One of the Holy Grails in the world of electronic badges has been full-colour PCB art as a logical progression from the limited palette of traditional PCB artwork. In the progression towards achieving this goal there have been a variety of techniques applied, and it’s become an expensive commercial possibility rather than an unachievable dream. Has it become practical for mere mortals then? [Tom Clement] has put together a write-up of his progression towards achieving full-colour PCB artwork on a limited production run , and it makes for a fascinating read.
The board in question is the Pixel badge, an improved commercial version of the CampZone 2019 I-Pane badge we reviewed last year. It’s a very bright large multicolour LED matrix that has caught the eye of campgoers at events ever since the original, and has generated enough demand for a new production run. As well as a few electronic enhancements it replaces the original’s dithered monochrome silkscreen rear art with a full-colour design, and it is that with which the write-up is concerned.
It starts with UV printing, and goes through the various iterations of the process until a satisfactory result is achieved. We learn about the effect of reflow temperatures on UV printing inks, it seems that white ink discolours with temperature and the inventive solution is to transfer all the whites to the PCB silkscreen layer. He closes with a discourse on alignment, and we start to appreciate the achievement behind producing this badge. A colour print isn’t necessary for the Pixel’s eye-searing light show, but the point of badges is as much to show off the cutting edge of the art.
As the regular Hackaday reader will know, colour PCB art is a long-time topic of interest here at Hackaday.
Quite apt timing, I just randomly came across this silkscreened board design
https://www.amazon.com/ElectroCookie-Leonardo-ATmega32u4-Arduino-Projects/dp/B07W8MKP2H
ah yes, than explains the arduinos with breasts that i have seen advertised recently.
Back when I did digital imaging and prepress related work for custom surfaces and media I would use my own test colour grids which were basically quantised gradients of different combinations of the process colours. Once that was printed on the new medium I’d run an optical densitometer over each square and note the RGB values, these were then checked against the values in the original art and a colour space transform was calculated from that. Sometimes you would have weird nonlinear effects and a few iterations were required but eventually you could map all parts of a pure colour space to what was possible in the target medium. These days the process is almost automatic if you have the right software and a good digital camera.
Before you rush to get one of those 1-2k $ aliexpress hacked epson UV printers- don’t. They are absolute garbage. They will rarely work, the quality will be sketchy to say the least and the ink will destroy the machine more often than you’ll be using it. Ask me how I know it.
How do you know this
OK now I got to… “How do you know this?” :-D