Building a UV exposure box

There are two methods of using etchant resist when making circuit boards. We use the toner transfer method that requires ironing on laser toner to the copper, but you can also use chemical resist that reacts to ultraviolet light. [Bogdan] decided to start doing more of the latter so he built a UV exposure box to make the process easier.

It is possible to use flourescent light bulbs for this, but he decided to use UV LEDs, a method we’ve also seen before. But there’s always room to innovate, and [Bogdan] built-in a couple of nice features that are new to us. Because the UV light can be bad for your eyes, he included a set of red visible-light LEDs on the bottom half of the box that are used to align two layers of exposure mask when making double-sided boards. There’s also a switch that automatically shuts off the UV light when the box is opened. And as the coup-de-grace, he added a programmable timer to regulate the exposure, using his newly created box when etching the PCB for it.

Comments

  1. Stu says:

    I made one of these too, out of an old flatbed scanner body, aluminium foil lined on the inside, with two 4Watt insect killer uv lamps, a lot safer on the eyes and skin. The lamps were on special offer too, 2 for 1, both cost only 13quid!
    I’m achieving very fine traces using uv sensitive copper board after 15mins exposure, could probably cut that time down too.
    The same shop I bought the uv lamps (Maplins) also sell proper uv exposure boxes, upwards of £135!!! Saved a bomb.

  2. macw says:

    come on, guys — you should know it’s “fluorescent”. I am this close to writing a little greasemonkey script to automatically replace all instances of “flourescent” because it bugs me so much :P

  3. waffles says:

    this would be good for screen printing. I have another use for this to. thanks for posting. you could also use reptile flourescent lights. they come in several differant spectrums, would be worth experimenting

  4. Brennan says:

    15 minutes? I built a box almost exactly like the one in this pulse and it only requires 1 minute. Those UV LED’s are a surprisingly good solution and you can buy them in bulk cheaply on ebay. My box cost me about $20-30 total

  5. Brennan says:

    *pulse=post. Gah! HAD why won’t you allow post editing!?

  6. Tom says:

    $10 facial tanning lamp (4x UV tubes) works great.

  7. aaron says:

    drill a hole at each end so you can insert clear piping and it could double as part of a portable water purifying system.. for when you are not being a mad scientist of course..

  8. Philippe says:

    I always wince when anglophones use French to sound smarter… and fail because they do it wrong.

    Coup de grâce (note the â) means “merciful blow”, the one, given by the executioner, that finally brings death after the condemned has suffered enough on the wheel. As you can see, it’s an expression with a pretty sinister origin.

    Is that what you meant? I didn’t think so.

    So, on a lighter note, if you meant the cherry on the sundae, it’s “la cerise sur le gâteau”.

  9. DA says:

    > I always wince when anglophones use French to sound smarter… and fail because they do it wrong.

    > Posted at 4:42 pm on Oct 20th, 2010 by Philippe

    Français râleur!!!

  10. absolutezero says:

    What photoresist products do you use with an exposure box such as this? The pre-treated PCBs seem to be a bit on the pricey side and the only spray I can find looks to be $40 per can. I find plenty of info on people building lightboxes, but don’t see much dealing with making or applying photoresist. A lot of the sites show a photoresist film that requires a laminator.

    I’m pretty sure the local office store can make a print that would work for toner transfer. What is the benefit of the uv exposure method over toner transfer?

    I have enough UV LEDs in order to construct a box, but I also own an iron. Its really a toss-up for me without having more information on the photoresist end.

    Would a UV box such as this be a good solution for exposing silk screens?

  11. bobdole says:

    UV LED’s are usually in the range of 405 nanometers or so, which isn’t even really technically ultaviolet. It’s a bit of a stretch to even call it UVA.

    Most etching boxes use industrial grade UVC lights, which DO cause cancer, and you DO want to shield yourself from.

    UV LED’s are less dangerous than a black light, wood’s glass peaks somewhere around 380nm. You can take off your welding glasses and SPF100 sunscreen, those LED’s aren’t going to hurt you. It’s not until you encounter light in the sub-200-nanometer range that you have to worry about cancer.

    absolutezero: I was under the impression that most silkscreening masks were *light*-sensitive, not necessarily just UV. I recall a girlfriend of mine doing some silkscreening stuff for a university art course, and I’m pretty sure she just used a 100W incandescent bulb. I’m not sure you need UV light for most silkscreening stuff. I’ll agree though, all things considered, that using the toner-transfer method is easier and more straightforward. Unless you’re planning on making hundreds or thousands of boards at a time, I wouldn’t bother with photo etching.

  12. macw says:

    @aaron — it’s not the right wavelength. Germicidal UV lamps operate around 254nm, while these LEDs would be in the 380-410nm range. As far as I know, LEDs in that deep UV range have a very short lifetime and cost hundreds of dollars each…all current systems use a mercury-vapor tube.

  13. Philippe says:

    @DA ah, mais je t’emmerde !

  14. Brennan says:

    @absolutezero:

    I get boards that are pre-coated with a positive photoresist from MG Chemicals. Less work that way, even though it is less DIY. The overall cost savings isn’t quite as much over a cheap prototype service like BatchPCB when you have to buy pre-sensitized boards, but the satisfaction you get doing it yourself is definitely worth it. And the turnaround time is an hour, instead of weeks.

    As for the advantage over the toner transfer method, it’s all about precision. The more precise you can etch the copper, the tighter trace widths and spacings you can get. I can get down to 5-10 mil consistently with the UV exposure method.

  15. threepointone says:

    guys, rtfm or insert other phrase here

    mercury vapor lamps are 350-400nm, so UV LEDs are well within the range of most photoresist cures. the problem with UV LEDs is simply that they cost a shitton for the little amount of UV they put out. UV blacklights are far cheaper for the power you’re getting for this application, even if they don’t sound as cool (which is the only reason anyone wants to do LEDs)

    the bigger problem, which i’ve seen in almost all exposure boxes, is the lack of collimation. the light coming out of most exposure boxes here is not well collimated at all–the LEDs spew light out at a crapton of angles. what does that mean? suppose you have a ray of light going at 45 degrees from the normal, which is very reasonable (at these distances, it’s probably even worse). your photoresist thickness is 1-2mil, so you easily get 1-2 mil undercutting on both sides of a traces. 8 mil becomes 4 mil–good luck with that. and it’s probably worse than that in real life. . .

    really, just build one of these: http://www.thinktink.com/stack/volumes/voliii/equipment/uvlamp/uvlamp.htm

    these guys have been around forever, and are pretty darn legit.

  16. nfs says:

    “3A protection diode”
    You realize that the max LM317 can handle is 1A and when theres 1A current running through it its already unstable so lets say ~800mA.

    And not to mention that this is not real UV.

  17. djrussell says:

    could also be used to test the UV stability of materials. how long will it take for x material to turn yellow?

  18. erksome says:

    I wonder if this would be good for explosing cyanotypes and other alternative photography. There are different optimal wavelengths for different processes and I wonder if the UV LEDs may have too narrow a wavelength range for being a good general solution…?

  19. IsotopeJ says:

    @threepointone, LEDs usually have fairly narrow viewing angles, and most light comes straight out, not at a “crapton of angles” so actually LEDs are fairly well suited for this application. ( I didn’t even know you could get 140-degree UV leds, so maybe undercutting is a problem here.)

    Incidentally, if you do use narrow angle LEDs the light tends to focus in a bunch of circles, so you need to have a good distribution of bulbs as well as make sure you have enough distance between the LEDs and the exposure surface.

    I made a box just like this, with 22-degree bulbs and a depth of about 10″ so the light hitting the glass at the top is pretty even.

    With 90 410nm leds in an 8×10″ area I can expose silkscreens in 10 mins. It’s my understanding that visible light won’t cure screenprinting emulsion. I still haven’t tried PCBs yet.

  20. Noobius says:

    @nfs
    ““3A protection diode”
    You realize that the max LM317 can handle is 1A and when theres 1A current running through it its already unstable so lets say ~800mA.”

    Maybe you should read up on how diodes work. If it’s a 3 amp diode that means that the max forward continuous current it can handle is 3 amps. It isn’t there to limit current. It’s there to protect the LM317 from negative voltage on the input.

  21. bogdan says:

    @absolutezero
    I use pre-sensitized boards because using spray on lacquer is just too complicated: it smells, you have to let it dry, keep it dust free, do special moves to get a uniform thickness. I did start using spray, it was called POSITIV 20. But it expired before i got the chance to finish it and i only found out when the board didn’t come out right. For such small scale the cost difference is worth the trouble.

    @hreepointone
    I have not noticed any undercutting, though i only made a couple of boards with 10mil tracks. I might check that in the future. What would you suggest as a collimator?
    As for your suggestion of 18″ tubes. It defies my needs and purpose: small compact box. Of course it is cheaper to make with tubes. you just can’t make it so small(ok, maybe you could if you could source 2W UV tubes, which i couldn’t).

    @nfs maybe you should read the data sheet of the LM317 better. Short answer: it can handle 1.5A. and it is stable at such current if you use it right. The whole box consumes about 1.2A. That three amp diode is the next larger thing i had after 1n4007.
    UV light is below 400nm. The leds i use have a peak wavelength at 405nm. It’s safe to say half of the energy they put out is below 405(almost all UV) and half is above, which is visible.

  22. bilbao bob says:

    @phillipe – tous le monde était interresser de connaitre… well, in english… You’re bent out of shape because the guy didn’t bother to add a circumflex? Or because he used the english spelling of grace instead of writing gras as many americans do? Heh.

    I hate to be the bearer of bad news, but your mastery of english idioms and colloquialisms is weak – it’s been common for americans to use this phrase to describe “the best feature/attribute/point” since the 1950’s, and possibly long before that. We usually use it as a synonym for the word “killer”, which shows our culture in an interesting light.

    Is it the same usage as a native french speaker? No. But I suspect that your suggested phrase “The cherry on the top” fell out of favor circa 1960, having been used for roughly 80 years.

    Languages are living things; therefore, y’all aint got no reason to be hatin on the boy’s grammar, ma petite ami. :)

    And yeah, it’s like that.

  23. bilbao bob says:

    @threepointone

    This guy knows what he’s talking about.

    Little boxes like these are OK for one-offs, but diffusion will increase problems with narrow traces if you’re doing really narrow traces for surface mount and then etching – especially if the UV exposure was weak.

    Old bulb based tanning lamps work pretty well, if you keep a fan on them to keep the heat down and use very thin glass to hold down the artwork. You can set them 6-12″ above the pcb.

    PS – An industrial wand style “personal massager” set on jack-hammer mode will do an excellent job of agitating a tray full of etchant. I always meant to try an ultrasonic etchant tray but never got around to building one.

    I still remember building a hundred and fifty or so populated PC boards by hand as a kid – from bulk pcb material to test. It was a Tom Sawyer thing, with friends coming over to do the work.

    Got them done in a weekend, and I suspect the combined Etchant/Lead/plastic fumes caused permanent damage.

    They looked like south china specials, and since I didn’t understand through hole technology or proper bit sizing, I used big drill bits and then used solder – lots and lots of solder – to fill the gaps. Awful, but they worked.

    The day I cashed a check for a week’s work that was more than both my parents had made in three months was the day that I decided to become a tech whore. I’m a little played out, but my clientele likes that kind of rough experience. Luckily I have a great Pimp. :)

    Don’t do drugs!

  24. Tom says:

    @bilbao bob

    Are you trolling? We all know that Americans exist only to destroy other languages and pass it off as acceptable with the strange argument you just used.

    I offer a counter argument to the mass stupidity and ignorance, read on…

    ‘coup-de-grace’ already has a fine meaning, don’t add another one. Just choose an arbitrary string of letters you can pronounce.

    If phillipe is a native French speaker, or indeed has any grasp on the French language, I’d say he has more than enough reason to call out terrible use of ‘coup-de-grace’, like here in this article.

  25. tim says:

    i am a native french speaking, and i confirm, it is perfectly ok to use “coup de grâce” here. Even if this sentence is “tombée en désuétude” because we have abolished death penalty a long time ago, it is still used in this exact context.

  26. bilbao bob says:

    @Tom

    “`coup-de-grace’ already has a fine meaning, don’t add another one. Just choose an arbitrary string of letters you can pronounce.”

    I say, Tom, do you study a lot of linguistics? :)

    If you can figure out how to stop adoption, borrowing and modification of meaning within host languages, you’ll be eligible for the Nobel prize, or at least honorable mention by L’Académie française They’ve been fighting it for a while with no luck.

    The french word for idiom is basically “idiotism”, and that’s what those little catch phrases are. If you want that kind of precise definition, you’ll need to use a dead language, like latin. And even latin continues to morph – not as quickly as living languages, but there have been shifts over the last 200 years as we lose the context those languages were used in.
    Even programming languages shift.
    Look at BASIC.

    Things change along with society.
    Skipping words like Freedom or Liberty, let’s look at the word “punk”. Do you think anybody who called themselves a punk circa 1977 knew that 40 years earlier it generally meant “young gay prostitute”? Ya wanna see some action? Indeed.

    Or the french word for hospital, which is pretty much like the anglo word, except that the S fell off a truck some centuries ago and was replaced by a circonflex that was hanging around. It used to mean a place where you could go to die and receive a bit of random attention, if you had money, but now it means… well, maybe that’s not the best example.

    I speak french, but generally in the context of picking up girls and arguing politics. I cannot offer definitive proof, but I have heard this phrase used to describe something cool in this exact manner by french people.

    I mean, at some point you have to allow that the phrase “Spanking the monkey” has nothing to do with monkeys or spanking.

  27. Garbz says:

    I like the idea here, and to all those people complaining about the definition of UV or the wavelengths, who cares? This box serves a purpose to expose PCBs, and having used these LEDs for photo resist PCBs myself it works very well.

    One thing about the diffuse lighting though, it does create a problem for small traces. You can use a grid made of densely packed black straws cut about 50mm long to “collimate” the light. Ok so it’s not really collimated but it does actually do a very fine job at decreasing stray light and making the rays hit the board perpendicularly.

  28. bogdan says:

    @Garbz, you made a good point, and in the end that was my requirement too, to make a pcb exposure box.

    35mm straws are not that ok for me, i’ve only got 35mm between the tip of the led and the board surface, but i will think of a way and do a test. it would be best to compare.

  29. Jawa.jay says:

    I want to make a UV LED exposure box for testing UV exposure on cultural materials (museum collections). I am looking for 340nm UV LEDs. Anyone know of a source?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 92,123 other followers