UPDATE: LED at SRI

[alan802]

I'm happy to finally be able to give an update to the long thread I started a while back.  Vastex did what they said they would do and bought the exposure unit back.  They deserve a HUGE KUDDOS for doing that and I do appreciate it in so many ways.  I don't want to get back into what got us to this point because it's over and done with, Vastex did right by us and I'm grateful for it.

We've had a Starlight in production for I guess about 3 weeks, maybe 4, can't remember off the top of my head.  I've ran all of the same tests and without getting into a "versus" thread I'll say that we're able to get an acceptable range of halftones in the 65lpi column, 5%-96% slightly underexposed, 10%-97% with proper exposure (Chromaline Proto).  The HVP performs better but I've only done 2 expo calculator tests on it so I'll wait before I post those numbers.  To be honest, we've only had a few jobs the last few weeks that we could test for stencil durability on press and on a 576 piece run we did have a 180/48 breakdown but the 2 280's held up the entire run just fine.  We also did a 1200 piece white-only print but we print white ink with such little pressure I wouldn't expect the screens to breakdown like when we use more pressure and harder blades like black ink through a 180 or 225. 

Over the next few weeks we'll continue to test emulsions to get the best one that fits the Starlight. Right now the HVP has given us the largest range of halftones by enough of a margin that if it performs ok on press then we might go back to it.  The Ulano Orange was ok, nothing special, Kiwo Mult-tex was awful in just about every category, I can't believe it did so poorly.  The Chromaline Prototype we've been using is really good and my screen guy is pushing me to keep it but he's not looking at the big picture, he's just looking at his part of the process.

A couple of Fridays ago I pulled the old Richmond Solarbeam out from the corner and did a few calculator tests on it, while we still had the Vastex plugged in.  We hadn't done much testing on the Starlight at that time so once we get more time with it it will be a better comparison.  There was a fairly even gap between performance of the 3 units using the Chromaline Prototype and I believe we would have gotten different results using a different emulsion but I feel that the standing between them would remain the same.  I feel though that the Starlight would close the gap with the Richmond with the HVP because I don't think we could get much better results on the Richmond because we held a 2%, 65lpi dot with the Chromaline.  The only thing that could have been better was the high range, and we're only talking about a few % so we reach a point of diminishing returns due to how well it performed originally. 

At this point we're around 15 seconds on all of our white mesh, and I'm trying to find the time that gets us the strongest stencil on press without struggling to get the lower % of halftones to spray out with our high mesh counts.  150 and 180 is doing well at 30 seconds and once we get a few more decent stacks of shirts through I'll know where we can go with that number. 

One thing I can say now that we've used 2 of the main LED units on the market is the latitude you have with white mesh versus yellow mesh seems extreme in their differences.  Others might not agree, but I don't think LED units work well with white mesh.  The latitude is very small compared to what you can get away with on the yellow mesh.  15 seconds is bordering on overexposure with our white mesh and they aren't coated thin, they are all around 20% EOMR.  10 seconds and you'll see noticeable underexposure, 20 seconds and you'll be lucky to get your regi marks sprayed out.  Yellow mesh exposures you can see very little difference in how the stencil behaves when spraying out from 10 seconds to 90 seconds.  I haven't been able to overexpose a yellow mesh yet but the longest I've gone is 90 seconds.  It's easy to see why white mesh performs differently than the yellow but the gap between them seems large to me.  Single point obviously handles the light scatter and undercutting better and I'm not surprised that we see a difference but it's the amount that gets my attention.  With all of the experience we have with a good metal halide we know there is also a difference in yellow versus white mesh but it's nowhere near the extreme that we see with the LED units we've tested.  Just something to think about and perhaps consider when choosing your mesh in the future. 

Sorry for the long read but there is much to talk about.  Again, I'd like to give Vastex credit for standing behind their product.  I am ready to see how our production is affected this year with the change in units and I'll continue to update our experiences and testing.  Within the next few weeks I'll have a more solid opinion of the Starlight and where I think it fits and how it compares to the others we've had.  I'll also give as much info as I can on the different emulsions we're testing and which one we end up sticking with.

[Sbrem]

I appreciate your work on this Alan, thank you. Regarding the lack of latitude with the white mesh, which make perfect sense, what about a timer that lets you enter 1/10ths of a second? Seems like the quickest fix for fine tuning; just a thought, now back to work...

Steve

[zanegun08]

Kiwo Mult-tex was awful in just about every category, I can't believe it did so poorly.

That is interesting, we use Kiwo Multi-tex as our only emulsion.  We switched from Kiwo Versatex before, and did initial testing before it came to market.

I will say though, that you must make sure your screens are completely cured before imaging, otherwise we ran into issues.  Once we were curing our screens more thoroughly we have great results, and only harden or post expose on water base jobs of over 500 pieces.

Maybe give a another shot and let the screens cure more, as I remember it being in the lower end of the emulsion price range, which is huge savings for us a year because we do a lot of screens.

[Mr Tees!!]

I find the photopolymers to be a bit too sensitive for the Starlight, at least for my liking. We use CCI WR-14 (a diazo emulsion), and get a good burn on any mesh at 40 seconds (really anywhere between 25-60 is good, YRMV). I have found the stencil to be more durable as well. we have done DC runs of up to a few hundred with no issues.

[Colin]

Alan:

I forget, are you still using film?

[alan802]

Alan:

I forget, are you still using film?

Yessir.

Zane, yes, it dried slower than the others and a full 24 hours in a 30% humidity environment didn't provide the results we're looking for.  It was very tacky on the shirt side and required a ton of baby powder to keep film ink from peeling off.  I still have some left and we can work with it a bit more in drier conditions if it performs well in other areas.  I just received a quart of Chromaline Chromatech WR from M&R and will put it through some tests this time tomorrow.  I've got a routine where I have 5 different mesh counts ready and I shoot them all the same way for each emulsion then analyze.  I also shoot the Chromaline expo calculator along with our own film that was done with our Epson to get a more accurate result to what we'd see in normal production. 

[GraphicDisorder]

I am with Sean. We like Diazo emulsion, never had a screen break down under several thousand prints before. Takes a little longer to expose but my guys get a flow going and burning is not the bottle neck.

[alan802]

I'm certainly not looking for an emulsion that will burn in 10 seconds or even 20 seconds necessarily and when we were using the Richmond there were times with older bulbs that 90-120 second exposures were the norm.  The way our flow works it wouldn't hurt us in the overall scheme of things to spend 1 minute or 1.5 minutes per screen for the expo process.  I have changed my tune the last year or two in regards to fast exposures and I'm a little looser on sticking with a pure photopolymer and will entertain a dual cure or diazo emulsion as long as we're not talking several minutes-long exposures.  If the level of detail is where I want, stencil durability on press is thousands of imprints like it was with the Richmond/metal halide, and reclaim isn't a bear along with 60-90 second TOTAL (vacuum and burn) then we'll have a winner. 

[GraphicDisorder]

I'm certainly not looking for an emulsion that will burn in 10 seconds or even 20 seconds necessarily and when we were using the Richmond there were times with older bulbs that 90-120 second exposures were the norm.  The way our flow works it wouldn't hurt us in the overall scheme of things to spend 1 minute or 1.5 minutes per screen for the expo process.  I have changed my tune the last year or two in regards to fast exposures and I'm a little looser on sticking with a pure photopolymer and will entertain a dual cure or diazo emulsion as long as we're not talking several minutes-long exposures.  If the level of detail is where I want, stencil durability on press is thousands of imprints like it was with the Richmond/metal halide, and reclaim isn't a bear along with 60-90 second TOTAL (vacuum and burn) then we'll have a winner.

We are burning screens at about 20-50 seconds (no vacuum) with Diazo. So I think your gonna be good on that time line if you add in vacuum.

[ABuffington]

Good stuff Alan,

Some things to consider which you are finding out.

1. LED = chaotic light.  Yellow works better than white to prevent undercutting, but white will expose stronger and faster for vector art.
2. Pure Photopolymers like HVP or SP-1400 expose better than a dual cure.  With a dual cure the SBQ portion will expose before the diazo can expose at all.  Avoid dual cures.  (For me info PM me on why, but our tests have proven the inside of the screen diazo portion will not be exposed.)
3. Each emulsion should be tested with a step test to determine exposure times, then use the derived time to test for resolution test using your 2-98% tonal gradation.  Are your tests using an  exposure calculator to determine halftones achieved?  Be aware that on LED's 2 layers of film blocks a lot of light.  Whereas a step test uses one layer.  The slight difference in time can amount to 1 second which can be a large percentage of difference, 1 second in 15 is 6.66% thats a huge swing in exposure on an LED.  Using tenths on the timer does help fine tune for lower tonals, plus durable exposure, plus capturing higher tonals that do not fall off in development.
4. POST EXPOSE all screens on the squeegee side for Pure Photopolymers.  No matter what time you use, the squeegee side of an LED exposure is soft.  We have had the luxury of having a Starlight for several months.  They can achieve good exposure, but the mesh counts need to be a bit higher, (S Mesh helps here for baseplates) and the emulsino thickness needs to be controlled.  The stencil exposes better with a little less emulsion.  6-8% instead of 15-20% EOM.  In all cases the inside of the screen is a soft exposure, and no amount of extra time helps, it will start undercutting needed details first before durability is achieved, and even then our tests on 300 mesh still benefited by post exposure.  We could print short runs without post exposure, but if sensitizer on the squeegee side is partially cross linked, screen openers, mineral spirits can chemically flash the sensitizer and lock emulsion in.  Highly recommend using your Solar Beam for post exposure with the vacuum off and light on, squeegee side down.  It is multi spectral and will create stronger screens that reclaim easier.  LED intitiallty helps speed up exposures.
5. HVP is not our best resolving emulsion.  For plastisol try Photocure SR for incredible halftones, or SP-1400 for all inks and more resolution.  Resolution is emulsion thickness dependent.  Coat less, switch to sharp edge, or thin out with up to 2-3 oz of water. You can thin out SP-1400 a bit more when adding the diazo for a thinner emulsion coat and finer halftones.  Switch to sharp edge with a thinned emulsion, expose and post expose well.  SP-1400 does contain a bit of SBQ type sensitizer, so it helps to post expose it as well for more durability and easier reclaiming.

Metal Halide is not broken folks.  LED is the future however, but with some trade offs.  Thick stencils and thick film can be difficult, but for most printing they work fine and the speed is admirable.  Curving halftones to stop at 6% and having 95% images easier with less vignette moire than trying to obtain halftones below 4-6% or holding halftones above 95%.  On most base plate meshes the thread is bigger than the <4% halftone we are trying to image, and on the upper end the mesh opening is bigger than the negative 95% tonal so it has nothing to hold onto but air when it lands in the center.  You'd have to print 280-300 for the base to hold these tonals and those days are gone with 225/40 and a good curve.

Al

[Orion]

A few thoughts on Mr. Buffington's post. Pertaining to pure photo polymers and dual cures my suspicion is that LED lacks the amplitude to completely punch through the emulsion layer before the effects of absorption and scatter occur. Even if you get no scum with the "blot" test, is that stencil fully exposed? Post exposure helps but complete exposure before developing is optimal because initial exposure is key. To compensate, a thinner EOM is suggested, which is not really a bad thing because I think that we should be more concerned with Rz (flatness) of the stencil.

LED is the future, but not quite all the way there yet.