PCB Exposure Box / Light Box

Hello.

I’m interested in the light exposure process of transferring a PCB design to the copper clad boards (with photoresist). I’m trying to do this on as little a budget as possible, so am willing to build one, or buy a “work light” that can accomplish the task with good results.

I would really like to build a box similar to the one here (http://www.scienceprog.com/ultraviolet- … -exposure/). The problem is that materials are slightly expensive.

I’d like to hear about others’ experience with light exposure methods for making PCBs (i.e., How well did it work? What light box, etc. and other supplies do you use? Do you have any suggestions about how you might have approached the task in a different way?, etc.).

I’ve not done this but have seen several others using old scanners for a light exposure box. Seems ideal and you can often find them for free or $5 at thrift stores.

Here’s my UV exposure unit:

http://www.leonheller.com/Pics/UV%20box.jpg

It didn’t cost me much to make and works very well.

Leon

mrbit:
I’m interested in the light exposure process of transferring a PCB design to the copper clad boards (with photoresist). I’m trying to do this on as little a budget as possible, so am willing to build one, or buy a “work light” that can accomplish the task with good results.

Why are you attempting to make your own PCBs?

There is a world-wide surplus of PCB capacity. Prices are

the lowest in history and the technical standards very high.

The QA level is amazing.

mrbit:
I would really like to build a box similar to the one here… The problem is that materials are slightly expensive.

This is a very sophisticated design and includes a uC based controller.

For this to be worth the time, effort, and money to build,

wouldn’t you need to be running a lot of PCBs?

mrbit:
I’d like to hear about others’ experience with light exposure methods for making PCBs…

I've attempted most hobby PCB methods. I needed quite a bit

of space and many tools. At best the PCB results

are “serviceable” and there’s a great sense of satisfaction.

However, my goal was not to make PCBs “to do” or “to have”,

instead I made PCBs “to use”. I was spending too much time

in the PCB process and loosing time in other areas of the

hobby. Also, QA was non-existant. Making two exact copies

of the same design on different days was impossible.

The home-made PCBs were two-sided but required

extra care in layout to end up with a design that didn’t

require PTH (plated-through holes) or fine pitch traces.

What I took away from the experience is that my time

was better spent doing other things, although I enjoyed

the homebrew PCB experience.

You might be much better at this than I was, and you

may want to be the “homebrew PCB king”. If that’s your

goal then copying the linked design exactly will bring you

one step closer.

It only takes a me 30 minutes or so to make a PCB (plus drilling), which is my main motivation for making my own. I can have an idea, design a PCB, and have it working in a couple of hours, instead of waiting a few days to get one made.

They don’t look as pretty as those I have made professionally, and they are a bit harder to assemble, but they work just as well.

Leon

leon_heller:
It only takes a me 30 minutes or so to make a PCB (plus drilling), which is my main motivation for making my own.

Bigglez:
You might be much better at this than I was, and you

may want to be the “homebrew PCB king”.

The OP asked for opinions. Your cycle time may be

short, and your output serviceable. It doesn’t sound

as if many of us (including the OP) have your resources.

In a nutshell, I’d prefer to do it myself since it can be relatively cheap, I’ll probably never completely stop making boards (and so my skill will grow and the initial cost isn’t super terrible).

I’m aware of the complexity of the linked lightbox, but think it’d be an interesting machine to build, and a nice one to use… that’s that.

It seems like even the cheap PCB manufacturers would be more expensive than what it would cost me to make the boards, and I’d catch up somewhat soon in savings to justify the initial investment.

I guess I should mention that a few other people will use this system as well…

bigglez:

leon_heller:
It only takes a me 30 minutes or so to make a PCB (plus drilling), which is my main motivation for making my own.

Bigglez:
You might be much better at this than I was, and you

may want to be the “homebrew PCB king”.

The OP asked for opinions. Your cycle time may be

short, and your output serviceable. It doesn’t sound

as if many of us (including the OP) have your resources.

I don’t need much in the way of resources: I use a cheap inkjet printer, the UV exposure unit, some plastic food containers, the kitchen sink, a Minicraft drill and stand. and some tungsten carbide drills. Consumables are Jetstar Premium film, caustic soda, ferric chloride and IPA. I can manage 8/8 mil design rules or less, although I usually use 10/10 mils.

Leon

When I used to do homebrew PCBs, I first tried using an array of UV LEDs, with poor results (uneven lighting and eventually I blew LEDs due to poor driver design). Then I moved to fluorescent UV tubes, which were available cheaply for lighting computer “case mods”. That worked pretty well. I also noticed that I could get quite usable exposures just by leaving the PCB under a regular fluorescent light for ~30 minutes!

I have done a lot of boards using pre-sensitized PCBs and regular fluorescent lights. Under a regular shop light (two 48" 40W T12 lights) I let them expose 20 minutes per side. I have a shelf with lights for starting seedlings so they’re at a convenient height. I print the artwork on inkjet transparencies using regular (cheap) black ink. I place it under a a piece of glass taken from a picture frame (nothing special about picture frame glass, it was just convenient) to keep the artwork pressed against the PCB, and place it about 4" below the lights. When there aren’t plants under the lights, I’ll swap out the tubes for blacklight tubes and decrease the exposure to ~3min per side.

I get very good results at 8/8. I’ve made a few test boards that go down to 5mil with no shorts or breaks, but I never use anything that small in real life – I’d expect a lot of defects at that size.

20 minute exposures sucks when you’re in a hurry. The 7ish minutes it takes with blacklight tubes is gives me plenty of time to warm up the etchant get the developer ready.

I’m not suggesting you just use plain fluorescent lights, but you can make a lot of trade-offs in designing your box – the pre-sensitized boards are pretty forgiving.

Why are you attempting to make your own PCBs?

I have to pipe in here - I completely disagree. Making your own PCB is a great way to get fast feedback. Frankly, the thought of waiting weeks for BatchPCB seems ridiculous. OK, maybe that’s an extreme statement but I really like being able to get something going measured in hours not weeks. Even faster turn around on the order of days seems painful to wait.

Sure there are limitations to the process - lack of pt holes is one of the biggest - but it really isn’t that hard or expensive. In addition, there is a real sense of accomplishment when you look at a board that you REALLY made.

Philba:

Bigglez:
Why are you attempting to make your own PCBs?

I have to pipe in here - I completely disagree. Making your own PCB is a great way to get fast feedback.

That's what SBBs (Solderless Bread Boards) and BOBs (Break Out Boards) are for.

Making a single PCB is very educational. Having to make the tools

(however crude your woodworking and lamp ballast skills are)

first is counter-productive.

Philba:
Frankly, the thought of waiting weeks for BatchPCB seems ridiculous.

Depends on how long the project process takes. Granted if one

wakes from a dreamy sleep and must build a widget by sundown

then homebrew everything is about the only solution.

Most projects go through a process of “thought design”, research,

POC (Proof Of Concept), Prototyping, debugging, code writing,

spec and functionality iteration, feature refinement, and

finally documentation and pre-production.

Protos are not just for high volume consumer products, museums

are filled with artifacts from, say, the Apollo moon mission but

so far only twelve humans have moon-walked.

If this is one’s hobby then multiple projects overlap. For example;

today I see that BatchPCB has cut a shipping label for my

recent order, I know I’ll be stuffing those boards in the next

few days, so I will make gathering a BOM and a shopping bin

of parts a priority. Also to clear a space on my bench.

This is an incentive to send out the next BatchPCB project,

which can be in China while I work on an existing prototype.

If it goes well I’ll have some time for experimenting with

another circuit idea I had after reading a magazine yesterday,

and so on. This where SFE has a niche - they source parts

that are not available through other channels.

The counterpoint is that as hobbyists we are forced to use

PCBs to prototype due to limitation on semiconductor

packaging, or interface to modules that are re-purposed from

other projects (cell phone displays, for example).

PCBs open a lot of doors for better workmanship. Prototypes

go together much quicker (minimum wiring or chassis-bashing).

Code authoring and CADCAM tools speed the work with

cut-n-paste, and make error free archived records for

repairs or mods later.

Philba:
In addition, there is a real sense of accomplishment when you look at a board that you REALLY made.

So are you constructing your own copper laminate? Did you

go out recently to find copper ore?

Another important aspect is that the hobby has risen to a higher

plane, where projects are closer to system engineering than

simple low component count circuits. I don’t get a sense that

visitors to this forum are building crystal radios, four transistor

radios, or budget analog bench instruments. When was the

last time someone posted for help with a non-uC related project?

A visit to the Maker Faire demonstrates that young hobbyists

are system builders - they don’t need workshop skills such

as PCB fab and metal-bashing to turn out inspiring works.

Good luck to the OP. If PCB fabbing is your thing enjoy building

an exposure frame, I’ve been there and tried that and I’d

rather check my work carefully and outsource it by email.

Non-hobbyists view my handiwork for what it is and what it

does - not the raw materials content. Those same folks will

throw away their cell phone if it stops working or get a new

one just to add a system level feature.

We live in interesting times!

Wow, yours is the only opinion that counts? I bet your co-workers really love you. Sheesh, “dreamy sleep”. I wasn’t sure before but now I know, you ARE trying to be rude and arrogant. and succeeding.

To be clear, SBBs and BOBs are fine but not the only solution. I’m surprised you can’t see it any other way. And I’m surprised you chose to denigrate creating the tools and expertise to make PCBs.

I suppose you LIKE waiting 3 weeks to get a board. Fine. Have it your way. Take the frickin slow boat. I’ll build my boards my way and I’m not alone on this one. To be sure there is nothing wrong with using BatchPCB but it isn’t the only way. I’m surprised you can’t see that.

Have a nice day.

Phil

and by the way, I’m quoting your posting here so you can’t go and delete it later like you seem to be want to do.

bigglez:

Philba:

Bigglez:
Why are you attempting to make your own PCBs?

I have to pipe in here - I completely disagree. Making your own PCB is a great way to get fast feedback.

That's what SBBs (Solderless Bread Boards) and BOBs (Break Out Boards) are for.

Making a single PCB is very educational. Having to make the tools

(however crude your woodworking and lamp ballast skills are)

first is counter-productive.

Philba:
Frankly, the thought of waiting weeks for BatchPCB seems ridiculous.

Depends on how long the project process takes. Granted if one

wakes from a dreamy sleep and must build a widget by sundown

then homebrew everything is about the only solution.

Most projects go through a process of “thought design”, research,

POC (Proof Of Concept), Prototyping, debugging, code writing,

spec and functionality iteration, feature refinement, and

finally documentation and pre-production.

Protos are not just for high volume consumer products, museums

are filled with artifacts from, say, the Apollo moon mission but

so far only twelve humans have moon-walked.

If this is one’s hobby then multiple projects overlap. For example;

today I see that BatchPCB has cut a shipping label for my

recent order, I know I’ll be stuffing those boards in the next

few days, so I will make gathering a BOM and a shopping bin

of parts a priority. Also to clear a space on my bench.

This is an incentive to send out the next BatchPCB project,

which can be in China while I work on an existing prototype.

If it goes well I’ll have some time for experimenting with

another circuit idea I had after reading a magazine yesterday,

and so on. This where SFE has a niche - they source parts

that are not available through other channels.

The counterpoint is that as hobbyists we are forced to use

PCBs to prototype due to limitation on semiconductor

packaging, or interface to modules that are re-purposed from

other projects (cell phone displays, for example).

PCBs open a lot of doors for better workmanship. Prototypes

go together much quicker (minimum wiring or chassis-bashing).

Code authoring and CADCAM tools speed the work with

cut-n-paste, and make error free archived records for

repairs or mods later.

Philba:
In addition, there is a real sense of accomplishment when you look at a board that you REALLY made.

So are you constructing your own copper laminate? Did you

go out recently to find copper ore?

Another important aspect is that the hobby has risen to a higher

plane, where projects are closer to system engineering than

simple low component count circuits. I don’t get a sense that

visitors to this forum are building crystal radios, four transistor

radios, or budget analog bench instruments. When was the

last time someone posted for help with a non-uC related project?

A visit to the Maker Faire demonstrates that young hobbyists

are system builders - they don’t need workshop skills such

as PCB fab and metal-bashing to turn out inspiring works.

Good luck to the OP. If PCB fabbing is your thing enjoy building

an exposure frame, I’ve been there and tried that and I’d

rather check my work carefully and outsource it by email.

Non-hobbyists view my handiwork for what it is and what it

does - not the raw materials content. Those same folks will

throw away their cell phone if it stops working or get a new

one just to add a system level feature.

We live in interesting times!

Philba:
Wow, yours is the only opinion that counts?

No it isn't, and it's disingenuous of you to think otherwise.

Philba:
Sheesh, “dreamy sleep”. I wasn’t sure before but now I know, you ARE trying to be rude and arrogant. and succeeding.

Your opinion. The OP asked for opinions. I doubt I'm the only

one here that’s had a daydream about a circuit or gadget. Judging

by some of the newbie questions it would be odd if anyone reading

this forum had not dreamed of one.

Philba:
I suppose you LIKE waiting 3 weeks to get a board. Fine. Have it your way. Take the frickin slow boat. I’ll build my boards my way and I’m not alone on this one. To be sure there is nothing wrong with using BatchPCB but it isn’t the only way. I’m surprised you can’t see that.

Once again, I offered my opinion, and it may be opposite of

yours. I explained in detail that a 20day turn for BatchPCB

boards is not a show stopper, in fact it provides a chance to

do other work on a project, or have more than one project in

rotation.

I use two other US based PCB houses in addition to BatchPCB,

but not for prototypes (both have higher $$ mins and cater to

impatient clients willing to pay expedite fees).

henryhallam:
I also noticed that I could get quite usable exposures just by leaving the PCB under a regular fluorescent light for ~30 minutes!

Good point! I've used quartz halogen 'work lights' from the

local home improvement center. Without the front glass

they produce quite a bit of UV.

A more valuable tool than a UV box would be a vacuum frame

to hold the board and artwork in register.

Another trick, this time for wiring harnesses (not that many

are used these days), is to take the handwiring out of

the prototype and lay it on blueprint paper, near a sunny window

for twenty minutes.

Run the paper through the developer roller of the

blu-ray machine. Instant template for production harnesses!

Thanks for your opinions, everyone!

I think at the end of the day, if homemade PCBs seems worthwhile to make to someone, for reasons based on education, experience, enjoyment, or whatever else, that they should just go ahead and do so until they feel something else is more suitable. Or maybe not.