PCB making at home

[akis]

I have been making PCBs at home. I am using pre-sensitised photoresist PCBs. They are covered with a thick black film to protect from light. The photoresist is green n colour. I have two types, single sided and double sided. I peel off the protective film, expose the board with UV light, and then develop in the photo resist developer. The development process removes the areas which have been exposed to the UV light leaving the rest untouched. It is easy to tell because the photoresist is green, so it is obvious when the development process works, and when it does not.

My problem is that even though the developer specifies 30 seconds at 18-22 C, it takes more than 60 seconds to remove the resist from the single sided boards and sometimes even 120 seconds to remove the resist from one of the two sides of the double sided boards. As the board has to stay in the resist for so long, some areas of it end up being removed as well, lines and pads become thinner.

Since the two sides of the board usually behave differently, or at best they both require so much time to be developed that areas which should be left alone also get affected,  I am putting this down to bad quality of photo-resist. I do not know what else to suggest.

Any ideas?

[PaulAm]

Have you done an exposure test strip?  Sounds like they're not getting exposed sufficiently.

[akis]

I have had good results with exposures from 3.5 min to 5.5 min. Last evening I upped it to 8 min, thinking that if the developer is not removing the photoresist quickly then it might need more exposure. But it still took a huge amount of time to develop and by then other areas on the PCB got compromised. On a two sided PCB you have the additional problem of one side being more ready than the other, so trying to develop the slow side affects badly the other one.

The problem happens mostly on the two sided PCBs, which leads me to believe that maybe they were badly treated to start with - maybe they used different resist on either side, or maybe the resist is too old and has aged? Can this happen?

Has anyone tried the spray-on photo-resist? Is it any good?

[poorchava]

What sort of laminate do you use (manufacturer & type) and what UV source. It might be that the peak sensitivity of the laminate is somwehere on the edge of you sources emission spectrum. This would cause severe underexposure.

Second thing: do you agitate the developer when developing the board? get a soft brush an gently rub the pcb to ensure that fresh developer gets to the parts that have not been developed yet. This is often a problem when clearances are rather small (say... less tham 12mils).

[akis]

I bought the laminates from mega UK :

*******************************************************
Fotoboard (2) Pre-sensitised 'FR4' Copper Board

The first choice for the production of quality printed circuit boards. Fotoboard is a positive working, Photoresist coated, glass fibre printed circuit laminate to BS4584 and MILSPEC MIL-P-13949GE GRN. The resist is applied 5 microns thick using a specially developed roller coating process which ensures a particularly even layer. The resist is high temperature cured which, combined with its tough plastic film laminate covering, results in a durable board which can be handled, guillotined and processed with a high degree of safety. Inspection at the development stage is simplified by use of a green dye in the Photoresist. Fotoboard is 1/16" thick, 1 oz copper. It has a minimum shelf life of one year and is available as both single and double sided. 2 oz and other thickness' are available on request.
********************************************************


The green dye makes it easy to see when it is properly developed so there is absolutely no guessing what will happen during the etching process. And yes, I agitate it for the full duration of its immersion into the developer. Maybe my light box's lights are not UV ? I will buy replacement tubes.

[Tac Eht Xilef]

I've found the mega Fotoboard to require a lot longer exposure in my setup than others e.g. 6~8 mins vs 90 secs ~3 mins for Kinsten. That said, I find it's harder to much harder to overexpose than many (e.g. Kinsten).

I also find it much harder to over-develop. 10g/L sodium hydroxide (yeah, I know...) @ 25C, gently agitate the board until the resist is washing itself off in sort of glutenous strings ("like blood clotting in water", as my sidekick says), then I pick the board up and use a dense but soft open-cell foam sponge to wipe developer on until all the exposed areas are clean. I've never timed it, but at a guess the agitating takes about 15-30 secs, plus another 20-30 secs of wiping depending on board size.

It sounds like your original 3.5 - 5.5 mins exposure is about right. I've never just agitated it in developer until the exposed areas wash clean, but my impression is that you could overdevelop if you did.

[nctnico]

I use a sponge to remove the resist after developing. The resist doesn't always come off by itself. There is no general rule on how long PCBs need to be exposed. I make test strips for each type/brand of PCB I have. Exposure times vary from 30 seconds to 7 minutes (especially for the CIF crap Farnell sells).

[akis]

I spoke to the man from Mega and after we went through the process he said the most likely problem is bad exposure, because I did not pre-heat the UV light box, so one side receives much less light than the other. He said UV light boxes are known to increase intensity by 30% when warmed up. I then carried out the test using my light meter and when cold I measured around 500-600 lux, and after 5 minutes warm up time I measured 1100-1200 lux.

So I made a new board and it worked fine, there was no difference between the two sides. It seems this was the biggest problem all along.

My final question would be, how do I test the strength of the Alkaline Metasilicate which is the developer that I use?

[nctnico]

I always use pre-made developer which I store and use at room temperature.

Regarding the UV box warming up: that is interesting. I think I'll repeat that test someday. OTOH the only double sided PCBs I make have a solid ground plane on one side so that doesn't need etching.

edit: I checked and my UV box starts at 1200 lux and climbs to about 1800 lux in a minute or so.