result he did in his experimentation even with a collimating lens? Even if you respond and tell us that information is top secret that would be better than no response. @Sam_Jacoby could you also point him here to this thread in case he hasn’t seen the questions @KevinHolmes asked? This is a great opportunity to start a trend toward more open communication.

1 Like

@RocusHalbasch A lot of our optics details is formlabs secret sauce, so I can’t say too much about it. What @KevinHolmes writes in the previous post makes sense to me based upon how our system works. There’s much that can be done with collimated laser light.

@CraigBroady I like that you wrote something, but that’s not a response, I just contradicted you - and you said “that makes sense” - come again?

Well that is unfortunate. It seems a little ridiculous to me as I’m pretty sure you are not doing anything that is groundbreaking, and keeping it “under wraps” is not preventing a relatively resourceful competitor from figuring out what you are actually doing. I’m guessing someone who knows a thing or two about lasers and optics, who had a Form1+ that they didn’t care about the warranty on, could tear it down and figure out what you guys did. If that is true the primary group being hurt by your companies silence is likely the community of customers who do care about their warranties, but want to explore the printer and options for making it better. But I suppose that is often the nature of “secret sauce”.

:frowning:

Thanks a lot for getting back to us though, I think we all appreciate it.

@CraigBroady [quote=“CraigBroady, post:33, topic:3398”]
optics details is formlabs secret sauce
[/quote]

seriously?! the optics is your secret sauce? not the software or the custom PCBs, but the “optics”.

Stuff that is easily quantifiable by anyone prepared to open up your laser and measure the focal length of the apparently cylindrical lens (perhaps there is another one behind it then) in the laser.

That’s your barrier to entry in the commodity SL game - that some company is not prepared to spend three thousand euro to dissect your laser?

That’s ridiculous.

@KevinHolmes I doubt that is there sole barrier to entry. From what I have divulged through previous conversations with Formlabs I would guess anything inside the printer or under the big orange hood, that is not overtly obvious, as well as any process they do to any part of it that could be done multiple ways is “special sauce”. Oh yeah and the software too. However yelling at them won’t change it, I’ve already tried. Hopefully with time we can convince them they are hurting their customers and themselves by being ridiculous with where they draw the line of what they can and can not talk about. However @CraigBroady is almost definitely not responsible for drawing those lines.

I think one major problem is Formlabs seems to be of the opinion that the Form1+ is somehow a huge breakthrough with lots of secrets that need to be protected. However in reality a large portion of what it is is a fairly simple and straight forward machine, that like most simple and straightforward machines is not something you can keep secrets about. They don’t seem to recognize or care that trying to keep secrets that take little effort or money to discover is a waste of time and possibly harmful to their users. I could say a lot more on this topic, and will if anyone cares, but I’ll stop there for now.

1 Like

@RocusHalbasch certainly I don’t expect @CraigBroady is responsible for delimiting the bounds of FL corporate IP - but in this case I see either one of two possibilities - either you just gave @CraigBroady an easy out with your previous post suggesting he respond “secret sauce” and he took it - or he genuinely thinks the optics are corporately sensitive IP.

In the latter case - I think it relevant to remind him that’s a ridiculous premise - because even if he’s right, and FL as a company thinks their optics are part of their secret sauce, he still cannot be without influence. He is a “Form1 whisperer” after all.

Plus I wasn’t yelling - I am irritated, and I was hoping that came across - after all he did have the gall to throw mirrors into this discussion.

1 Like

There is a (maybe not so) fine line between purposefully mystifying your product and protecting the “secret sauce”.

I might be completely wrong on this, but i think Formlabs is on the former (no pun intended) side of that line more often than not. Which, if true, is quite likely due to corporate culture hastily imposed on a startup.

2 Likes

I assure you I am not a Formlabs shill. :smile: I just have run into Formlabs employees dropping vague hints and then backing away claiming they can “say no more” on multiple occasions. I was wondering if his lack of response was because he had “said too much”.

LOL I suppose that’s true, carry on.

That’s 7 years of bad luck right there.

@CraigBroady Thank you for trying to respond.

@KevinHolmes I think what that was meant to imply is that the photo is consistent with how the optics assembly would appear if photographed as you have it. I imagine the following two scenarios could be at play here.

There is every possibility that the definition of collimator to us as a two lens system is only one to them. After all they only need a beam to be more or less coherent until it hits the resin. To us that would be focused for that limited context though it would fit some definition of collimated.

It could also be that there are two lenses in there and that one or both of them is at the wrong distance to properly collimate. Manufacturing tolerances are as much a function of price as they are design.

Yes ether one of these means that the spot size will drift as the beam moves off the center of the resin tray.

Ether way @CraigBroady has said the following which means we know they are taking our issues seriously. I guess all we can push for now is for formlabs to change how they QA the laser’s collimator & spatial filter. In the meanwhile I think they deserve a chance to show some progress on the issue.

@CraigBroady seems to think this issue affects a minority of users. I am curious to see how small a fraction is. I know of a fraction that is. My gut says that this is the kind of bug most users do not understand that they even have. I mean if you are coming from the world of poorly built FDM machines even a bad day with a Form1+ would look pretty damn perfect.

@KevinHolmes and others I wonder if we could find more users and convince them to print your giant plus signs. I think more data would help narrow down just how wide the tolerance on these parts is.