LED UV Exposure Box

[Spork Schivago]

Sorry if this is in the wrong section. I'm interested in UV reactive PCBs and making a UV Exposure box. Essentially, you print your boards traces / pads on a transparent / semi-transparent paper, put it over a UV reactive PCB (I think they're called Photoresist PCBs) and then shine UV light on them for a certain amount of time.

I'm interested in hearing from people who have experience with this. I'm working on building a unit but I'm not 100% sure on the type of LED to use. I do want to use LEDs as well. I'm pretty stuck on that idea. I do NOT want to use a bulb.

The question is what's the different outcomes from the different UV LED's wavelengths? On e-Bay, I can purchase 100 UV LEDs with a wavelength of 390 - 395nm for 3.30$. Or I can purchase 100 with a wavelength of 365nm for 49.50$. I've seen how-to's that say to buy ones that have wavelengths around 400nm - 405nm. I've seen howto's that say you need the 365nm's. I'm assuming the more expensive shorter wavelengths mean less exposure time?

Also, to get a nice even exposure, I planned on building a diffusion filter. I've seen some UV LEDs with wide viewing angles (the ones I'm thinking of purchasing have only 20 degrees) but some out there have 120 degree! If I went for LEDs with an ultra-wide viewing angle, would I still need the diffusion filter to disperse the UV light and make it all even like on the PCB? Thanks and again, sorry if this is in the wrong spot!

Spork

[stj]

Re: LED UV Exposure Box

UV-A
very dangerous, there was only one western supplier of those and you had to sign a disclaimer to get them.

as for a defuser, you need to use glass, either alkaline-etched or sandblasted.
if you use plastic the UV rays will destroy it.

get datasheets for the led's and we can look more closely at it.
your going to do this much cheaper using UV anti-bacterial tubes though.

[budm]

Re: LED UV Exposure Box

https://en.wikipedia.org/wiki/Photor...ght_wavelength

http://www.eevblog.com/forum/manufac...sist-exposure/

http://www.electricstuff.co.uk/pcbs.html

http://www.technick.net/public/code/...p_dp=guide_pcb

[Spork Schivago]

Re: LED UV Exposure Box

UV-A
very dangerous, there was only one western supplier of those and you had to sign a disclaimer to get them.

Thanks for the reply! Are you sure here? I thought UV-C were dangerous rays, UV-B were "burning" rays and UV-A were the ones that mainly reached the earth from the sun. Tanning rays I think people might call them. My source is from OSH. Do you possible mean UV-C type rays or do you think OSH has it wrong? Either of the wavelengths I picked are in the UV-A range I do believe, the most safe out of all of them. Correc me if I'm wrong here. http://www.ccohs.ca/oshanswers/phys_...radiation.html

as for a defuser, you need to use glass, either alkaline-etched or sandblasted.
if you use plastic the UV rays will destroy it.

I was planning on using clear acrylic that was sandblasted to provide a "frosty" type effect. Will this not work? I was under the impression the acrylic type glass was okay...

get datasheets for the led's and we can look more closely at it.
you're going to do this much cheaper using UV anti-bacterial tubes though.

I'll try finding datasheets for you. I don't know how well that's gonna work though. I'll contact the guy selling them on e-Bay. I believe they're coming from China so they might not ever get back to me. Sometimes the sellers don't really understand English so well. I tried manually searching for the datasheet myself but I don't really have much info to go on. The item shows F5 5mm Round Ultra Violet LED UV Light and I can pick the wavelength. My choices are 365nm, 390nm, 395nm and 405nm. I mainly just wanted to know if I needed to use the more expensive 365nm wavelength LEDs or if I could use the cheaper longer wavelength ones and what the down fall would be. Thanks for answering. Here's a link to the item on e-Bay. I know I shouldn't be purchasing them from e-Bay, but finding two hundred 365nm LEDs for sale on places like digikey just isn't really an option.

http://www.ebay.com/itm/F5-5mm-Round...item33a12f0033

[Spork Schivago]

Re: LED UV Exposure Box

https://en.wikipedia.org/wiki/Photor...ght_wavelength

Thanks for the reply! I finished reading this. I don't want to use the bulbs at all, I really want to stick to LEDs. However, from what I gathered, from that article, it's just the G and I lines from the mercury-vapor lamps that affect the photoresist boards, is that correct? The 436nm and the 365nm wavelengths?

0
http://www.eevblog.com/forum/manufac...sist-exposure/

I've ran into this website before. This is where some of the confusion comes from I guess. One of the articles you link me to further down was an article I was following on how to do this. I read it a few days ago. It says you do NOT want to use black lights that you get at the part stores. However, in this link here, Balaur says, "Those look like "BLB" (i.e. blacklight/night clubs) with wavelength peaks at 350-370nm.

Most pre-coated PCBs have a good response for a 350-450nm spectrum. While "white" UV tubes or LEDs are recommended, your tubes may work."

From what I've read, unless I'm misunderstanding something, I don't need a range of UV wavelengths for these boards. Just one. According to his post, if most pre-coated PCBs have a good response with UV-A wavelengths somewhere between 350nm and 450nm, I'd think the "night club" black lights would work fine, with a peak wavelength of around 350nm - 370nm.


Later on in the forum there on EEVBlog, Carloscuev says they work best with 365nm which is arguably the center of the UVA range.

http://www.electricstuff.co.uk/pcbs.html

I've read this link a couple times now. This was the first source I used a few days ago when I started my research. It's a great article but it's meant for people who are using tubes, not the LEDs. It doesn't really tell me which UV LEDs are the ones I should be purchasing.

0
http://www.technick.net/public/code/...p_dp=guide_pcb

This site kind of says the opposite of "Mike's Electric Stuff (the previous link). It says I can use the disco light tubes or the tubes that I keep calling "party" tubes. "The UV tubes you need are those sold either as replacements for UV exposure units, insect killers or 'black light' tubes for disco lighting etc".

Like I've said though, I really have my heart set on using LEDs and not tubes. From what I've read though, so long as I'm within range I should be good? Maybe the the 365nm works the best, but the other much, much cheaper ones are okay as well? Maybe they just take a little longer to develop? Does that sound about right?

Thank you for sending the links. I read all of them, even if I had already read the previously, I re-read them.

I've seen so much conflicting info, I'm not sure what to purchase at this point! Some say I want the 400nm LEDs, other sources say 365nm, etc.

Thanks for the help everyone!

[goontron]

Re: LED UV Exposure Box

Chuck a mix in there then

[Spork Schivago]

Re: LED UV Exposure Box

Chuck a mix in there then

That could be disasterous Goontron. From what I was reading, the reading I need the diffuser filter is so the UV is dispersed equally throughout the board. Imagine 4 LEDs hooked up in a square like fashion shining on a board. Imagine holding the LEDs close enough to the board so a solid area gets covered by all four LEDs (you'd end up with something like a ven diagram but with 4 circles instead of 3). The areas where the light overlaps the other LEDs won't develop at the same rate because they have more light on them. If I were to use different LEDs or a mix of them, I'd imagine the same would happen. We'd have different brightnesses, different viewing angles, etc. Even with a diffuser filter, I doubt I'd get the traces to develop equally throughout the whole board.

Correct me if I'm wrong in this thinking. Thanks for the idea though.

[eccerr0r]

Re: LED UV Exposure Box

I was playing around with some "UV" LEDs and they turned out to be ~400nm LEDs and that I could easily see the light from the lamps (UV should be invisible, these I wouldn't even call UVA!). These were distinctly blue-purple but were barely sufficient to fluoresce uranium glass just like blacklight fluorescent bulbs. On the other hand these LEDs are insufficient to erase UVEPROMs that require UVC to erase. Someone tried it and it just didn't work.

Currently, as far as I know, it's not cost effective to build UV lighting fixtures with LED as these deep UV LEDs are quite expensive. If you're set upon it I'd think you'd at least need the 365nm LEDs. However I'm not sure if you can get the brightness you need with UV LEDs at a reasonable cost - UV tubes are cheap, even the germicidal UV-C tubes. Wikipedia mentions that these 365nm LEDs have power efficiency issues too, being half as efficient per power than 400nm LEDs.

Blacklight fluorescent: UVA (little visible)
Blacklight incandescent: UVA, "DO NOT USE THESE" due to efficiency
Bug zapper bulbs: UVA and visible <- I'd say these are probably "best" IMHO.
Germicidal: UVC, UVA mixture (and some visible)

-----------------
EDIT:

Actually know what... NO DIFFUSER. If you really need a diffuser, UV needs to be made out of crystalline quartz as even amorphous glass will absorb UV and get warm. Since you will lose light through the diffuser, you'll need even more LEDs to make up for the brightness.

Germicidal bulb tubes are made of quartz for this issue (and is the main reason why these bulbs are more expensive than regular CFL bulbs.)

[Spork Schivago]

Re: LED UV Exposure Box

I was playing around with some "UV" LEDs and they turned out to be ~400nm LEDs and that I could easily see the light from the lamps (UV should be invisible, these I wouldn't even call UVA!). These were distinctly blue-purple but were barely sufficient to fluoresce uranium glass just like blacklight fluorescent bulbs. On the other hand these LEDs are insufficient to erase UVEPROMs that require UVC to erase. Someone tried it and it just didn't work.

Currently, as far as I know, it's not cost effective to build UV lighting fixtures with LED as these deep UV LEDs are quite expensive. If you're set upon it I'd think you'd at least need the 365nm LEDs. However I'm not sure if you can get the brightness you need with UV LEDs at a reasonable cost - UV tubes are cheap, even the germicidal UV-C tubes. Wikipedia mentions that these 365nm LEDs have power efficiency issues too, being half as efficient per power than 400nm LEDs.

Blacklight fluorescent: UVA (little visible)
Blacklight incandescent: UVA, "DO NOT USE THESE" due to efficiency
Bug zapper bulbs: UVA and visible <- I'd say these are probably "best" IMHO.
Germicidal: UVC, UVA mixture (and some visible)

Thank you for the response. I was thinking of using around 70 UV LEDs on each side, maybe 100 (I want to have two modules, so I can do double sided boards). I was either going to do a 10 x 10 grid or a 10 x 7 grid. And yeah, I wrote the China seller on e-Bay asking for a datasheet and asking if these are guaranteed to have a wavelength of around 360nm - 365nm. I can buy 200 of these LEDs for 100$ or so. That's a heck of a lot cheaper than Digikey's 17$ for one! I'm okay with spending 100$ on the LEDs if they're going to do what I want. I guess I just need to wait for his response. I implied that I had a way to actually test the wavelengths to try and encourage the seller not to lie and just sell me 400nm LEDs.

As for being able to see your 400nm UV LEDs, it's not unheard of. I guess people have reported seeing them with even shorter wavelengths. People contributed it to great eyesight. Whether they had crappy UV LEDs or where just lying, I dunno. BTW - Wouldn't Deep UV be something 240nm? I didn't think 365nm was considered Deep UV. Thanks for the input!

[diif]

Re: LED UV Exposure Box

Not UV LED tubes but have you seen this http://m.instructables.com/id/Printe...g-UV-Nail-Cur/

[Spork Schivago]

Re: LED UV Exposure Box

Thanks Diif. I'm not sure if I've seen that one before or not but I do appreciate the link. Mostly what I've read up to this point involved how to make the exposure box, not really how to use it! I originally saw a video on it maybe a year or two back though. Thanks! This is the main tutorial I'm following (it's not so much a tutorial, just more or less a gentleman's journal).

http://pcbheaven.com/projectpages/PC...c=presentation

He's a real nice guy, doesn't have a lot of free time to answer questions though. I believe he's a CNC machinist for a living. Anyway, I'm mainly copying his project. Mine is going to be a bit different but for the basic idea, I'm going to follow his design. If you go there, check out the worklog. He gives ya a bunch of info on how he did, what materials he used, etc. I asked which UV LEDs he used but he hasn't gotten back to me yet. That's where I got the idea for the diffusion filter. He seems to use a piece of 4mm thick acrylic transparent glass, sand blasted to give it a frosty like appearance.

[eccerr0r]

Re: LED UV Exposure Box

I think the 400nm are actually considered "purple" IMHO and most people should be able to see it. The 365nm should be deeper in the UV range and one would expect people should start not being able to see it. Yes 365nm is not "that" deep in the UV region but it should be in the range where it could become a problem (mostly due to invisibility).

Dang I really should look into doing PCB litho this way, I'm sick and tired of these toner transfer methods, it's really tough to get it to "stick"...

[Spork Schivago]

Re: LED UV Exposure Box

I think the 400nm are actually considered "purple" IMHO and most people should be able to see it. The 365nm should be deeper in the UV range and one would expect people should start not being able to see it. Yes 365nm is not "that" deep in the UV region but it should be in the range where it could become a problem (mostly due to invisibility).

Dang I really should look into doing PCB litho this way, I'm sick and tired of these toner transfer methods, it's really tough to get it to "stick"...

Ahhh! I gotcha! Sorry if I came off the wrong way here, I'm trying to learn stuff and I guess I don't have the best communication skills. Sometimes people think I'm being a bit aggressive when I really don't mean to be at all.

So if I ordered these 365nm LEDs and I got them, I shouldn't be able to see anything, right? Only if it was UV reactive, like perhaps the strip in the bills (1's, 10's, 20's, 50's, 100's) where the strips light up under a UV light, right?

I mean with my "party" UV black light, I can see a purplish colour from the bulb while it's on, I can see purple on the bill and I can see the strips hidden inside light up. With the 365nm, I'd only see the strips light up, right? I shouldn't be able to see the LED actually glow and I shouldn't be able to see purple on anything, right? Is the 365nm wavelength dangerous? Should I not look at it without proper filtration?

[stj]

Re: LED UV Exposure Box

i think you need to decide on a pcb supplier and then ask them what the chemical sensitivity frequency is before going any further.

go look at somewhere like MG chemicals.

btw, instead of 100's of leds, how about using an old flatbed scanner.
you can have a single row of led's or a single tube, and just glide it across the board slowly.

[Spork Schivago]

Re: LED UV Exposure Box

i think you need to decide on a pcb supplier and then ask them what the chemical sensitivity frequency is before going any further.

go look at somewhere like MG chemicals.

btw, instead of 100's of leds, how about using an old flatbed scanner.
you can have a single row of led's or a single tube, and just glide it across the board slowly.

I was hearing a lot about MG. I know there's a few ways, you can buy boards already coated or you can coat them yourself. From what I've read, it's really difficult to coat boards yourself and I should buy them pre-coated. If this is still true, I was thinking of something like this to start with: http://www.mouser.com/ProductDetail/...Fc4WHwod5pYKbg

Just a pack of 5 (around 6 x 9 in) boards. I could go for a smaller size, I'm sure. I found out the original guy in the tutorial I was following used 400-405nm, 2000mcd, 5mm LEDs. Reading through the comments, he now suggests going for better LEDs. I wonder if the 365nm ones would mean I'd need less LEDs? Someone said they used 63 1watt UV LEDs but couldn't get the boards developed correctly. The original owner of the article said 1 watt wasn't enough. Turns out the guy meant each LED was 1 watt...

[stj]

Re: LED UV Exposure Box

personally i would like to see a project for a homemade 2D milling machine to proto pcb's
maybe based on an old HP plotter, they used to show up at auction a lot.

i did the acid thing years ago and it's a pain.

[Spork Schivago]

Re: LED UV Exposure Box

I was interested in the homemade milling machine to carve away the copper to make traces and all that for a bit. What finally turned me off from the idea was the whole 20 pins in a 1/4 inch space thing. You know, really small components. I looked into using a dremel for the spindle instead of a router to mill the PCB but from all the research I did, there just wasn't anyway to get fine enough traces for what I was going to be working with.

I'm definitely down for other ideas besides UV if there are any better ones out there that don't cost more than a few hundred to build. I know it can cost a bit in materials for controller boards for stepper motors. For example, just to control two steppers taken out of DVD players via USB, we're talking 50$. That's just for the controller board. That doesn't include belts, the actual stepper motors, clamps, wires, material to make the frame, etc. You can use wood to save some cash when it comes to building the frame, but keep in mind, you don't want anything that can warp. I think you'd need to use MDF board or something similar. I hope if you decide to build one, you keep us posted on your progress and overall results.

[Spork Schivago]

Re: LED UV Exposure Box

Someone says you can use metal halide for these boards. If that's true, maybe I'll just use m metal halide bulb for this. I also have a high pressure sodium and a digital ballast that can handle both. They're smaller bulbs, 400 watts for one, 440 for the other I believer.

[Agent24]

Re: LED UV Exposure Box

Cheap UV LEDs from eBay work fine, they're what I use anyway. I use a sheet of photo-frame glass to hold the artwork flat against the board. Use cheap negative UV resist film I apply with a laminator to blank PCB. Expose for about 15 minutes.

My LEDs are on protoboard in a 3x3cm spaced grid. They could probably be closer but it's what I tried for a first test and it's worked well enough so I've just left it like that despite some darker areas. I have them at a distance of about 18-20cm from the board being exposed.

I develop the board in 2 cups of warm water and a spoonful of washing soda crystals (~40% strength), and brush it with a toothbrush to remove the unexposed resist.

Make sure you double-up your artwork printouts or use a super-high quality professional print shop or such, Toner-Transfer has pinholes but so does toner printed on transparencies! I couldn't get the artwork exposed properly until I did a double layer. Either the traces were underexposed and fell off during developing, or they were OK but the clear areas were full of little dots.

That PulsarProFX Toner-Transfer paper is great though. I haven't tried their TRF (Toner reactive foil) which they say will fill in the pinholes, but if it works as well as their paper does, then I think if you use that the Toner-Transfer will be just as easy or easier than UV.