Decided to take it up....Making an LED Exposure unit.

[Steve Harpold]

Hey,

Out of respect I will remove the post  and apologize for the inconvenience.

I will leave the below as a follow up (less shameless)

In the thread many questions were asked on why a machine wasn't built a certain way or why current manufacturers don't make things science based. A light analysis was done and posted about current machines in form that doesn't shed a good light on current manufactures. You mentioned not wanting to build a "copy" of something that exists @ price point that was more economical. The details and the light analysis I laid out was something that already exists within the industry. A science based benchmark, take it a challenge to see if you can make it better. Good luck on you endeavor I will be excited to see how far you can push limitations of LED exposure units.

Thanks!

[SoccerMom]

  FWIW,   I saw it as helpful usefull information that now is edited and gone for eternity.... Guess your on your own now Pro.

[IntegrityShirts]

Yeah I didn't take it that way, Steve. I saw it as an industry professional HELPING out the small time guy with information he may not already have on-hand. Especially taking in to consideration that Pro hasn't seen ANY of the industry's production LED units in person. Best of luck and I will be following this as far as it goes!

[Frog]

I also didn't see it as a hijack especially since it was in a discussion thread in the DIY section, rather than in a particular manufacturer's promotional thread.
You may remember a recent similar occurrence with a squeegee handle post that due to forum "newbieness", actually was a hijack, and was split to it's own thread.

[jvanick]

  FWIW,   I saw it as helpful usefull information that now is edited and gone for eternity.... Guess your on your own now Pro.

I'll summarize:

405nm (which is the accepted light range for most thorough penetration of the emulsion, frying vs. baking).
LED's are 4" from the screen.  (not sure this matters as much) -- however it does play into more consistent light, and speed.  -- Inverse Square rule applies here as the the light strength is inversely proportional to the square of the distance from the source.

I like a slower exposure as it makes for more latitude.  Our starlight was literally exposing some emulsions at 3 seconds!... 2 was undercured, 4 was over...

[Prōdigium]

Hey,

Out of respect I will remove the post  and apologize for the inconvenience.

I will leave the below as a follow up (less shameless)

In the thread many questions were asked on why a machine wasn't built a certain way or why current manufacturers don't make things science based. A light analysis was done and posted about current machines in form that doesn't shed a good light on current manufactures. You mentioned not wanting to build a "copy" of something that exists @ price point that was more economical. The details and the light analysis I laid out was something that already exists within the industry. A science based benchmark, take it a challenge to see if you can make it better. Good luck on you endeavor I will be excited to see how far you can push limitations of LED exposure units.

Thanks!

Steve, as a person who long ago built a business using forums to promote my business (sometimes a bit shamelessly as well), there is a way that you could of discussed the topics without turning it into a shameless plug for your product. Your post was only missing some fancy graphics and PDF brochure to qualify as a full on advertisement. You spent all of 2-3 sentences talking about the topic, then the advert...surely you can understand why I replied to it the way I have.

The details I laid out certainly make a case for what I will offer, and a about why I think its better than some current designs..but lets please note that this is the proper forum for me to do it. I am not a mfg company (yet) and even though I have been granted access to resources that are typically reserved for big companies that does not make it any less of a DIY project.

The questions asked to me, were not why "a" machine was not built to copy others, it was asked why "I" chose a certain design path. The light analysis you provided was merely part of your sales pitch, complete with a pat on the back for your design staff.  It was not my intentions to come across so harshly, and really I am not trying to make an enemy here, but your post was a shameless advert and I called you out on it.

I hope as well that my research and design has pushed the technology (at least a little bit further) but in reality it was not my intention to build the "best" mouse trap or even push the limitations as you stated... I just want to offer professional performance on an affordable level and prove that a decent exposure unit does not need to cost so much. from which the initial idea came from a prior thread on the same matter that you may or may not have read about?

[Steve Harpold]

Hey Pro,

I am excited to see research in action and believe that there is a place in the market for everyone. I read this forum and as many others as I can daily to keep pulse on the industry as well as watch for ideas, or reoccuring themes.  I will rarely chime in as this website is a community that is moderated well and provides great discussion amongst printers. I only jump in when manufactures are being put out there in a negative light (For unfair/unlikely reasons)  While you were promoting your concept and your way of thinking, you routinely mentioned manufacturers  by name and pointed out what they were all doing wrong. Including posting a model of light analysis done without and actual exposure unit just an assumption made for reading their literature. The light analysis i provided is an actual analysis of a current unit in the field. (Side note: I ran Pierre's idea through light simulation software in Atlantic City roughly 2 years ago, it has promise but as all ideas it would need to be perfected)

I play on a team, the patting on the back is for my team members as i expect them to follow the forums as well. I also believe when they have created something that is truly exceptional they deserve to publicly congratulated (I enjoy reading stories on here about a employee(s) who have done something that makes a difference in shop whether a print shop, a manufacturing facility or a supplier.)

Current units in the field range from:
$2,500 - $15,000+

If you call out a company that is retailing @ $5,000 and than remove all the features (important to some or not) and and build it for less you are no longer building the same unit and discrediting there efforts (Phone Apps, Light Integrator chips, lensing research.....)

If you are building a basic exposure unit @ a more economical price than your goal is to compare to those that are already at economical price.  The unit and features are on a unit that retails @ $2,795. I believe there is another commercial unit @ $2,500 as well.

If you simply promote your ideas and plans, I will stay silent and wish the best of luck and continue to push my team to make the best available unit in each price category. If you call in question to efforts of other manufactures and downplay there inventions than I will chime in by posting the science behind some of the efforts and inventions.

As always best of luck, i am firm believer that competition brings out the best in all people. It can get a little chippy at times but that simply drives people to try harder!

thanks
 

[SoccerMom]

Your patents, analysis reports, and sequential manufacturing capabilities, have more than likely pushed your project beyond the scope of the typical DIY (*IMHO).... A build to me is born of love and merely a creative  outlet utilizing multiple designs, all done late in the night as the world sleeps . Its not to save money, be cheap or competitive, but as a marriage of ideas and concepts. Buying materials and equipment is quite expensive, fabricating and welding high carbon steel / aluminum is a daunting task that few have the patience or skill for, not to mention the safety aspect in terms of engineering and electrical....  One should almost ask themselves, why bother?, I could just go buy one.  

[Prōdigium]

Steve,
Its all good and this is more along the lines of a beneficial discussion. To be fair I was merely noting that my approach to design was meant to resolve some inherent design flaws that in a $5000+ exposure unit should NOT exist, such as warped or distorted UV strips. It would be like having a Mercedes's C class with faux leather seats. Its just something that should not be done in that price class....at least from my perspective.

To be fair, I did state clearly that my tests were done on the layout design, NOT on the actual units. Fully admitting that either mfg could of corrected this flaw with special lenses, but that I doubted that this was the case based on evidence provided to me. If I am incorrect, than either mfg named is more than welcome to step up to the plate and set me straight. I will edit or correct my post with an apology as well. I recreated in a testing environment the layout and was provided the image by the lab who performed the tests. I did not mention your product at all, so it was never meant as a comparison. But of course you can easily create a uniform light coverage by simply lowering the lights until you get even distribution, as you have done but that can have an impact on performance. What is an average exposure time on your unit with a dual cure?

However now that current LED designs are in the near flash of any eye, some screens burning in 5 seconds (or less) the performance measurement really comes down to measuring screen quality. Again, I stated that I am merely at the stage where I am beginning to validate my design and that until its tested in the real world.."it might suck" Also like you, I do not think its so great to have near instant exposure as your latitude is so small that 1.5 seconds may be the difference in proper exposure....that would be for most, an annoyance to say the least.

As for removing features, that is not exactly the case. In some models for example, having the lights turn on when the vacuum blanket has drawn is called a "feature"..in what I am building its just how it works and its not an option, so its a fair comparison to make and I did. Its a $2.00 sensor switch, so there really is no valid reason why it should not be standard on any unit over $2500. AN LCD screen to some may be called a feature, I call it an accident waiting to happen. Anyone who has cracked a cell phone can attest to just how easy it is to break and when your handling a roller frame its bound to get broken sooner or later and I am positive that the cost for replacement will not seem like a feature when you cant burn a screen. Sometimes a good feature is just plain simplicity, and reliability. Like being able to replace the LED without a service tech or the bill that comes with it, again its a fair comparison to make.

I cannot compare some of the other base units to what I am building as they lack the performance (I hope to achieve in my initial design) and in most cases ANY features so it would be foolish to compare my design to say a Lumitron unit, which even falsely claims as a feature that the "light never degrades" This is factually incorrect as told to me by the UV LED mfg directly as we discussed the life span and luminous output. UV LED's do degrade over time, especially in the ranges below 390nm which degrade at nearly double the rate of 400nm wavelengths, with 365nm typically being rated for a mere 8000 hours due to the ingredients used to make them, regardless of the company that makes them.

My talking points were not to compare what I am trying to accomplish simply based on price, but on performance and serviceability as well. Its like comparing the performance of an American muscle car such as a Mustang Shelby GT350 to say a Ferrari. They can be compared in terms of performance, but they are in wildly different price ranges. In many circumstances the Mustang will trash the Ferrari and where the Mustang really wins is being able to afford to still buy your groceries and drive them home in it as well. I am just making the same type of arguments.

[Prōdigium]

Your patents, analysis reports, and sequential manufacturing capabilities, have more than likely pushed your project beyond the scope of the typical DIY (*IMHO).... A build to me is born of love and merely a creative  outlet utilizing multiple design concepts, all done late in the night as the world sleeps . Its not to save money, be cheap or competitive, but as a marriage of ideas and concepts. Buying materials and equipment is quite expensive, fabricating and welding high carbon steel / aluminum is a daunting task that few have the patience or skill for, not to mention the safety aspect in terms of engineering and electrical....  One should almost ask themselves, why bother?, I could just go buy one.

Agreed on so many levels. I have built many projects not because it would save anything, I just like to tinker with stuff and sometimes a better product is the result even if its not the intended goal. After the thread about me complaining of the costs, I just purchased a bunch of stuff and started to play with it not serious about building a unit to sell. But I was astonished at how easy it was to achieve fairly decent results for pocket change, so I decided to see what I could do with the plan to keep the costs inside a specified range. The end result is such that I would be foolish NOT to build it, even if I can only sell 10 units a year.

Although the patent I seek actually has nothing to do with the LED exposure system itself. There may be something in my design approach that is "patent-able" but its not something I will spend any efforts to protect. In fact I am of the Open Source type, but the reality is that sometimes you stumble as I did across a really good idea when trying to resolve a problem. So that is why I am getting a patent on the idea...but I will gladly sell the idea or license it to anyone once its protected, well all need to make a living.

Also to be clear, I do not think I have anything that could be called "sequential manufacturing capabilities" The parts will be made by various small companies here, many by people I know from my pizzeria here and I will hand build each one myself in my little 150 square meter space until such time as it becomes impractical. Help with board design is by a friend here whose company builds radiant gas heaters. The LED companies are of course wanting the business and its normal to provide testing, especially when developing custom made LED's. After all they need to do in house tests to ensure they are designed to specs as well so many of the tests were ones they would of done anyway.

My initial developments were in fact all DIY...but when you plan to mfg something you bring in professionals to make sure your not overlooking something and ensuring that its done in a cost savings manner. So as Frog mentioned in the beginning, it may not qualify as a true DIY project..but until they are being sold as a viable product its just me, my ideas and my one off test product at this stage...and its certainly a labor of love working late at night on ideas. When I go into production, then I will post it under a more correct section.

[Steve Harpold]

6 months ago:
Diazo exposure time 30-40 seconds
Now:
Diazo exposure time 20-30 seconds

In order to expose photo polymers use a variable power supply to run the LED's 50% power (Increase life for all those using Photo Polymers as well as control exposure times)
Photo Polymer Time without variable power supply: 1 - 5 second
With Variable power supply: 15-30 seconds (Depending on power supply out put)

Degradation of LED's are exponentially increased by excessive heat, which is why you keep them as far away for the glass as possible.  The heat has somewhere to go and glass does not act as a insulator increasing the life of the LED's.  LED's also degrade faster if ran @ 100% capacity.

Optimal: size the LED's 25% greater than what is necessary. Use a light integration chip to drive them harder as they degrade.

Like posted in the Vega video we have unlensed LED chips that produce up 8w/cm^2. (Two expensive for this industry and complete overkill, but will give you and idea of how it is possible to use 28 LED's and achieve the above stats)

Cool LED things to look @ With lensing/dibonding a few companies have reached heights of 24w/cm^2  (Different industry but cool to watch)


As far as the Ferrari Vs The Shelby, it is something we fight in every industry sometimes perception is better than  performance.

[Prōdigium]

Steve,

All good information...thanks.

I tested higher power LED's, a couple were actually rated at over 100 watts as one idea was to develop a single point style, the problems were that UV LED's in that range are stupid expensive and degrade too quickly not to mention heat issues. Ironically with my current design, heat is NOT even a factor to trouble with. The bulbs so far have never got hot enough that its even hot to the touch. This is most likely due to my LED board which is in fact a copper clad aluminum heat sink. Maybe we have a different idea about "affordable" LED's, but even the more powerful ones 10-Watt Cree units I purchased were under $2 each in bulk. I just found that its too much power, and heat. More LED's of lower power resolved all such problems.

There are no current plans to use a "dimmer" for different emulsions, at least not in the typical analog sense. As stated I have found that more power is not always better. In testing my ideas, uniformity of light won out over sheer power, unless maybe you need to burn a 400 micron Film, but still that should not be a big problem? Of course testing will tell. That said my power supply is rather unique for my unit (I think) in that it will not provide constant line current, and that will help keep temps low. However my plan is to operate at 100% of the rated LED power.

I am not troubled at all by LED lifespans. Even at 10,000 hours your talking about decades of life @ 30-60 seconds per screen, so to spend much time worrying about it is nonsense. The company that is mfg my LED's states that the spectrum does not alter very much with age, but wattage does drop over time. So if your using an integrator I hope it measures wattage and spectrum.

[Steve Harpold]

The above I posted about the 8w/cm^2 example is a poor example as it does not apply to this industry or exposure units. Just a cool way to see how far LED's have come.

The example you posted of a 100 watts or 10 cree units (roughly 70 watts) is missing a area to compare to the latest technology.

Assume the following 100 watt led chip, that is going to illuminate an image area of 16" x 16".  256 in^2 or roughly 1,651 cm^2.

100 watts/1,651 cm^2 = .06 watt/cm^2 delivered to the area.

You can a buy a grow light off the shelf that is 300 watts over the same image area 16x16 for $375, nearly 3 times the output. Ironically the grow lights are in the 410-420nm, which does work on many emulsions.

The 8w/cm^2 I was referring to is
.06w/8w = 133 times the amount power the 100 watt chip would deliver over the same 16x16 area.

Edit: There is a distance not being accounted for in the above formulas, when I get second I will go back and recalculate. (Since I didn't include a distance in either one the comparison is accurate but the concept impossible without an array)

Edit (2): traveling again had time to fix the equation:
Power is reduced by a square of the distance
Above example of you had a single chip that was 100 watts at the chip like he above example the further away the chip got from the source the less power it would deliver by a square of the distance:

Ie. Let say I took a chip that was 100 watts and put it 16" away to achieve uniform light, the chip would have to be 4 times as powerful as the chip that was pressed against the glass to deliver the same amount of watts/cm^2.


What is cool about this:
We are looking for the same thing how many watts/cm^2 does it take to expose a screen @ 405 NM wave length.  It is irrelevant how you deliver those watts (852 chips or 1 chip the science equation remains the same) now the questions is which one delivers longevity, ease of use and a unified light pattern. Let the best design win!!!

Side note: the answer to above is very different when considering both photo polymer vs. Diazo's.

[ZooCity]

6 months ago:
Diazo exposure time 30-40 seconds
Now:
Diazo exposure time 20-30 seconds

In order to expose photo polymers use a variable power supply to run the LED's 50% power (Increase life for all those using Photo Polymers as well as control exposure times)

Degradation of LED's are exponentially increased by excessive heat, which is why you keep them as far away for the glass as possible.  The heat has somewhere to go and glass does not act as a insulator increasing the life of the LED's.  LED's also degrade faster if ran @ 100% capacity.

Optimal: size the LED's 25% greater than what is necessary. Use a light integration chip to drive them harder as they degrade.

Smart.  0.5s exposure times are not beneficial in any way. 

I'm also impressed that you are addressing the very real fact the LEDs will degrade over time.  I didn't know about the heat expediting the degradation.  We have been running our unit with glass and film and it's over 100˚F on the blanket during production.  One has to wonder what that's doing to the array.  I'm pretty sure ours has become dished/warped from the heat, the glass certainly has warped significantly from the heat forcing us to reset our "flat" on our roller table, screens and presses.   I'll be confirming if the array mount is flat this weekend as we're removing the glass in prep for a cts install.  I think an LED expo with variable power and integrator is brilliant if you'll excuse the pun. 

Love threads like this and interested to see what you come up with on your project Prodigium.  Good luck. 

[Steve Harpold]

Hey Zoo,

We have a couple units in the field with temperature sensors on the boards to measure the max operating temp. It is unknown what temperature will ultimately cause the boards to fail and what the life span is @ different temperature values. Currently we have the machines set to alarm around 40 c or 104 F.  The best formula we can put together is that 50 c or 122 f will not harm the boards or shorten the life span, but in complete honesty I can only stand behind 2 years of data. It is a pretty safe assumption that these numbers are congruent across all LED exposure unit lines.