TSB
screen printing => Equipment => DIY - From master engineered marvels to cobbled together jury-rigged or Jerry-built junk! => Topic started by: Pangea on January 20, 2018, 09:49:44 AM
-
I'm in the process of rebuilding my LED exposure unit and figured I would document it here.
Through looking at some tech sheets it looked like most emulsion crosslinks at about 395 nm wavelength. I was surprised to find this wavelength was available in inexpensive strips and 2 density layouts. I opted to get the densest at 120/meter rather than 60/meter because more = better, right?
I ended up using 11 34" strips mounted on a piece of coroplast that was then stapled to a box made of 1"x4"s. It looks pretty janky and like a fire hazard which is the reason it's getting rebuilt.
The power supply is 120v AC in and has two 12v DC outputs that are rated at 10 amps. The time controller was purchased on eBay and pretty much interrupts the 12V DC circuit when the time is up, shutting off the LEDs. The switches on the laser cut control panel turn all the power on, start the timer and turn on safe lights (yellow LEDs). I intended to purchase glass at some point but it was never needed because I use a space saver vacuum bag for exposure. For the new build I'll probably end up adding glass and a proper vacuum lid in phase 2.
I plan on reusing the time controller/switches and the power supply but not the LED panel or box. This time around I ordered 4200 LEDs because I want the strips to be much closer together like the M&R Starlight/Workhorse Lumitron. The strips will be soldered this time instead of using the crappy snap connectors and will be done in parallel to avoid differences in brightness. If they're wired in series the end strip has voltage drop and the whole exposure would be uneven.
So now for the before pictures:
Wiring Diagram
(https://i.imgur.com/L5cpW3gl.jpg)
Turned on
(https://i.imgur.com/MuRwEysl.jpg)
Janky Wiring
(https://i.imgur.com/2XjXnB1l.jpg)
Control Panel Mock Up
(https://i.imgur.com/HS44u0ul.jpg)
Control Panel Finished
(https://i.imgur.com/ccHXNGTl.jpg)
First Tests (WBP Emulsion)
(https://i.imgur.com/TvyDjItl.jpg)
Switched to Cryocoat with 18 second exposure
(https://i.imgur.com/oWd5TiMl.jpg)
Final exposure time using a Stouffer 21 step is ~80 seconds. I'm curious (and I'm sure someone here can answer) when companies are advertising their exposure units can expose in 10 seconds, are they using the 21 step guide?
Anyway, I'll update this thread when I start the rebuild and make sure to add any updated documentation (wiring diagram). Until then, ask any questions if you have them!
UPDATE:
Redrew the schematic during my morning coffee
(https://i.imgur.com/7ke9jYHl.jpg)
-
Thanks for sharing. I built one similar to to this a couple of years ago. I need to go back & solder the connections - I do have a few places where brightness isn't consistent. The clips were a simple way to get the thing put together, but aren't great quality.
-
Looks good. I went down the same path a few years ago. Final results were ok, but finding a reliable quality LED source was the variable I ran into. The M&R LED's are not the same as the ones we can get off AliExpress or from Amazon. I built mine as an experiment, but it also can serve as a back up if my 5k Olec decides to die on me. I hope as technology progresses we might find a single point source LED in the 1000W range that is reasonably priced with enough wattage to REALLY replace a 5k traditional source.
-
Just finished putting together my updated version:
(https://i.imgur.com/5kfKiE8l.jpg)
(https://i.imgur.com/EjHh1wvl.jpg)
There are a few dead LEDs because I had to pull them from my old unit due to a math error on my part. Soldering took about 3 hours so I'll probably post a proper update tomorrow.
-
try to eliminate the gap between the sections. It causes the light field intensity drop in that area and potentially weaker exposure.
Also, see if you can get a current regulating power supply. This will stabilize your light output and prevent overheating.
there's more to think about, but all together looks like a pretty good start. Looking at the exposed screen images, they look pretty good!
pierre
-
Based on the last image, isnt the glass pretty far from the bulbs? Arent most of the LED units closer to like an inch or two from the glass vs 4 to 6 like in the image?
-
Based on the last image, isnt the glass pretty far from the bulbs? Arent most of the LED units closer to like an inch or two from the glass vs 4 to 6 like in the image?
glass distance has to be calculated based on the cone angle. It should be set up so there is a correct overlap from each LED without undercutting from the side bulbs. This is part of the reason why some of the manufacturers are just making really crappy units. They have field intensity variation of over 20% (in one case almost 40%!!!). Imagine having a stencil with weak spots throughout the image from being underexposed by 40%!
pierre
-
I know it requires calculation, but I thought one of the major benefits of using these LED strips is the dense clustering of the LED meant the sweet spot was significantly closer than something like a flo bulb unit. I cant find any useful info on the angle of the cone from these kinds of strips, but I would assume its somewhere around 20 to 30 degrees, which should result in something between 2 and 3 inches away max. All speculation of course without actually testing.
-
I know it requires calculation, but I thought one of the major benefits of using these LED strips is the dense clustering of the LED meant the sweet spot was significantly closer than something like a flo bulb unit. I cant find any useful info on the angle of the cone from these kinds of strips, but I would assume its somewhere around 20 to 30 degrees, which should result in something between 2 and 3 inches away max. All speculation of course without actually testing.
I’m guessing you might be able to get something as low as 45, but almost all are around 120. That makes for a lot of undercutting if the distance is too far.
Pierre
-
if the angle is 120, that would mean the ideal distance sits closer to an inch away max unless im missing something
-
if the angle is 120, that would mean the ideal distance sits closer to an inch away max unless im missing something
it would depend on the LED spacing. The further apart they are, the further the glass has to be . . . Vastex unit still has the best field I've measured (data is kinda old now though) and their unit is about 6" away.
pierre
-
This post caused me to look a little closer at my exposure unit. I'm using a DIY LED that is about 300 watts of 5050-SMD diodes (1500 total)... that are mounted 1.75" below the glass. The spacing between diodes and rows of diodes is 16mm (5/8") and covers about 18" X 32" in total area, so it's a pretty tight pattern. I don't know what angle the diodes broadcast at... but I think the 1.75" distance between the diode and glass is a good compromise between even exposure, without too much undercutting.
The nature of my work is predominantly simple spot color stuff... so I've never really investigated it's maximum capabilities.
For a long time I burned S-mesh screens at 15 seconds.... but changed that to 45 seconds to increase durability (I catalog a lot of my screens for re-use and the removal of tape was pulling emulsion from the edges of the screens). The longer (45 second) exposure time fixed that problem.
I've just burned a 180S, using a Chromaline test film, and I would say the dots between 10%-90% we're really good... and a 6 point font looks good too.
The 90% to 100% range looks the same (fully Open) and the 0% to 10% range looks the same (fully Closed).
Maybe I could squeeze a couple percent better by adjusting the exposure time, but that's a capability that I haven't needed, yet.
So.... I'm guessing that a decent DIY LED will require about 75 watts of power per square foot.
And, the spacing between diode rows should be the same as the spacing between the diodes on the strip.
The quality of LED strip is questionable.... I needed 15 meters of strip to get 10 meters of uniform product, but it's cheap so that isn't a deal-breaker.
-
Based on the last image, isnt the glass pretty far from the bulbs? Arent most of the LED units closer to like an inch or two from the glass vs 4 to 6 like in the image?
Yes, I had the panel about 4 inches from where the screen rests for exposure. I've never seen an LED exposure unit other than photos and those weren't too helpful in figuring out the proper distance. I exposed a new test screen this morning and raised the panel up 1.5" higher to see what would happen. The Stouffer test came out at a pretty solid 9 with 25 seconds exposure on a 305 with Cryocoat emulsion so I could probably cut it in half and hit a decent 7.
Here are some photos:
(https://i.imgur.com/fCSH4TOl.jpg)
(https://i.imgur.com/KFCAbvLl.jpg)
-
going to 7 from a 9 requires doubling the time rather than cutting it in half. Or am I missing something?
pierre
-
Other way around. I'm aiming for a 7 but I'm at a 9 (which is slightly over exposed).
-
Other way around. I'm aiming for a 7 and I'm doing better than that because I'm at a 9 (which is slightly over exposed).
I must be missing something. On a Stouffer scale, 7 is exposed longer than 9. If you don't do anything, all the numbers will wash out (that would be 21). With a 10 second exposure, you should get something like a 12 (based on your numbers). Going to 50 seconds, would reduce the number and leave the 7 intact.
Can you post a picture of the strip and what it looks like on the screen after exposure?
pierre
-
Absolutely, check out the photos below. I lost the sleeve for the strip so I may be totally wrong and misunderstanding the guide from the website:
http://www.stouffer.net/using21step.htm (http://www.stouffer.net/using21step.htm)
(https://i.imgur.com/QmFLvQ1l.jpg)
(https://i.imgur.com/wTcFcG1l.jpg)
I should probably get a new strip soon too ;D
-
DUH!!! You are right! I was thinking backwards. Glad I asked for a photo, it cleared it instantly. The picture is worth a thousand words!
pierre
-
DUH!!! You are right! I was thinking backwards. Glad I asked for a photo, it cleared it instantly. The picture is worth a thousand words!
pierre
Whew, I thought I was going nuts haha. Thanks for double checking my work!
-
At the risk of resurrecting a dead post I did have a question for you on this.
Do you have safe yellow LED's running in between the groups of UV LED's? It kind of looked like you had something running down those gaps in one of the pictures.
I am working on my own design and am considering using yellow LED's around the perimeter of the light chamber.
Thanks
Jon
-
I do but I never hooked them up as I don't really have a use for them. I have a little jig I built to line up positives on the screen, it's not fancy but it only cost $30.
-
I'm in the process of rebuilding my LED exposure unit and figured I would document it here.
Through looking at some tech sheets it looked like most emulsion crosslinks at about 395 nm wavelength. I was surprised to find this wavelength was available in inexpensive strips and 2 density layouts. I opted to get the densest at 120/meter rather than 60/meter because more = better, right?
I ended up using 11 34" strips mounted on a piece of coroplast that was then stapled to a box made of 1"x4"s. It looks pretty janky and like a fire hazard which is the reason it's getting rebuilt.
The power supply is 120v AC in and has two 12v DC outputs that are rated at 10 amps. The time controller was purchased on eBay and pretty much interrupts the 12V DC circuit when the time is up, shutting off the LEDs. The switches on the laser cut control panel turn all the power on, start the timer and turn on safe lights (yellow LEDs). I intended to purchase glass at some point but it was never needed because I use a space saver vacuum bag for exposure. For the new build I'll probably end up adding glass and a proper vacuum lid in phase 2.
I plan on reusing the time controller/switches and the power supply but not the LED panel or box. This time around I ordered 4200 LEDs because I want the strips to be much closer together like the M&R Starlight/Workhorse Lumitron. The strips will be soldered this time instead of using the crappy snap connectors and will be done in parallel to avoid differences in brightness. If they're wired in series the end strip has voltage drop and the whole exposure would be uneven.
So now for the before pictures:
Wiring Diagram
(https://i.imgur.com/L5cpW3gl.jpg)
Turned on
(https://i.imgur.com/MuRwEysl.jpg)
Janky Wiring
(https://i.imgur.com/2XjXnB1l.jpg)
Control Panel Mock Up
(https://i.imgur.com/HS44u0ul.jpg)
Control Panel Finished
(https://i.imgur.com/ccHXNGTl.jpg)
First Tests (WBP Emulsion)
(https://i.imgur.com/TvyDjItl.jpg)
Switched to Cryocoat with 18 second exposure
(https://i.imgur.com/oWd5TiMl.jpg)
Final exposure time using a Stouffer 21 step is ~80 seconds. I'm curious (and I'm sure someone here can answer) when companies are advertising their exposure units can expose in 10 seconds, are they using the 21 step guide?
Anyway, I'll update this thread when I start the rebuild and make sure to add any updated documentation (wiring diagram). Until then, ask any questions if you have them!
UPDATE:
Redrew the schematic during my morning coffee
(https://i.imgur.com/7ke9jYHl.jpg)
Hi! I'm wondering if you used 5050 uv leds. I will try these and I'd like to know if they'll work fine:
Specification
Input volts: DC 12V (12V 2-3 A for 5meters 3528 300LED light strip)
Wavelength:395-405nm
Long life span 50,000+ hours
Protection Rank: iP 65waterproof
PCB Color: White
Is Dimmable: Yes(dimmer required)
Is Trimmable: Yes(cuttable every three LEDs)
Drive Mode:Contant Voltage
Beam Angle:120°
Viewing Angle: adjustable by mounting position
Working Tempreture:-20° to 50°
Size: W1.0cm x T0.25c
Did you tried the exposure unit on diazo?
Thanks!
-
I used 3528s that are spaced about 3 per inch, the densest I could find. WBP is a diazo emulsion and it worked just fine with longer exposure times (45 seconds-1 minute if I remember correctly). I've since switched to photopolymer and expose at 20-25 seconds but may switch back again to compare details.
-
Fiodor,
Thanks for posting the pics. I think you'll be much happier with the leds closer together. The pic you posted of the lights spaced out like that looks a lot like the lawson unit I bought "I've since went back to metal halide". The lawson was fine for the majority of the work I do but I wasn't happy with the halftone exposures. I know the starlight has roughly 10 times the leds which eliminates the cold spots.
Another thought: I admire you guys that have the knowledge to even attempt building your own because I wouldn't know where to begin.
-
Hey, Doug, how are you? The work is from Pangea, not mine, so props to him. I will be starting mine next week.
The widht of the strip i'll be working with is about 1 cm, so I'll put one strip (1 cm. empty space) another strip, etc. until i cover 1 mt. lenght.
As I measured, the space between every led light is about 0.65 mm. and the distance between the glass and the lights will be about an inch. What do you think?
-
Hey, Doug, how are you? The work is from Pangea, not mine, so props to him. I will be starting mine next week.
The widht of the strip i'll be working with is about 1 cm, so I'll put one strip (1 cm. empty space) another strip, etc. until i cover 1 mt. lenght.
As I measured, the space between every led light is about 0.65 mm. and the distance between the glass and the lights will be about an inch. What do you think?
if the distance between LEDs is 0.65, the distance between the strips should be the same.
the distance to the glass will depend on the angles of the light the LEDs are producing.
You will also have problems with heat if you are only 1" away. Heat will kill your light source and it will also degrade the light output much faster.This in turn means you will have to adjust your exposure times often.
pierre
-
Hi, Pierre, I was considering what you said before on the thread. When you say " angles of the light the LEDs are producing"
you refer to the beam angle of light? the manufacturer specifies "120º beam angle", I don't know if this is useful information.
Is there an accurate formula to calculate the distance between the lights and the glass?
Thanks a lot for the advice, is very helpful for the beginners!
-
Hi, Pierre, I was considering what you said before on the thread. When you say " angles of the light the LEDs are producing"
you refer to the beam angle of light? the manufacturer specifies "120º beam angle", I don't know if this is useful information.
Is there an accurate formula to calculate the distance between the lights and the glass?
Thanks a lot for the advice, is very helpful for the beginners!
yes, the beam angle. 120 seems to be the most popular. . .
yes, there is a way to calculate the spacing, unfortunately I don't have that formula. Easy way would be to get a UV meter and measure the field than keep tweaking until you get uniform readings.
pierre
-
I know it would increase the exposure times, but wouldn't a diffuser help with the hot and cold spots?
It'll be a tough sell to get me away from my 7500k Amergraph. I do screens in about 15 seconds...2 up.
-
I know it would increase the exposure times, but wouldn't a diffuser help with the hot and cold spots?
It'll be a tough sell to get me away from my 7500k Amergraph. I do screens in about 15 seconds...2 up.
You might consider the power draw with the Amergraph. We just got the Saati ProLite 450, to replace a Violux 5000S (5K) and so far exposure times are similar to the Violux, but we still have considerable testing to do. Currently, we use Saati HU42 and Ulano Orange, and we're about to text Murakami T9... The ProLite is closer to a point source than a bed of LED's, but not as small as a MH bulb.
Steve
-
The theory of a point light source seems to have been thrown out the window.
Not only are people using banks of LEDs but also have the screen moving over the LEDs.
I would like to know if anyone has tested exposure with a bank of LEDs opposed to a MH.
I’d also like to see a test of static leads and LEDs with a moving screen.
-
Hello Maxie, MH was never broken, and LED didn't fix it, LED is just an alternative system to an already great light source in MH. MH does have some quirks that LED did help out, like lamp costs, electrical draw (which I can argue the savings are lost on your first breakdown), but are they equal in terms of exposure quality? We have done head to head tests to help develop new emulsions for both. LED=fast and inexpensive to operate and often the best emulsion is a diazo or pure photopolymer emulsion while dual cure tells another story. MH = the standard by which resolution and stencil strength are judged by can expose diazo, pure photopolymers, and dual cure emulsions (perfectly). Science tells the story. Histograms (Spectral Output), amplitude of light, how parallel are the light rays?, all tell different stories when compared. A lot of these annoying science facts make a difference in durability and resolution. However the two systems can work together beautifully. One takes care of the majority of work, but the other fixes what the first can't do. I would have both, I like LED for speed, I like MH for resolution and stencil durability, especially on thick stencils. I work with many shops that utilize both constantly. MH serves as a good back up to the sun on a rainy day (probably not an issue for you!) when post exposure after LED may be necessary for a high solid acrylic base destined for a long run. In any event, you can't go wrong with either, some issues that may not be important to one shop really shows up in another. Some print mostly spot, some 85 line concert shirts. To each his own, nothing wrong with either system if you like the prints and your customers are stoked. Just print.
-
Alan, thanks for the reply. I'd guess that most people on TSB are going to have one system only. In theory there is no doubt that a point light source is better. How much does this effect us in practice? As you mention, for most the LED will work fine.
I think a good compromise is something like the Saati lamp mentioned, it's closer to a point light source and LED.
I'd like to see someone post a test chart exposed with LED and MH? Maybe you can post one of your tests.
-
to throw my 2 cents in, the point source will make better halftones, I went from Carbon Arc (really the best, but poisonous) to Metal Halide, and we are now using the Saati, still testing exposure times. We don't do very much discharge or water base, so haven't had to consider the post exposure yet. However, as Alan mentions, we're keeping the MH for just that purpose when needed, or if the LED fails. I'm also making (as a pure hobby, not as a need) lamp using a 200w UV LED, I have all the parts, except for a case, which seems to be the hard part for my needed dimensions. If I get that to work out, it could become my backup, maybe redesigned a bit, say, with 2 200w, or 4 100w. I personally have never warmed to the bed of lights idea.
Steve
-
I dont see any examples of what LED stripes you used. Can you share a link?
I have seen problems after 1 year with LED's that had weatherproofing added. They are duller as have yellowed.
What spectrum did you use? I have seen 365nm which I think is to low. I think the optimal is 390-405nm.
-
Got them on Amazon, the densest configuration they had (around 3/in if I remember correctly). No need for weatherproofed type and 395-405nm is optimal for based on emulsion tech sheets I read.