Cheaper LED's and a look into LCD for Exposure Split from Saati Screen Toaster Thread

[jvanick]

The Saati unit is at 450nm, seems a bit high on the scale for most emulsions.

this is not true.  Their units are 405nm. See the previous post with the lamp in it.

According to Saati's techs, they've found the best and deepest exposure to occur on their emulsions with a 405nm source due to the fact that the most sensitive area (395nm) also absorbs some of the light, and the areas around most sensitive areas allow for a slower 'cook' time.

[Prōdigium]

The Saati unit is at 450nm, seems a bit high on the scale for most emulsions.

this is not true.  Their units are 405nm. See the previous post with the lamp in it.

According to Saati's techs, they've found the best and deepest exposure to occur on their emulsions with a 405nm source due to the fact that the most sensitive area (395nm) also absorbs some of the light, and the areas around most sensitive areas allow for a slower 'cook' time.

A type error, 450 WATT, 405nm....to clarify that sounds better, but still 405 is just outside the range of true UV. But what your saying is that 395 is TOO efficient and works too quickly. BY going into the white light spectrum a bit more they have in fact given a wider latitude for exposure calculation errors, which makes sense as correct exposure times are an art & a science that most people never get right. So 405nm is more "User Friendly" but not as fast.

I will still play with my 395nm box because its what I already told my friend to make me and in the end I think will produce better results. If exposure latitude becomes an issue, there is this thing called a dimmer and the wattage can be dialed down to slow the process, kind of like a medium setting on the microwave, not everything cooks best at full power.

[jvanick]

The Saati unit is at 450nm, seems a bit high on the scale for most emulsions.

this is not true.  Their units are 405nm. See the previous post with the lamp in it.

According to Saati's techs, they've found the best and deepest exposure to occur on their emulsions with a 405nm source due to the fact that the most sensitive area (395nm) also absorbs some of the light, and the areas around most sensitive areas allow for a slower 'cook' time.

A type error, 450 WATT, 405nm....to clarify that sounds better, but still 405 is just outside the range of true UV. But what your saying is that 395 is TOO efficient and works too quickly. BY going into the white light spectrum a bit more they have in fact given a wider latitude for exposure calculation errors, which makes sense as correct exposure times are an art & a science that most people never get right. So 405nm is more "User Friendly" but not as fast.

I will still play with my 395nm box because its what I already told my friend to make me and in the end I think will produce better results. If exposure latitude becomes an issue, there is this thing called a dimmer and the wattage can be dialed down to slow the process, kind of like a medium setting on the microwave, not everything cooks best at full power.

the Starlight is a 395nm source...

you can also use distance... we found that even adding 1/4" of distance on the starlight slowed down the exposure time due to the inverse square law.

we expose with the starlight and then post expose with the previous generation saati 300w (the multi one that looks a LOT like the pic you posted a few days back)...

[Full-SpectrumSeparator]

An exceptionally interesting concept, and accepted proof of concept. The issue will remain that most, if not all LED/LCD screens will simply NOT produce enough power (per pixel) unless specifically manufactured for this process. An expensive endevour to say the least. This would especially be true for trying to expose a thick cap film, but could have application in the printed circuit boards field?

I think there is still a bit of confusion as to what is going on.

-"The issue will remain that most, if not all LED/LCD screens".... 

Hold on...  LED/LCD are not interchangable terms.   These are not LED screens, OLED, etc...  it is a typical computer/TV monitor, which even when they say "LED" they are talking about the light source which changed over from miniature cfl tubes to LED strips along the sides or bottom of the monitor panels years ago, but the LCD - Liquid Crystal Display - panel, is still just a transparent panel of RGB - Red, Green, Blue - LCDs.   

You simply take apart any regular computer monitor or LCD tv, and separate the parts, you just don't use the LED backlight strips at all or the diffuser sheets or glass....

So the other part of the sentence.. " will simply NOT produce enough power" --- The LCD panel is not producing ANY power output of light, at all.   It is PASSIVE, and we are not using the backlight of the LEDs inside the monitor - we are DISASSEMBLING the monitor, and then putting a new light source behind it -- these high-power UV LED chips etc. 

" (per pixel) unless specifically manufactured for this process." -- But there is no need to change the manufacturing process to UV LCD's, because for one - the BLUE LCDs allow much of the UV light to pass, but the Green and Red pixels still allow it, and the entire panel as a passive semi-transparent screen - even though there are polarizing filter-films on each side of it that also have some UV-blocking... it is not absolute... and the amount it allows through is still enough to cure the entire emulsion in about 12 minutes - this is with a small 3-watt battery-powered flashlight UV LED that is just directly behind the LCD panel...    It goes  (HIGH POWER UV) ---> LCDs --->Emulsion.     The distance and how much affect it has shows just like anything that the power "through-put" needs to be ramped up more than when you have no LCD panel between the UV light and the emulsion.    But this is not some crazy science or advanced thing you have to do...  Maybe picture it this way...   Put some tinting on your glass on your exposure table, or if the glass were like sunglasses.   You'd have to add more UV power, maybe move the lights closer, etc... to get your screens to cure.   But I bet you could get screens to cure in the sun even with tinting or a pair of sunglasses on them, and so adding more UV power to a light table to get it to cure through another layer that is absorbing a lot of the UV light, but adding a digitally-controlled masking-technique with resolutions capable of spot and halftone with details, so I'm only trying to test whether the larger industry would benefit from the trade-off.    Or if the cost-effectiveness for me to even build my own DIY table for my screenmaking and printing purposes would work with these newer UV LED high power chips etc.   The LCD panel is just taking the already current cost-effectiveness of older cheaper LCD monitors and Tvs and it is very easy to take apart and then you can just tape the panel to your glass, maybe I will post another picture of some earlier tests trying it right on the fluorescent-bulb light table... which was clearly not enough power.   But I am saying maybe even these units like the Ryonet table you could just tape a LCD panel on and you've got digital CTS masking with no film etc, and only have to cure it longer like a minute or two probably.    I tested 3 watts up close to 12 minutes, what would 450 watts from a foot or two be like?  Or maybe just move it closer - get a smaller area but thats one of the things with the panels and actual-size vs. resolution etc... but still, this is not something expensive to test and get going, the manufacture of new types of LCD panels is not required at all, and might even be pointless because of how much UV is actually getting through the RGB LCD panels anyway.      But I will test further once I get my hands on some of these bulbs and chips of higher power and come back with more specific data and videos etc.     Exciting stuff and thanks again for what people are sharing here, this thread is getting deep... I'm all about the LED curing but I'm just throwing in a whole other variable for myself and making it harder obviously, but its only a simple thing I NEED MORE POWER, lol. 

Ps those photos are from over a year ago, and those tests didn't expose with just the fluorescent bulbs... then I tried UV LED strips, those cured in a long time and sort of not enough the screens would fall apart washing more because curing all the way through would take too long or just the little LEDs on the strips are very low output,  so when I tried the LED flashlights battery-powered and up close to the panel is when I got it down to like 12 minutes, but still just the area of the LED light behind the panel, so either I have to build lots of LEDs into a grid-panel like the other one you posted, (or purchase existing units to test), or try the point-source bulbs from various distances and maybe multiple chips in combination to put a very strong and large-area spot of UV to go through the LCD pixels... but they certainly pass the light in the UV spectrum at times and distances and powers that are easily pushed into the acceptable range by further testing with higher power.   This is going to get exciting.   Maybe I have to find that other example of the ways people are using it successfully in 3D printing which uses UV curable resins in the same ranges as we do....  screenprinting is just one other application... I feel it is a transitional stage between doing fully 3d-printed stencils if those are ever feasible.

http://www.lcd3dprinter.com/

http://3dprint.com/5429/rayee-lcd-sla-3d-printer/

http://www.buildyourownsla.com/forum/viewtopic.php?t=84

https://www.youtube.com/watch?v=MS1t3tHsx9k

[Full-SpectrumSeparator]

One more note about the LCD digital masking technique...   

If we are eliminating the printing of ink or wax onto film or the screen with an inkjet/plotter print head, which can take a few minutes to complete,  and even the CST DLP machines plot over the screen...   then we are eliminating those minutes - not just the consumables of film and its registration/placement time, or the ink/wax etc, or eliminating using a machine with moving parts...     so if the exposure takes maybe 2 to 4 minutes (although I'm sure with enough power or closing the distance it will come down to under a minute and in those 30-second or less ranges) -- but even if it takes 4 minutes but you're able to load a screen and just expose,  not having to wait for RIP or printing the film/ink/wax onto the screen... then you are saving a significant amount of time over even the current DTS machines, let alone the LED exposure where you still have to print a film.      If the exposure is 10 seconds on  a powerful UV LED table, but you still have to wait 3 or 4 minutes for a film to print, or you have to wait 30 seconds for thermal to output and then tape and place on the screens etc... it is just going to add up over screens until it is hours per day and more being saved.... but that is just one potential time/labor/materials savings with this technique.... even if the exposure is a few minutes. 

[Gilligan]

Don't forget about the tape... THE TAPE!!!!

[Frog]

Don't forget about the tape... THE TAPE!!!!

Isn't that what they told Nixon?

[mimosatexas]

This is a pretty similar idea to the DIY projector builds that were all the rage when I was in college (old school overhead projector for transparencies, but with a torn apart lcd monitor on them instead of the clear glass above the fresnel lense.  They worked "ok". 

As far as replacing film or dts, the main issue you are going to have regardless of any of these proofs of concept you've done with LED's is that the LCD is not going to block enough UV to get a quality stencil no matter what you do.  You can prove this with any exposure light source, and the LED flashlight isn't going to prove this better than a MH or flo tubes, etc. You also have other issues with LCD, like the fact that lots of UV is destructive to the LCD, it will scatter light like crazy and cause undercutting (even with adjustments over time like you discussed).  The LCD will act like a really faintly printed film that is not vacuumed to the emulsion and exposing through a super thick not optically clear glass and over time will fail due to the UV exposure.  Neat idea, but a long, long way from realistic.

This stuff was actually being discussed at least a decade ago for etching circuit boards, but I am pretty sure it was pretty much definitively abandoned.

[ebscreen]

Thanks for saying succinctly what I was trying to convey. We're not the only industry that wishes this would work.
It's been tried.

Using a projector at close distances looks interesting however. I forget the three letter acronym they came up with
for that one.

[Frog]

I seem to remember emulsions or stencil films of different sensitivity to be used with a projector at the flat stock shop I worked (though they went with this around 1978 after I left) I do not know what the light source or power was.

[Full-SpectrumSeparator]

This is a pretty similar idea to the DIY projector builds that were all the rage when I was in college (old school overhead projector for transparencies, but with a torn apart lcd monitor on them instead of the clear glass above the fresnel lense.  They worked "ok". 

As far as replacing film or dts, the main issue you are going to have regardless of any of these proofs of concept you've done with LED's is that the LCD is not going to block enough UV to get a quality stencil no matter what you do.  You can prove this with any exposure light source, and the LED flashlight isn't going to prove this better than a MH or flo tubes, etc. You also have other issues with LCD, like the fact that lots of UV is destructive to the LCD, it will scatter light like crazy and cause undercutting (even with adjustments over time like you discussed).  The LCD will act like a really faintly printed film that is not vacuumed to the emulsion and exposing through a super thick not optically clear glass and over time will fail due to the UV exposure.  Neat idea, but a long, long way from realistic.

This stuff was actually being discussed at least a decade ago for etching circuit boards, but I am pretty sure it was pretty much definitively abandoned.

Excellent.    It just makes me want to do more tests and take it further.    I love the critical thinking it is important, and the feedback about previous attempts is great, don't want to repeat mistakes just take what is being done and improve it.

Did you see the 3d printer links?   It is working excellent in 3d printing... same specs.   

I got a cure and a printable screen, I could make it work with just a larger stack of the flashlights and a little more distance and longer exposure time.   

Was there this kind of UV LED technology a decade ago??

So one thing I want to be clear on in your post was "regardless of any of these proofs of concept you've done with LED's is that the LCD is not going to block enough UV to get a quality stencil no matter what you do."....

Did you see I already got a quality stencil with my test??   Yes it was only in the areas exposed by the light and blocked by the LCD stencil being charged in those pixels to block the light.   

I think it is fascinating but typical when presenting new concepts and obviously needs to be repeated and shown in a full-area screen exposure, and others will need to repeat and print with the screens... then maybe you will realize it is not some novelty useless gimmick.     But all the critical thinking and attacking it from every angle is necessary to prove or disprove potential values in various segments.     I already know from my tests I can get a working unit put together myself and burn screens for jobs either spot color or halftones and print with them... sooo.... what is it that I'm not getting about having a cured stencil that prints??     

Yes all sorts of tests about the quality for duration, and the max "resolution LPI detail etc" -  etc, but why not post-expose it in the regular high-power UV after washing if breakdown is a concern then?     

I understand that adding power to the UV will increase the amount of light still getting through the LCDs that are set to black pixels which cut the light significantly... but if I am simply going for the same amount of cure I already got, within a shorter amount of time perhaps but the light power and distance factors will be checked to find the best curve....   but how is it going to be radically different than getting a cure in the areas you want and no-cure in the areas you don't want, as already shown in a low-power-long-time-close-distance scenario?   

The amount of actual wattage in a given area might be similar with the 3watt flashlights up close behind the panel, compared to the 450-watts spread from a distance over a larger area...  I'm not exactly sure about this... but are you getting 450 watts of UV light right at the surface of the emulsion in every area of the screen?  Or is the point-source of 450 watts spreading the 450 watts of output light and with distance and the final area-coverage of the spot-light of rays actually cutting down the amount of watts per square-inch for example?   Or are you getting 450 watts in every square inch or whatever dimension is measured?     

It should be apparent that I am getting photons to pass through the pixels to expose the emulsion, while the LCD's that block the light are preventing enough photons from curing in that area so it works just like putting UV through a film.    Also, it is not going through some thick layer of glass after the lcd panel, there is no gap any different than that of a film really...  the scatter of light through the black pixels is still dark enough while having the higher power uv light through the clear pixels to prevent it from curing, and it will wash out... so you dont even care about that part, it washes out, its not going to "break down" later lol.    It breaks down when you wash it out, and the cured areas don't break down, and if I want maybe its a longer cure or a post-expose at just a few seconds without the LCD panel, then you're definitely creating a "quality" screen.    I think the really high-end quality will take higher-resolution LCD and more testing to achieve,  and obviously the CST machines use a DLP chip focusing on a small area and plotting across so they get 2400+ dpi resolution.... but most printers honestly will be able to work with a 100 dpi exposure and even get low LPI halftones and detailed small text and lines/shapes to hold, and cured stencil screens that you can print with, post-exposed if the cured areas are not truly strong enough but they behaved as strongly in wash-out as any other screen i've done.    We post-expose anyway.     Don't get me wrong, I appreciate all the feedback and it is an open project or idea for anyone to test.    I just think giving up without truly testing a full-area exposure would be too soon to call it "definitively abandoned", but  I am not satisfied until someone shows me tests that were done and I try it myself and can't repeat what I've already done for some strange reason or another, I'd like to find out why, not just take "pretty sure" and "pretty much" as definitive scientific proofs that it is "impossible to get a quality screen" --  especially when I've already got a quality screen and only needed to put more flashlights if I wanted and a little further away for longer duration.   I even tested some of that and haven't shown those pics yet because I want to video the process and the results for that step by step to make it more clear the extent at which it has already been proven to work... it is just kind of pointless to spend hundreds on a bunch of $10 flashlights when I'll get a larger panel or single-point UV bulb to test.   

But I will be the first to show and accept any results I get, good or bad, as I continue with the testing and improvements, and accept defeat if it truly turns out to be a dead-end.   

" The LCD will act like a really faintly printed film that is not vacuumed to the emulsion and exposing through a super thick not optically clear glass and over time will fail due to the UV exposure.  Neat idea, but a long, long way from realistic."

--- The LCD acts like a sufficiently dark ink printed on sufficiently clear film.  This is shown and proven in the areas the 3watt LED flashlights exposed and where the LCD pixels that were black blocked the UV light from the flashlight.    Did you see the screen and the image of the printed sample?    That will work for MOST printers these days, as long as it is obviously the full-area of the screen not just spot-lights where the flash-light LEDs were hitting through the LCD panel to the emulsion. 

--- It can be vacuumed to the screen just fine,  it sits ON TOP of the thick-optically-clear glass if you want, or no glass at all...  but there is direct contact from the LCD pixels through a thin piece of polarizing film but the undercutting it not as bad as you seem to think...  what is the part that will fail due to UV exposure?   I understand that maybe the LCD panel will have issues where the RGB pixels perhaps break down or something due to the UV exposure??   Are you just saying this or do you know that UV light breaks down LCD panels over time??    The panels are cheap, unless talking about 4K and larger size monitors...  but you realize this also will work at oversize dimensions... maybe poster printers it will save the most time and costs and labor with printing large-format films etc....   just get a 42" monitor/TV and you can do 24" x 32" screen exposures probably at smaller DPI/  larger LPI but that is normal for poster printing.       

In my own test I have shown it is not "a long, long way from being realistic" -- it is realistic already, its real, a screen was exposed and printed with it, even a larger full-front design but it had some minor issues due to just using lower power lights trying for a greater distance... but on both examples you can see that it only takes another test with the full-area consistent UV light and the right distance/exposure time to get a fully exposed screen stencil of a design at normal print dimensions, and I suppose it will make more sense when I show that a full-area exposure can happen and a screen taped and printed would hold the design successfully.    Maybe I have to use it on a real customer print job to show it works successfully to make a "quality" screen that can print a normal job with it.     I'm just not taking any of the critical assessments as "definitive disproofs".   But I appreciate all of the critical feedback we need to attack it from every angle to see if it will have any potential value for any potential market segment.   Thanks again, love this type of discussion and experimentation and it is already making progress through this.

[Full-SpectrumSeparator]

Thanks for saying succinctly what I was trying to convey. We're not the only industry that wishes this would work.
It's been tried.

Using a projector at close distances looks interesting however. I forget the three letter acronym they came up with
for that one.

Umm... I won't keep contending anything until I show more proofs and evidence and you can repeat it yourself. 

But you do realize from the other links I posted, that 3D stereolithography machines are already being built and sold and work with this same exact technology??

We aren't the only industry who has tried it, perhaps the one who gave up on it, but obviously it works for other purposes and other industries, using the same variables - UV curing photopolymers and UV LED spectrums through LCD panel masking.   

I find it very interesting that there seems to be a consensus that this "doesn't work / will never work" -- when it does work, and already is working.   

I will wait to show more and produce a regular customer's print job with a screen made from the process and video the process so you can maybe have more opportunity to push for more of the "quality" that we would all like to see.    But there is no "not possible" with this at this point.    Maybe others have given up on it in the past, and I am not talking about "projection" with DLP or through LCD projectors either.     This is a regular LCD panel, up against the emulsion/mesh, and with UV light behind projecting through.   

Is there something I am missing, other than a few "consensus opinions" that it still is somehow "not possible" ...??

But I do understand what I have shown looks crude and not usable at all,  the test was not for a printable screen,  I was testing the cure-through and dark-blocking aspects to be sure that indeed, it would work... and it does.   The only next step for a "usable screen for a real job" is just the full-area coverage of the UV light at the right power and distance for the right timing....   but I guess some will not "believe" it until they see a real job being printed with a screen made with this process?

That will be fun, I'm just sharing this because others already have the high-power methods to try it out if they want to take apart an LCD monitor and test it themselves.

I'm saving up for the lights now, maybe I will just get more flashlights or put them together in a way that tests a more connected area exposure,  perhaps if I just show a left-chest area next printed successfully for a real job?? 

[mimosatexas]

Not going to write a novel since I dont have the time, but basic points:

• No one is attacking you because you are doing something new or outside the box.  The point is the concept is not new and the pitfalls have been tested and discussed for a long time.  Don't let me or anyone else stop you from further testing, but understand that we are responding based on a context that is much larger than your specific experiment.
• I don't consider what you have shown as a usable or quality screen, which is why I said what I said.  You have failed to prove it is viable beyond what you have shown, which is lacking.
• 3D stereolithography is not the same, so I'm not sure I understand why you're bring that up.
• you misunderstood my description of the issues with the LCD, as I was specifically discussing the downsides of the panel itself in relation to how it passes the UV, not whether glass or a vacuum are required or how they would interact with the panel.  the panel itself is akin to a thick piece of non-optically clear glass, and the way it blocks UV is akin to a film that has not been vacuumed tight to the emulsion.

[ebscreen]

The bottom up 3d printers specifically want to cure the thinnest layer possible at a time. We essentially
want the opposite. Their medium is much much more UV sensitive as well.

I tested it awhile back, the results weren't usable, I looked into different panels and found that the very
composition of an lcd panel precludes what we are looking for. That was where I ended.

By all means continue your testing. We wouldn't have any of the inventions that we have if everyone quit
when someone said it wasn't possible. My only reason for commenting was you  had asked if anyone has
ever tested it before. I had, it sucked.


My comments on the projection idea sort of takes into account what Frog is referencing, I believe Murakami
still sells an emulsion meant for projection? But the idea that the 3D printers use is to remove/replace the focal lens
and use the projector at super short distances. For us probably 3' or so. In 3D printing the light from the projector is
enough to cure their medium. We would likely require significantly long exposures, but as you've mentioned, it
bypasses an extra masking step.

[ScreenFoo]

FSS:  No one is saying you can't successfully expose a stencil with LCD tech, but are implying the quality is not there compared to traditional methods, rather akin to a lot of spirited discussions on here over the last couple years. 

FWIW, saying 'sufficiently dark' and 'sufficiently light' are extremely subjective terms when it comes to stencil making.
Plenty of people who once used halogen lights and oiled paper positives would agree with me... 

Finally:  Just a suggestion mods, but how about a thread split with a realistic title for all this stuff that has little to nothing to do with Zane's project or Saati's LED unit?