A Pictorial Tour of PCB Manufacturing with Joey, Chad, and Jacky!

Beginning on April 24th and extending in to the wee hours of April 25th, Chad Weider, Jacky Leung, and myself went on a journey where we’d meet caustic chemicals, raisined fingers, and shiny things. The goal was to make the printed circuit boards for the Position Detector (PDs) in the infamous galvanometer project. The PDs consist of a stationary plate and rotor. The stationary plate is on standard thickness copper clad board and square shaped, with four smaller squares of copper on the surface. The rotor is smaller, and circular with two opposing corners copper, the other two blank. The rotor is on super-thin PCB, not for it’s flexible property, but instead for it’s low mass, do minimize rotational inertia when it is moving on the galvo’s axle.

With guidance from Easy Printed Circuit Board Fabrication we began, making modifications to the process on the way. I shall walk through the process step-by-step inside the post, for there are too many photos for the front page. If instead you’d rather jump straight to the pictures, check out the album immediately below.

I also have CAD drawings of both the Stationary Plate and Rotor in PDF format available for download.

The Process

The project included creating two PCBs sequentially, first the stationary plate, then the rotor disc. For illustrative purposes, I will use pictures from each interchangeably to highlight the best pictures from each board.

The first step involved preparing the transfer drawings. I used Adobe Illustrator to create the proper masks, then the venerable ACM front office printer, Scribble, to print on to some Office Depot brand 4*5 photo paper. I had some trouble with halftoning in the black, but seemingly every other page I printed came out dark and solid. I blame ColorSync for the troubles.

Once the masks were printed, I went on to prepare the Copper Clad board by cleaning it (using some alcohol), cutting the thin PCB to size, and placing the mask on top. I procured the regular stuff from Fry’s and the thin copper clad board from Surplus Sales of Nebraska (the same place I bought my magnet wire), Part# (PCB) N4105.

After placing a paper towel on the back of the paper, the boards required much ironing. Turning the iron all the way up seemed to completely melt the paper, so 75% power for as long as 10 minutes seemed to get the job done quite well. This was probably somewhat overkill, but worked well. While the board was still hot, we then threw it in to a tray filled with water to hear wonderful sizzling sounds. Hot water may have worked better to melt away the paper, but warm allowed us to reach in to the tray and agitate.

Next we rubbed off the paper backing on the photo paper. The knife allowed for the first layer to be pulled off, then finger friction to rub down to the very thin plastic-y layer.

The final layer of paper was more like plastic than paper. It was somewhat melted to the copper-clad board, but not in the same way the toner actually bonded to the board, allowing us to peal it off. This was really quite fun, but somewhat scary, because it seemed like we were going to rip the toner up from the copper. Fortunately it was bonded quite well, but we pealed very slowly at an extremely low angle to get best results.

After a little cleanup of pieces of the plastic-y layer that didn’t quite come off, we were left with a board ready for etching; we were impressed at the quality of this kind of hacked-together method. Some of the masked areas appeared as if they had tiny cracks in them, so we tried to cover up the imperfections with a Sharpie, but I’m not sure how much that did.

Next we drilled a hole in the board so we could suspend it in the solution of Ferric Chloride. Ferric Chloride is nasty stuff, so keep metals away. The green container holds the solution vertically, and contains a heater which speeds the etching reaction. A fish-tank style bubbler agitates the solution again, helping in the etching. The first regular thickness board took almost exactly 15 minutes, the thin PCB took longer because it was double sided and we needed to completely remove the back side, which wasn’t cleaned well, impeding etching progress.

Finally we were able to remove the board from the etching solution and reveal the etched pattern. It turned out pretty well, considering we didn’t buy any special paper or anything. Once rinsed, Acetone easily removed the toner mask from the board, revealing the beautiful lustre of the copper. It unfortunately stained the thin PCB with black toner, but shouldn’t functionally be a problem. That acetone step was performed outside due to the terrible fumes generated. I probably should have worn gloves though…

(You’ll notice a sneak peek of the coils I’m working on in that last picture. More to come about those soon.)

Over all it was fun, and I’d do it again, at least a couple of times. I look forward to making my own test PCBs, but I think I’m going to get the amplifier boards for the galvos done professionally–maybe I just like that green color…

*edit*
Thank you Chad, for taking all the photos documenting the process. Without your help, this would have been a really bad pictorial step-by-step guide, for there would have been no pictures.