I have been doing a lot of circuit prototyping lately and have had to make printed circuit boards (PCBs). In order to save money and time, I started making them in-house, although it took several weeks of frustration, the purchase of three different systems, and some tinkering before I arrived at a method that worked for me. There are already numerous guides on how to make PCBs at home, but the purpose of this is to compare three popular systems and evaluate their claims to (hopefully) save you some trouble. Each of the three systems has strengths and weaknesses–you must simply know when to use each.
The first system I tried this year was Pulsar’s “PCB Fab in a Box”. It was attractive because they claimed to offer a selection of toner transfer paper, foils, copper clad boards, and laminating machine that worked together flawlessly to quickly produce excellent, repeatable results. Unfortunately, despite carefully following the directions and using an approved printer, I got poor results.
The first image shows that the traces simply would not adhere to the copper board. I was finally able to achieve some success by “tacking down” the paper into place with a regular iron before running it through the laminator, and then running it through a total of 6 times, rather than the recommended 2. The next image shows the toner adhering to the board much better than before. I think the added heat helped. Unfortunately, you cannot stop at this point and etch because toner is very porous and the etching solution will still eat at the copper beneath unless you seal it. Pulsar’s solution for this is their green foil which “seals” the toner. Unfortunately, after applying the foil with the laminator, it didn’t stick everywhere and in some areas, it actually pulled up a few pads and traces from the board. Overall, the resulting board was not useful. Perhaps my little HP laser jet P1006 is the problem, but testing with a large Xerox “mainframe” office printer offered similarly poor results.
Next I returned to the Techniks Press-n-Peel system, which I had used when I first made some PCBs around 2001. It worked reasonably well with only occasional breaks in traces. Unfortunately, a clothing iron was the only solution at the time to apply heat and pressure. However, with a fancy new laminator from the Pulsar kit in my possession, I decided to try the PnP system again with the hope the laminator would bring about consistency. Pulsar claims on their website the laminator is not compatible with the Press-n-Peel system, however this is not true. I got very good results by simply running the board through the laminator until the black artwork could be seen through the blue coating. This turned out to be 5-6 times. The best part is that after peeling the substrate away, no additional work is necessary. The blue coating on the plastic sheet sticks to the toner and accomplishes what Pulsar’s “green foil” does, but in one step. I found the PnP system with the laminator to be excellent for quick, simple, one-sided boards that did not require especially fine resolution. After attempting to create artwork for a TSSOP (a surface mount chip with pins just 0.6mm apart, the artwork was a bit too fuzzy.
The limitations of the Technics system led me to explore photo-based development–the use of photosensitive boards to create the etching mask. I avoided this route for some time as presensitized boards can be costly due to the additional coating. Furthermore, if you only make a one-sided board, an entire side goes to waste. After discovering the MG Chemicals Negative Dry Film Resist, however, I decided to try a photo process. The beauty of the dry resist film is that you only apply it when and where you need it, which allows you to buy any kind of copper clad board you wish–double/single sided, varying thicknesses, etc. The instructions for its application did not work as expected. I kept getting bubbles when I tried to apply the film to the board, and warping after running it through the laminator. With some experimentation, I was able to apply it perfectly every time.
Unwritten directions for applying the dry resist film: First, lay the board on top of the film and cut around it with an exacto knife–no need to leave several centimeters around the edge as the MG directions say. Next (and this is the key), after you peel the soft coating, place a few droplets of water on the copper board before applying the film. The water helps the film to adhere to the board evenly. Center the film, pulling it up and repositioning if necessary. Lastly, squeegee the water out working from the center outwards. I used a small, rectangular piece of nylon with a hard edge to do the trick (you can use the edge of a credit card). Work out all the water and bubbles. Lastly do not use the laminator as recommended (unless you can set the heat very low). I used a clothing iron and put it on the acrylic setting (the lowest it would go). Placing a sheet of printer paper between the iron and the board covered in film, I used moderate pressure for 30-60 seconds. This was adequate to adhere the film to the board. Absolutely no warping or bubbling occurred. You will know you’ve used too much heat if the surface is not even. The next image shows a laminate applied with too much heat on the iron.
After the film has been applied, the rest of the process is straightforward. Directions are on the MG Chem website. With the photo-based system, I was able to get incredibly high resolution artwork. The following image shows artwork for two TSSOP (0.65mm pitch) chips being developed under a standard broad-spectrum (also sold as “daylight” spectrum) compact florescent (CFL) bulb.
||MG Chem. Dry Film Resist
||transfers toner well to bare board for labels
||Simple, straightforward process (print/apply/etch); good results
||very high resolution possible; clear masks allow easier alignment for making two-sided boards
||additional step of adding green foil (assuming you can even get the system to work)
||artwork can be too fuzzy for fine-pitch SMT devices
||Need darkroom environment; applying film takes a bit of extra time; added chemical step of developing the film before etching; must invert artwork first (toner goes where copper should be removed)
||Use the laminator from the Pulsar system for added consistency
||Use water when applying film to avoid bubbles; use very low heat to avoid warping
||Use the transfer paper for silkscreening labels, part outlines, etc. to bare (i.e. copper removed) board
||Simple, single-sided projects that do not require high resolution; when you don’t have the time for the photo method
||Double-sided boards, or when high-resolution necessary
When making double-sided artwork, include four small vias at the corners of the board’s artwork. Tape each transparency to a piece of glass with the presensitized board sandwiched between. Align the four holes and the artwork should register perfectly.
If you choose to etch with Ferric Chloride, many suggest using a sponge to wipe over the surface of the board. This is to bring fresh etchant into contact with the copper. I would suggest simply wearing latex disposable gloves and gently wiping the surface with your finger–this seems to be a bit more gentle (less chance of ripping off part of the mask), and just as effective.
When using a photo process, always place the side of the transparency with the toner against the board. The acetate sheet is actually a few thousandths of an inch thick and is just enough to allow light to get in from the sides and cause fuzzy lines if the toner mask isn’t directly against the board.
Cleaning the board thoroughly is necessary for the Press-n-Peel system, however scotch brite pads are not necessary as many suggest. I feel they are too abrasive and leave large scratches. I have found dish soap, followed by comet cleaner to be effective. If you don’t use too much water with the comet, leaving it as a moist paste, it will be sufficiently abrasive.
Use a magnifying eye loupe to check the toner prints, masks, and etched boards for problem areas, especially when working with fine-pitch surface mount artwork.