Direct imaging of solder mask with UV printers

[cjb2051]

Hi folks,

       I oversee a prototyping lab for fab/assm of PCBs and I’m exploring ideas to simplify the process of applying the solder mask treatment. Mostly I’m trying to remove the need of practice/experience from this process in a prototype setting so users can produce a precision solder mask without the headaches that are encountered with the manual application methods and inexperienced hands. Along these lines, directly printing/imaging the solder mask layer onto the board is attractive. There are some companies (Limata, Orbotech) that appear to offer this specific capability but it’s way beyond my budget.

I’m wondering if anyone has tried (or maybe can tell me why its not suitable) to directly image the solder mask layer onto a PCB with a more generic UV printer. I realize these printers are also not cheap (particularly the Epson/HP brands). But there are some Chinese brand UV printers (Wer, Wayprinter, Funsun) that are within my budget of ~$3-4k and they’re happy to provide test samples. The stated printing resolution should be better than 1 mil and they claim the ink will adhere to copper and fiberglass. Board registration is a potential sticking point but I don’t feel this is insurmountable with a well designed board fixture.

Is there any reason to believe these UV printed inks would not function as a suitable solder mask?

[Berni]

Make sure these UV inks can stand up to soldering temperatures. They might make a lot of awful smoke in a reflow oven, or melt off the tracks when a soldering iron is applied to a pad.

But soldermask is not that important for prototype boards anyway as pretty much all common component packages solder just fine without a mask. Only exception might be advanced packages like BGAs and similar