UV lightbox cheap equivalents?

A

alband

Senior Member
All stuff was from rapid (except the nail thing).

Developer: http://www.rapidonline.com/productinfo.aspx?&tier1=Tools,+Fasteners+&+Production+Equipment&tier2=PCB+Equipment&tier3=Etching&tier4=Photoresist+developer&moduleno=72545&catref=87-0724
PCB (smallest one): http://www.rapidonline.com/productinfo.aspx?&tier1=Tools,+Fasteners+&+Production+Equipment&tier2=PCB+Equipment&tier3=Copper+Clad+Boards&tier4=Economy+photo+etch+PCB&moduleno=29439&catref=34-0160

I kept half the black backing on it then put the white insulation tape on to graduate it.

I'm not sure about the lines iether, but they were etched away before the rest of the copper was etched so I think it iether needs more developing time or more lightbox time.

Nothing real to be said, I'm just keeping posted before I forget, and in case anyone does have a particular comment.

Now, I'll see what I can do with round 2. ;)
 
M

Mycroft2152

Senior Member
some suggestions:

Some people use an aquarium heater to heat the etchant, along with an aquarium air pump to circulate the etchant.

PCBs can be scored with a knife and then snapped. Score heavily on both sides, You may have to clean up the edges with a bit of sand paper. It's not elegant but it works and you don"t have to woory about breathing the fibreglass dust.
 
S

SilentScreamer

Senior Member
alband said:
I think it iether needs more developing time or more lightbox time.
Click to expand...
From the little etching experience I have I'd say the opposite. Either less UV time or less developing time. The developer is very fast and quickly removes too much, an extra 30 seconds and then all the photoresist has been removed from my experience.

Having said that your paper/pen needs improving. If possible use OHP transparent sheets and/or a permanent maker. Make a nice big black square to see if it works, then refine the technique for more precise lines.

The blue ballpoint pen would have done barely anything to stop the UV, make sure to use something that will fully black out the paper.
 
A

Andrew Cowan

Senior Member
I'd agree with SS and say less development time and less exposure time.

Maybe try 1 minute, 1.15, 1.30, 1.45 up to 3 minutes?

I don't time development - I just slosh it about in the developer until all the resist is gone, leaving shiny copper.

A
 
D

Dippy

Moderator
Sorry, I'm being thick today.

I just can't see where the white insulation tape comes into it.

:) I was always taught at school to describe the experimental setup c/w a drawing and the experimental method before showing the results :)

So, you cut a piece of photoresist board.
You then strip off half the black film along the length, giving (visually) a black strip and an exposed strip.
Then you stick white insulation tape somewhere...?
Ah well, as long as you know what you mean;)

Ultimately you will need to know the exposure time with the transparency-material / printing -method / etcher that you plan to use.
AND the quality of UV distribution - otherwise larger boards with thin tracks will be a pain.

If you plan to use transparency with a laser, then doing a test with a felt-tip marker is of limited use - the ink distribution is not so even and not so dense.
If you print some lines with your intended printer that really is the only sensible test - unless you like wasting board.

Question:
After the developing, does the remaining photoresist still look shiny?
If it looks pitted or matted then that usually means excessive UV exposure or the 'negative' ink/toner layer is not dense enough.
 
Last edited:
A

alband

Senior Member
Hok;

Perhaps less development time. I doubt the exposure from the nail thing is too fierce because I'm still using the bulb meant for nails, not a special UVA one.

I reckon (although am clueless as to how) the it picked up some spear ambient light between the exposure and developing.

I too normally develop by eye but couldn't see any change so left it in for ages (this could be the critical fault me thinks).

@Dippy, I'm aware my story started with the finish and has seems to go on chronological reverse! - apologies, I'll see if my second attempt at story telling goes better too.

About the second attempt though, I am making the masks using the printer and special printer stuff from rapid so that can't be the problem. I've done a few thickness lines and am about to go and do it now.

The aquarium heater sounds good (and expensive :S), doubt I'll need the pump or need to work about UV distribution because I am literally making PCBs that are stamp sized. I may in the future need to do distribution tests however if I do "biggen's" but by then I should have a bit of experience with it and should have "tamed the beast".

Thanks,
David.
 
A

alband

Senior Member
Success, of sorts...

It took about 5 more tries, clearly the first go was luck. After one where all the copper came off, one where none came off (and one before all that, that just went to pot) I did a third. It still had way too much copper but it showed clearly that it needed less time in the UV (suprising since it is still using the suplied bulb) and quite a bit in the developer (though SS you were right, leave it a tad too long and it's a goner).

So I did a fith with a real circuit on it. Attached is the result. :D

Sure, the 20X2 wont quite have all of its pins partnered with a pad but I think that is where the beauty of having almost completely selectable pins at you disposal. Also there are clearly a few that need bridging, but as a second batch effort, I'm very pleased!
 

Attachments

D

demonicpicaxeguy

Senior Member
ALband,

what ever your doing for exposure, you look to be on the correct track if you can post clear picture of what the photo resist layer looks like after it's developed, i normally use plain paper a sheet of glass and a 4ft fishtank uv light or an energy saver at 1/2ft for 10 to 15 mins

when you're developing the photoresist with small tracks, which i define as anything smaller than DIP down to 0.05mm, heating the developer is probably not the greatest idea yes it does work quicker but it also makes the photoresist layer softer and even part of the unexposed area to desolve or seperate from the copper,which looks to be one of your problems, it could also be the type of developer, personally i use "draino" (sodium hydroxide) 1 tsp per 250ml or water at cold tap temperature

it does take a little longer but 1 key part to getting working properly is agitation
i always have the pcb clamped on a pair of locking foreceps and swish the pcb around in the solution, until it's finished then rinse it with cold tap water,

the beuty of using draino is you can simply flush it down the sink :)

etching, using the locking forceps, i use ferric chloride ,again at room temperature swishing the pcb around providing plenty of agitation, and removing frequently to check on the progress you'll find heating the ferric chloride is ok if your removing large amounts of copper quickly but finer tracks for smd components are a different kettle of fish entirly,

this is where the function in diptrace called "place copper pour" comes in very handy because while your normally waiting for the large areas to completely desolve your smaller already finished traces arn't being attcked where they shouldn't be as is evident on the photo, the copper pattern should ideally be a replica of the printout, i normally use a clearance setting between 0.5mm and 1mm
 
A

Andrew Cowan

Senior Member
Developer concentraion/temperature is very important.

A while ago, I exposed my last bit of photosensitive board, dropped it into too strong developer, and within 5 seconds the whole board had been stripped of photoresist.

Better for it to be slow than too fast.

A
 
D

Dippy

Moderator
Well the manufs of the 'better' brands of PCB (and the better suppliers) can supply information on Developer types and concentration and temperature guides.

The ability to read helps too :)

The time issue is important too. Basically , you've got to get that exposed stuff off as quickly as possible without pitting the unexposed bit.
Again, some suppliers and manufacturers can give good guidance - IF anyone can be bothered to read.

I'm afraid a lot of it is down to trial and error using your own intended printing methods.

And, yes, I've screwed up in the past too. Goes without saying :(

Alband:
Nearly there, but won't win any prizes at the moment.
As said before, best you post some pics post-developer and pre-etcher.
To me that looks like over-exposure or over-devlopment resulting in damaged photo-resist and the etcher penetrating.
It's a pity you haven't got a decent camera.
it also looks like uneven exposure - which is not surpring from a non-specialist UV source.
But keep at it. I'm sure 1 square cm boards will be fine ;)
 
Last edited:
A

alband

Senior Member
It just so happens: ;)

1st is now my new way of securing the mask on, using an old CD case. It's perfect for the job.

2nd is post develop pre etch and pretty shoddy quality I'm afraid. :(

There are still 3rd and 4ths
 

Attachments

A

alband

Senior Member
3rd is during etching and shows the way it etched. It took of those pads at the end very fast and I had to leave the other end in one of the troughs of the tray, with the end that is etched stuck out.

4th is pretty much the same but a bit later. I've just realized I don't actually have and post developer and pre etch only in the developer (which is what 2 actually is :()
 

Attachments

S

SilentScreamer

Senior Member
If that board results in failure the plastic could be blocking the UV light (some plastics block most UV light).

EDIT: Looks like it did work :), I don't know if you've already posted it but how long are you leaving it in the developer and the UV light for?
 
Last edited:
A

alband

Senior Member
SilentScreamer said:
If that board results in failure the plastic could be blocking the UV light (some plastics block most UV light).

EDIT: Looks like it did work :), I don't know if you've already posted it but how long are you leaving it in the developer and the UV light for?
Click to expand...
It seemed to be getting too much at 3 minutes so I did 1 minute for that board.

EDIT: and developing it by eye, I think that time it needed a little longer. The solution is too week rather than too strong and as you can see from the "en etching" shots, there was too much photoresist in most places. About 5-10 minutes I think though?
 
Last edited:
A

alband

Senior Member
alband said:
{UV} It seemed to be getting too much at 3 minutes so I did 1 minute for that board.

EDIT: {developing} and developing it by eye, I think that time it needed a little longer. The solution is too week rather than too strong and as you can see from the "en etching" shots, there was too much photoresist in most places. About 5-10 minutes I think though?
Click to expand...
10 charact
 
D

Dippy

Moderator
Sorry, the image of post-dev pre-etch board was just too fluffy to make any comment.
Get a Canon :)

Developing for 5 to 10 minutes? FAR too long. The developer will start attacking the un-exposed areas of photo-resist, cause pitting and thinning ad lead to a poor etch.
I develop mine in about 30 seconds.

The pictures in the etching solution give the indication that areas of photoresist (top-ish right-ish in image) still remain partially.

I assume you have the toner/ink side of your negative up against the board?

It also shows that the UV distribution (i.e. 'even-ness') is poor.

When you agitate in the developer does the resist come off like an ink in water, or does it come off like a skin?

The pressure of your negative to the board should be quite firm and even. I'm really not sure that a CD case is enough. I'd be looking for something heavier - maybe the weight of a D cell or more.

Even the cheap 'proper' UV boxes press quite hard with sponge inside a clip-down lid. The better ones have a vacuum pump as well.

Is the base of your UV thing perfectly flat?
I mean PERFECTLY flat.

I see the etched result is as uneven as your previous attempt. Does it make a difference if you move the whole thing within your UV device?

And, finally, is the developer you use properly compatible with the board and at the correct concentration?

(In the past, I had several problems with this until I found the correct board/UV-Time/Developer&Conc.).

PS. Best to call the image on the film as 'artwork'. Masks usually relate to screen-printing of pastes and resists. Some people may get confused :eek:
 
Last edited:
A

alband

Senior Member
Surely is there is still photoresist then it would need to stay in the developer longer?

I haven't thought about the side it is on and now that I do it's on the wrong side, so that is one thing I can improve. :)

For the UV distribution, is there much I can do? I can't easily move it away from the light and the inside is coated with reflective... stuff.

I'm still not sure about the developing, I can't really see it come off. What I have normally experienced with developer is that the PCB goes very dark in area's then if you shake it the photoresist is just wafted away like ink. It does neither of these and doesn't come off like a skin, it just changes to blue/green and yellow for the corresponding track and non-track area's. I may get some of the developer, make a week solution and very slowly develop another and study it carefully (also try and get some decent shots with a macro lens).

You can just see in the case I have some slivers of modeling foams to make sure it is a firm fit, but I can probably pack more in and put rubber bands round the case if necessary.

No it isn't. Why? :confused:

All of the pictures are from the same attempt, but seen as it was late I posted the results then plodded off to bed.

I don't know about the compatibility and didn't know there could be such an issue but it sounds very likely. :) The links to them are in post 161 (p17). The developer say it should be at 20 deg. But in that last attempt I did it as is, from the can of stuff I mixed.

re-PS right, I though there was a better word.
 
S

SilentScreamer

Senior Member
alband said:
Surely is there is still photoresist then it would need to stay in the developer longer?

He meant photoresist in the areas that are intended to be kept as copper.

I haven't thought about the side it is on and now that I do it's on the wrong side, so that is one thing I can improve. :)

For the UV distribution, is there much I can do? I can't easily move it away from the light and the inside is coated with reflective... stuff.

Try moving the board round by 180 degrees half way though and have a slightly shorter exposure time.
Click to expand...
10 char limit
 
D

Dippy

Moderator
Apologies in advance for such a loooong post, but you seem to be overlooking some real basics.

"Surely is there is still photoresist then it would need to stay in the developer longer?"

- no, not neccesssarily.

As you know the UV 'weakens' the resist and the developer 'dissolves' it.
So, obviously, if the resist isn't coming off like ink and leaving shiny copper (as it sould do) then you have a number of issues to investigate.
1. UV exposure not long enough.
2. Developer too weAk.
3. Devloper incompatible.
4. Your 'tranaparent' medium is not transparent enough to UV.
.. or a combination of course.

NOTE:
If you require longer UV you may run into the problem where all the resist comes off during Dev.
If you can get a different medium (I use Staedler OHP laser transparency film) which has less UV absorption it will mean that you can have a shorter time which is better.
Alternatively you will have to prnt with a denser ink.
The answer is Black and White haha.


Having the ink/toner on the 'contact' side is a MUST. This is quite basic.
Even the slightest diffusion effects through the transparency will make fuzzy edges.

For the UV distribution, I'm afraid there is only ONE proper answer.... so I won't state the obvious.

No need to post photos of it developing. You should just see the resist coming off like ink exposing lovely shiny copper.
You should NOT need to leave it developing for more than a minute if everything else is correct.


"No it isn't. Why? "
- I may have wrongly assumed the UV came through from below, so I was referring to the sheet between UV and board.
If it's from above you MUST ensure the negative-board contact is flat and even.
Proper UV boxes use standard float glass about 2 or 3mm thick.
Thinking about it again, I guess to do your nails you want it from the top.

So, is your precision CD case perfectly flat?

You've got to sort out the perfect flat contact exposure for yourself. Even with dozens of photos we can't help you with that apect.

I don't know about developer/board compatibility.
This is YOUR job to investigate.
If there is no proper detailed Data Sheet then PHONE Rapid Technical and ask.

Just think; if this had been costed in £ / hour you could have bought a proper one by now and be merrily making boards :)

I've got to make a board this afternoon. Oh I guess it'll take about 20 minutes all in.
Is there an emoticon for smug? ;)
 
D

Dippy

Moderator
Postscript

Honestly, I think its time for you to do some intensive testing - AFTER you have the physical aspect sorted.
You may also have to do some research re: Board/developer issues.

It may be worth doing that before posting any more images as there is a point where the pictures don't help us IF you haven't sorted the basics.
 
A

alband

Senior Member
UV exposure

(happy sigh) well, I don't know how your's went Dippy (although I could probably guess it went fairly well) but mine went exactly "ok".

I changed the "artwork" (;)) so that it was mirrored so I could have the printed side on the board, and used that 2add copper" thing on Diptrace. Attached is the result.

After UV exposure of the first board (yes, :( FIRST) it looked hopeful. Attached is the result. This board did actually eventually work, but there is more to come...
 

Attachments

Last edited:
A

alband

Senior Member
Development

I developed the first board but not for long enough unfortunately. It looked done and I didn't want to do much more in fear of over developing the tracks. Attached is the result.
 

Attachments

A

alband

Senior Member
Etching

I etched the first one with no luck what-so-ever. It turned out to be a combination of a tad too little UV exposure (but still just enough), too week developer and not enough time developing. I left it in the etcher and prepared another board. I exposed it (still not quite enough time as it turned out) and after developing it, it became apparent that unfortunately the artwork was too far away from the board so it failed completely. I prepared a third and final board, this one got just the right amount of UV but on developing had suffered the same fate as its previous brother of not being in close contact with the artwork.

2nd attachment = 3rd board, 2nd then 1st, left to right. 1st board looks like it's etch but that is because of the colour of the ferric chloride, the board is actually white and hasn't etched at all.
 

Attachments

A

alband

Senior Member
Re-Etching

At this point I was somewhat stuck, I had near run out of board.
I decided to put them back in the developer with a bit of fresh developer (from the bottle into the solution). This had a good effect. I got the first board, which was the only one with correct UV exposure to a state where it was starting to etch. I kept re-developing and re-etch (washing it each time or course). This was, also, of course quite risky because, it wouldbe developing everything, because the tracks would have gotten a sizable dose of light.
Never-the-less, I prevailed (for 2:30 hours :eek:) and managed to get it to work.

I now have, a very usable board (which I might add is eagerly awaiting the 20X2 chips which have been even further delayed (check the microchip web site)) and the knowledge of timings and such to make more first time:

1:10 in the UV CD case.
About 40 seconds in the developer by eye (which I am now sure is compatible when used to the right concentration).
About 10 minutes in this concentration of etchant (which must be week but I can't make it more concentrated.

The finished board still does need a few bridges, but has managed features such as traces running through SMD resistors very well; once again, see attached. :D
 

Attachments

D

Dippy

Moderator
Well its getting better but shows some nasties.
Looks like penetration of photoresist.
Too strong developer? Too long development time? Ink/Toner density not good enough? Over exposure?

How did mine go?
Well, from print to board took 30 mins.

This shows it. Sorry if blurred but the excitement made my camera shake.
The area circled shows tracks of 0.325mm width.

1. Post dev.
2. post etch with resist still there.

UV exposure 1min 40 seconds
Dev time 30 seconds @ 20oC
Etch time 20 mins (Well used Sodium Persulphate) @40oC in bubble etch.
 

Attachments

D

Dippy

Moderator
PS. Most people seem to be tempted to have drill size set too high.
Make it smaller for home-brew: less chance of failure (like you have) and better centralisation for home drilling. Set it correctly only for production.
 
A

alband

Senior Member
Those pictures have ruined my satisfaction :D

Serious note though, the first picture is particularly useful for what I should be aiming for, thanks.
 
D

Dippy

Moderator
Thats good.
The blotchiness is because board was still wet.

PS. The board was FR4 Fotoboard2.

The 1 litre £15 bottle of Sena concentrate makes up 50 litres of working solution so quite cheap.
 
D

demonicpicaxeguy

Senior Member
0.325mm. i can do smaller :)

the traces in this pcb are 0.2mm and is meant for a samsung nand chip
it was produced using the photoresist method using the artwork laser printed on to plain white printer paper and exposed for about 10 to 15 mins

it is one of the better scrapped ones i scrapped due to very fine bridging between
pads this was due to the design and layout of the traces,which i've since changed

if you look at the holes pads at the top right , this is an example of a finished area being eaten into by the ferric chloride while waiting for the larger areas of copper to finish etching

Dippy said:
Well its getting better but shows some nasties.
Looks like penetration of photoresist.
Too strong developer? Too long development time? Ink/Toner density not good enough? Over exposure?

How did mine go?
Well, from print to board took 30 mins.

This shows it. Sorry if blurred but the excitement made my camera shake.
The area circled shows tracks of 0.325mm width.

1. Post dev.
2. post etch with resist still there.

UV exposure 1min 40 seconds
Dev time 30 seconds @ 20oC
Etch time 20 mins (Well used Sodium Persulphate) @40oC in bubble etch.
Click to expand...
 
Last edited:
D

Dippy

Moderator
Oh dear, I didn't mean to start a "Mine's Smaller than Yours" competition, it was merely an example.

And yes, I can easily beat that DPG and without all those scruffy bits :)
(Indicative of poor black:transp ratio of artwork negative and the need to overetch)
I've done TSOP circuits that make you weep with envy...

One of the 'secrets' is to produce artwork with the following criteria.
1. 100% dense 'ink'.
2. No pits/holes
3. 0% UV absorption by media.

Impossible of course. Only the 'pro' laser photoplotters approach that.
As we know, ink isn't perfect. Inkjet being worse, even on expensive coated film.
Therefore if we can select a medium which has less UV attenuation then the exposure time can be reduced.
And with a reduction in exposure time then less UV will go through the 'ink'.
This leaves the covered areas of photoresist less affected by UV.
Result: a better board without all that pitting and track failures shown in all the photos - apart from mine (thankyou fans).

Almost analogous to Signal to noise ratio or contrast.

All pretty obvious of course - after you've been told ;)

And before I get too smug; yes, I've had many failures in the early days until I settled on the method I use now.
I still have the occasional wobbly/iffy track when using cheaper 'economy' boards.
And I like to play safe and use thicker tracks where possible... just in case. I'd sooner a 100% board rather than trying to show off.
Girls aren't impressed when your chat up line involves PC manufacture - however, Alband might be more successful as he can offer to do their nails for them.

Tip of the Day.
Where possible/practical do plenty of copper fill. Less copper in your etchant means your etchant lasts longer. (Only applies to those who mix up a lot and re-use).
 
D

demonicpicaxeguy

Senior Member
i didn't mean to try and start a "Mine's Smaller than Yours" competition
i was just merely showing what can happen to the copper traces after they have finished etching and are still immersed in the etchant while the rest of the board finishes

the poor black:transp ratio is caused by the fact that i don't use over head transparencies, apart from not being able to get them at a reasonable price plain paper is simply cheaper and i still get the desired effect at resolutions down to 0.1mm very reliably i also print onto the paper with a laser printer on the darkest setting

the pcb in the picture was never meant to be used in a circuit, it was one of the many off cuts of the stuff that i used for a quick etch test to see how well the layout would etch, parts of the photo resist were lifted along the top

as you can see between the fine tsop pads all that scratching is from my knife where i can checking to see how bad the tiny bridges were
i've since changed the layout entirely and added the copper fill
the board is intended as a breadboarding adapter for a samsung nand flash chip which worked out very well in the end
 
D

Dippy

Moderator
I was just teasing :)
Glad it worked.

Can't you get some transp film mail/emal order?
You'll notice the improvement.

I know its expensive - but so is board.
 
You must log in or register to reply here.
Top