solder mask - UV solder mask procedure

[r00n]

Hi. Lovely Day for PCBs.

I am trying to do my first solder mask  ---> I have had about 20 failed attempts and I am down to my last bit in the tube. It is green of course. All the information I have on masking obtained from you-tube videos so I thought if I documented my recipe you may be able to determine where I am failing.

Here goes:

1. photocopy the mask in duplicate onto 2 pieces of what we used to call Overhead transparencies, the plastic sheets that lecturers used.  I tape the two sheets together being careful to line up the black masks to block out more UV.

2. sand copper with a bit of really light grade wet and dry paper. wash the copper with isopropanol.

3. put enough UV solder mask down on the copper to cover the board and place another transparent sheet over the top. Use a credit card to squidge out the UV solder mask with an even covering over the copper.

4. place my duplicated photocopy mask over the top plastic lining up with the solder pads.

5. place under UV light for about 4 minutes.  (I completed a test exposing the UV solder mask to UV for varying times incrementing in 30 secs.  For 7 minutes the porridge was too hot, and 2 minutes it was too cold)

6. remove the plastic masks and sheet and wash the pads with a small amount of alcohol and a paint brush being gentle.

should be ready to solder then...................................but no.

The error occurs during removal of the mask and plastic sheet. The mask sticks to the OHP transparency and not the copper. Not all - but most.  For the UV mask left on the copper, the pads are still moist enough to clean off.
Any tips you have would be well received. Or perhaps there is a better method altogether?

[thm_w]

https://learn.adafruit.com/how-to-solder-mask-pcbs/applying-resin

I don't think the transparency print needs to be duplicated?

What UV light is used?
Photos?

[coppercone2]

printing on transparancies is a little dodgy. I put it this way, I put a piece of kapton tape on a PCB during exposure to keep the transparency on. The area under the kapton tape was nearly perfect but the traces were a bit like swiss cheese for the same exposure time. Set with maximum darkness on the printer.

But I did do a long etch, that is.. no polygon on top. So it was in the solution for a lot longer then necessary. If I put polygon I could have drastically reduced etch time and there would be no problem on the ground plane.

my result was acceptable, but I can see it being improved.

i use presensitized boards though.

I heard people using some kind of translucent tracing paper with better results then transparencies. BTW transparencies are not so transparent to UV anyway. For instance a clear acrylic sheet, put under the UV light, you can see strong attenuation. SO paper is not that much worse, apparently. You do use glass to press it down on the pCB?

[r00n]

thm_w - thanks for the link. im not certain on the light that i am using but it is strong enough to cure the resin.  After looking and reading the link I tried a shorter exposure time of 30 seconds and a thinner spread of resin.  The resin cured and the mask didnt - which is exactly what is required, however it was stuck to the transparent film and not the copper.

I heard people using some kind of translucent tracing paper with better results then transparencies. BTW transparencies are not so transparent to UV anyway. For instance a clear acrylic sheet, put under the UV light, you can see strong attenuation. SO paper is not that much worse, apparently. You do use glass to press it down on the pCB?

Coppercone2, I'm certain that the transparencies are the fault.  If you have any more info on that tracing paper that would be useful info. not using glass but will give that a try

edit - maybe i could just paint it on with a brush and leave the solder pads

[shapirus]

I don't think the transparency print needs to be duplicated?

I guess it depends on the printer, but in my case (for etching with photoresist, not for solder mask, but the reasoning is I think the same) I absolutely had to print it twice, normal + mirror, then, after drying the ink (I'm using an ink jet epson L805) put the two together, ink sides facing each other, align, fix them together using sticky tape, and it's good to go.

With a single print it was a consistent failure because of very low contrast: black areas were half-opaque with multiple small transparent dots without any ink in them, so it was impossible to expose uncovered areas enough and not expose the covered ones at the same time. However, with two prints combined it became a consistent success (the biggest problem now becoming the reliable application of the photoresist film onto the board), and with a very decent resolution, too: 10/10 is not a problem at all, and, judging by how well the copper text with component reference markings serving as poor man's silk screen is etched, even thinner traces and clearances should be possible.

Can also print twice on the same surface for the same effect (after thoroughly drying the ink which I do in a ventilated oven at 50 degrees C), but, even though the printer does a very good job at consistently aligning the image (or the sheet?) to print in exactly the same place, second printing spoils thin white (transparent) lines, which get even thinner, because the already printed areas can't soak enough of the new ink and it spreads a little outside of where it's supposed to stay. So I abandoned it and kept using two transparencies, at least where traces are thin and dense.

Firm pressing (with some glass) of the transparency against the board is vital, too.

I wonder what can go wrong using the same technique for solder mask. I've never tried it, and am now tempted to try it too.

[shapirus]

however it was stuck to the transparent film and not the copper

If it cures properly, then it either lacks adhesion to copper, or adhesion to the film is too strong (or both). For the former, you want to make sure that copper is absolutely free of any grease residue. IPA doesn't do a good job of removing it. Acetone or purified gasoline (light fractions of oil refining, usually sold as solvents), whatever is more available, will do it. Need to make sure that the stuff doesn't leave any residue after evaporating. Of course don't touch the copper after cleaning.

For the unwanted adhesion to the film, maybe try a different material? For example, polyethylene film (used as food wrap etc). PE is well known for its inability to be glued with (most?) glues, which in this case is an advantage.

[coppercone2]

how to do it
1) set printer to maximum density in advanced settings (mine goes to +6)
2) set the quality to max (1200HQ on mine)
3) set the whatever setting to super-something (it has options like 2x, 4x, and super). Its like alliasing

I think you set it to transparency mode too.

Then it does better. I got 0.5mm traces OK on mine (MSOP package) on yesterdays etch.

I do 4 min with my 750mA 16V single die UV diode at a distance of like 12cm

I have some film. Its old. I have not tried it. Maybe I can try for you because i got a blank copper (failed exposure). I have been meaning to try the film I Have on something. Wow. I got super derailed from my etching projects!

vacuum the inside of your printer and wipe the rollers and whatever you can too.

[abeyer]

If your film on the first layer against the resin is the printable stuff, make sure you're not putting the printable side down (and that it's not double-side printable), too. It's treated to make ink/toner adhere, so likely could do the same for resin.

[coppercone2]

I put my printed side on the board so there is less reflection. If its flat on it then its flat on it.

[sleemanj]

See this post I made some years back with various DIY soldermask info...

https://www.eevblog.com/forum/beginners/diy-solder-mask/msg920711/#msg920711

It has been a long while since I have used it, but as I recall you need the right type of plastic to ensure it doesn't bond to the plastic, and if you are not curing down to the PCB then you are not curing long enough (or you have it on too thick). 

In short, the ink stuff is a total pig to work with at a DIY level.  Best if I recall to thin it (with
Isoamyl Acetate I believe, good luck finding it), spray it, bake it (don't remember temperature), then expose it, and wash it.

[coppercone2]

oh haha i just realized this thread is about solder mask

ok

my experiance with it is shit. paper would be worse. i thought it was for the photoresist for some reason, because i have been working on doing this board for like 3 weeks

my experiance is that I tried it once and it pissed me off because you really need a clean board to do it on. if you have any solder on it or bumps it really gets bad. i did not want to super clean the board i was repairing. i got solder mask on it but it looks like someone hit it with a painters roller brush. but its enough to shield the mainboard of the rf generator i was repairing from leaked battery damage. i ended up using a paint brush

I think mine was sticking to the acrylic too. I thought maybe a solution might be to put a thin layer of wax on it. I saw some guy use like a roller to push it around but for me it was like poking a turd with a stick. But I had some components and stuff on the PCB still, I made like a patchwerk cutout to put over most of the traces to try to squeeze it around

my theory on doing this right
1) you need some kind of special roller for that goo shit, or just experiance rolling goo. I had some art store one. Applying even films of goo is hard.
2) some kind of release agent for the film


i almost bought more goo but fuck i see you can buy a mask Ima try the mask I dont want to deal with that mother fucking goo anymore holy shit that pissed me off so hard its like trying to smear shit evenly into a picaso. lets call it tomato paste  . serenity now serenity now


bought a film mask.
ill let you know how that goes in some amount of time. I will polish that old PCB up with neverdull to get the tarnish off, apply the photoresist mask, etch it, and then if it works try the photoresist mask. if for some ungodly reason the silk dry film is worse then the goo, I might buy another tube of goo. but i would be happy to never see the goo again.

[coppercone2]

is anyone here a professional silk screen goo roller? i wanna know how that is done in the factory. its one of those processes probobly where there is fine details and extremely neutrotic attention to tool choice. or im just fuckin incapable. or there is a old master doing that stuff with a wooden spoon like in a pizzeria. some how those guys get gauge block level of sauce flatness on a pizza using a 1/3 cup spoon. the lighting in those places always sucks so you know hes seeing in monochromatic from the malfunctioning flourescents

the other idea I got is to get another sheet of transparancy, cut a square out of it, put it on the PCB, and then run a oiled blade over it to smooth it out, like a long razor, or a starrett level

[sleemanj]

For soldermask they use a silkscreen process to apply the ink.


Or it will be spray applied




For the silkscreen text layer they can use a silkscreen and imaging as with soldermask, but also they can use direct printing these days like a massive inkjet.

[sleemanj]

This is a good overview from PCBWay on the Soldermask process...



Clean board, apply the ink with a silkscreen/spray, oven dry, expose with mask, wash in alkaline, oven bake.  As above the drying temperature is somewhat critical, and low, and ink dependant. 

[coppercone2]

how do you spray that goo like a billion psi??

[sleemanj]

I don't know if it is done in factory but supposedly the goo can be thinned with "Banana Oil", which is Isoamyl Acetate. Maybe other solvents. 

[sleemanj]

Thought I had written about this before, a mere decade ago...

https://www.eevblog.com/forum/eda/diy-problems-with-an-enormous-508mm-board/msg396222/#msg396222

I got some Ethyl Acetate. It doesn't smell as nice as Banana Oil but it works.

[coppercone2]

oh I have some too, that replaced MEK as paint stripper in home depot

why do I suspect total shenanigans with dilution of soldermask using solvents? I diluted epoxy before, it dont' work too good.

[Jwillis]

First off. Don't try this in broad daylight. That will make the job harder. Set up a room that you can make dark and purchase a red or yellow LED bulb. I prefer yellow because it makes it easier for me to see. Incandescent bulbs will work as well. All work is done in the "Dark Room". I use UV activated mask form Aliexpress Vendor.
I don't use roller because the crap goes on to thick. A silk screen and squeegee work the best for me. You may have better luck with the roller.
Put a thin layer of mask on the PCB and Dry using a heat gun or other source. Dry at around 160^oC. That's an arbitrary temperature, just don't burn it.  It must be completely dry before the transparency is attached.I use double transparencies to achieve maximum shielding from the UV. One transparency allows to much leakage. Expose with UV light. Time depends on the light source and intensity. I use UV LED strips set up side by side on an  8 x 12 inch Aluminum plate under a sheet of glass. Usually it takes about 6 to 10 minutes at high intensity. Let the exposed PCB rest in the dark for 30 minutes. Wash with the appropriate solvent. This is tricky because the solvent I use is from a silk screen vendor on Aliexpress and I can't get an English translation of the label. But it smells like Methyl Ethyl Ketone. I won't guarantee that this is without some possible health risk. I haven't grown any new appendages as of yet but do use caution and good ventilation. 
Silk screening takes practice with lots of failures until you get it right. It makes a really nice job once the procedure is perfected. But it's time consuming and messy. So being set up with all the necessary supplies and equipment before you begin makes the task much smoother. Use silks of 300 count or more for lettering. To low and it doesn't look nice.

Regards
 

[coppercone2]

lol just general ranting about PCB... my board.. I drilled all the holes, mounted all the PITA smt parts, etc. Then I decided ima just test the regulators before i power the whole thing up with a jumper

Ok regs dont work. hmm. not minimum load problem..

turns out I unmirroed the mirrored part I randomly chose to check as a sanity test and I mirrored the bottom instead of the top so my PCB was backwards! so the part orientation was wrong but it made my design seem right.

at least rev #2 I will wait till I get the silkscreen film, and put polygon on top, for more even etching, reduce UV exposure time too.

maybe I will get a beautiful board this time, or at least the correct orientation.

 

i made a orange light bar powered by 2x d-cell for the dark room work BTW. glued a bunch of orange power LED modules to a aluminum square piece i cut into and made a diffuser for it. I set it up standing in the corner, its enough towork from

also for alignment, this time I will buy a 1/16 sheet of acrylic and use it to line up the top and bottom layers. I lined them up perfectly on top of each other but when I etched the board i got more then a mm of offset! i think its because the thickness of the board ? like it was dead on...... :'(

[shapirus]

orange power LED modules

Please elaborate on these.

I've been working with negative photoresist with a regular 9W 3000K bulb (in a meter or so from it) and it seems, only seems, not to create any negative impact, but I guess I might want to try something that's considered better.

also for alignment, this time I will buy a 1/16 sheet of acrylic and use it to line up the top and bottom layers. I lined them up perfectly on top of each other but when I etched the board i got more then a mm of offset! i think its because the thickness of the board ? like it was dead on...... :'(

The board must have slipped when you flipped it to expose the other side. I've done 2mm thick boards with no offset between the sides, but it does require attention to how you grab the board and the transparencies that cover it from both sides to flip it.

Of course the transparencies must be aligned and reliably fixed against each other with e.g. some scotch tape before inserting the board between them.

[Solder_Junkie]

For the past 47 years, I have been making my own PCBs. Initially hand drawn with a Dalo pen and then using cheap Chinese laser toner heat transfer paper. I spray the board after assembly with printed circuit lacquer (conformal coating), it doesn’t look as nice as solder mask, but it protects the copper from tarnishing.

While I like the idea of having a board ready to solder in an hour or so, recently I bit the bullet and used JLCPCB. OK, I have to wait a couple of weeks for the board to arrive, but zero mess and the cost is minimal.

Perhaps not the answer for those who like to make their own boards, but I’m a convert.

SJ

[coppercone2]

orange power LED modules

Please elaborate on these.

I've been working with negative photoresist with a regular 9W 3000K bulb (in a meter or so from it) and it seems, only seems, not to create any negative impact, but I guess I might want to try something that's considered better.

also for alignment, this time I will buy a 1/16 sheet of acrylic and use it to line up the top and bottom layers. I lined them up perfectly on top of each other but when I etched the board i got more then a mm of offset! i think its because the thickness of the board ? like it was dead on...... :'(

The board must have slipped when you flipped it to expose the other side. I've done 2mm thick boards with no offset between the sides, but it does require attention to how you grab the board and the transparencies that cover it from both sides to flip it.

Of course the transparencies must be aligned and reliably fixed against each other with e.g. some scotch tape before inserting the board between them.

I don' know. I aligned it on a light table. I used 3 pieces of aluminum tape to secure it. I cut a hole on one side and used kapton tape to stick it to the PCB so it does not move when I flip it. I put a 1/4 inch thick piece of glass on top for the exposure too. no idea how it moved. it was aligned enoguh so that the 0.3mm holes all over the PCB are all shining light through (drill marks)

[Jwillis]

orange power LED modules

Please elaborate on these.

I've been working with negative photoresist with a regular 9W 3000K bulb (in a meter or so from it) and it seems, only seems, not to create any negative impact, but I guess I might want to try something that's considered better.

also for alignment, this time I will buy a 1/16 sheet of acrylic and use it to line up the top and bottom layers. I lined them up perfectly on top of each other but when I etched the board i got more then a mm of offset! i think its because the thickness of the board ? like it was dead on...... :'(

The board must have slipped when you flipped it to expose the other side. I've done 2mm thick boards with no offset between the sides, but it does require attention to how you grab the board and the transparencies that cover it from both sides to flip it.

Of course the transparencies must be aligned and reliably fixed against each other with e.g. some scotch tape before inserting the board between them.

The orange , red or yellow light is for work purpose and not to activate the mask or photo resist. Light on the low side of the spectrum won't activate the the UV materials while your working. Broad spectrum lighting will. Also the photo resist film will begin activation as soon it is exposed to broad spectrum light. I lost entire rolls because of this. Now I only unseal rolls in the darkroom under yellow or red  light. If your photoresist roll has turned from a pale blue to darker blue or purple, Don't expect good adhesion or etching.
If you are using emulsions for making silk screen prints, the emulsion must be used within a month after the activator powder is added.

To align the transparencies ,I drill a couple holes to insert pins through the board . That holds the transparencies from misalignment. You can use existing drill points in the circuit art if you like.

[shapirus]

Light on the low side of the spectrum won't activate the the UV materials while your working. Broad spectrum lighting will.

Yeah, I kind of know that, was just wondering whether there were some specific LEDs that could be recommended, or, rather, what configuration would be more convenient, from experience, than others. I've just searched aliexpress for "orange led light", and it did show a few results, but of those potentially suitable were only some 600nm (which is yellow/orange) LED strips. I guess I'll search for some higher power LEDs of this part of the spectrum, maybe even our local stores have something. Strips could work just as well though.