PCB Press
How many ways are there to make your own PCBs? Let me count the ways... Well, actually, there a quite a few - thermal transfer from a laser printer then use a household clothes iron or one of those laminate machines. Some homebrewers use the photo method, or even direct print using a modified inkjet printer. Even home built CNC machines have been built by the really adventurous homebrewer to machine away the copper. Some homebrewers simply outsource their PCB manufacturing to a company.
All of these ways to make PCBs have their pros and cons. I do like the simplicity of thermal transfer and have tried using the iron method but never really got good, repeatable results with it. Never used the laminator method. I didn't really want to invest in the photo method, especially dealing with the sensitized boards, light box, etc. Wouldn't even consider building a CNC machine, just don't have the room for one, let alone the cost involved. I needed a better way to make boards and after some brainstorming, the PCB Press was born.
Assembly
Starting with an old Black and Decker waffle iron and a sheet of 6061 aluminum, I cut the aluminum into two squares, roughly six inches on a side. I could have cut the aluminum to fit the entire griddle portion of the iron but I only had a single 12 inch sheet to work with. It's also important to use a waffle iron that has an expandable hinge. Some waffle irons have a fixed hinge and won't sit flush when the aluminum sheets are attached and the PCB board is sandwiched between.
Black and Decker waffle iron/griddle with aluminum sheets
Removing the griddle, it was back braced then clamped down for drilling. The aluminum sheet is centered before drilling and it's very important that the griddle doesn't flex from the clamping. The sheet needs to be perfectly flat on the griddle. I drilled four holes near the corners then used sheet metal screws from the back of the griddle to fasten the aluminum. Using a cutoff wheel on my Dremel Tool I trimmed off the protruding screws. It's imperative that both upper and lower griddles fit perfectly flush when closed.
Ready for drilling
Here's the assembled PCB Press. It takes about 20 minutes to reach full temperature.
Completed PCB Press
Experiments
This is where the fun begins! Figuring out the process of transferring toner to a blank PCB. The laser printer I have is a Brother HL-5370 DW, nothing fancy. Almost all the experiments were conducted with the PCB Press set to maximum temperature. I preheated the board for about a minute before putting the "transfer paper" on to the PCB blank. I set a 10 pound weight on top of the lid then start a four minute countdown timer. The same two boards were used as well as the same circuit artwork. This artwork was a PNG file so some of the traces are stepped. All boards were cooled to room temperature then soaked overnight in plain room temperature water to release the paper.
In operation
Baked PCB
Experiment #1
This experiment used regular laser printer paper with the laser printer set to defaults. I would have been shocked if it came out perfect but it was very close! Plain laser paper really held on to the toner. The toner did stick to the board and passed the finger nail scratch test. It takes a good amount of effort to scratch off the toner. The missing portions of the board in the center happened when I took a Scotch Brite pad and tried to remove the stuck on paper. It took a hard effort to clean all the toner off the board for the next experiment.
Results Experiment #1
Experiment #2
Number 2 used plain magazine paper with the laser printer at default settings.
PCB Blank and Magazine Paper
Magazine paper released better than plain laser printer paper but it still took using the Scotch Brite pad to get the paper removed from the copper. A few of the pads lifted, some of the ground pours got spotty, along with a few broken traces.
Results Experiment #2
Experiment #3
Number 3 used a thicker glossy paper from the cover page of a magazine. This did much better. I'm attributing the toner not affixed in the upper lift corner to not cleaning the board well enough. This still took the Scotch Brite pad to remove most of the paper but it was almost impossible to remove any of the paper in-between narrow traces without removing the toner. A few of the traces are broken but much less than the first experiments.
Result Experiment #3
Experiment #4 & #5
Experiment 4 used regular laser printer paper and a 5 pound weight, experiment 5 used glossy front magazine page and a 5 pound weight. Both of these were total failures. With 4, the paper did release much easier but the toner didn't stick to the PCB as well. The photo doesn't really do it justice on how bad it is. Five was not as bad but nowhere as good as experiment 3. More of the paper would come off between traces easier but toner didn't stick as well to the board. I was testing to see if I was using too much weight causing the paper to stick too hard to the toner making it difficult to release.
Experiment #4
Result Experiment #5
Experiment #6
Running out of freely available paper, I did some research and a few homebrewers had very good results using HP brochure paper. A quick trip to my local Office Cheapo and I bought a ream. Although it feels the same as the glossy magazine cover paper, the results are much better. After soaking in water for about 30 minutes, all I needed to do to remove the paper was to rub it with my thumb. The paper would nicely roll up and off the toner and the board - if the traces were wide enough. I was able to remove the paper from the toner in the very right bottom corner of the board too. The toner actually stuck over the board a few millimeters and stayed!
Result Experiment #6 - Paper Removal
I put the board back in the water for another 60 minutes to soak. I was able to get even more paper off with my thumb. This is by far the best result. A couple of the pads are light, probably due to not being cleaned well enough. By this experiment the PCB blanks were starting to show the fiberglass weave under the copper. So, I'm thinking the copper is thinning and the board is no longer a smooth surface as it would be when new.
Result Experiment #6 - Finished
Experiment #7
Number 7 used the same HP brochure paper but I reduced the baking time to 3.5 minutes. The results were rather rough. The edges could be due to not cleaning the board well enough. This really is pretty good, much less paper is sticking to the copper between the narrow spaced traces.
Result Experiment #7
Experiment #8
This experiment I started playing around with settings that would reduce how well the toner would stick to the paper. I found a setting called "reduce paper curl" and enabled that option. According to the printer manual, it slightly lowers the fuser temperature so the paper won't curl as it's passed through. I'm saying "jackpot!" on this one!! The upper left corner came off but I think that's due to the board being used so many times. Some of the text in the upper right corner did the same thing. The rest of the board looks great. Few small spots I would touch up, use a bamboo toothpick to clean any paper stuck in-between the narrow traces. Any manual cleaning should be done with the board wet. I'm thinking the white stuff left behind is some kind of clay. Not sure what affect the etchant will have on it.
Result Experiment #8
Conclusion
Why all this effort to make PCBs at home? For me, it's cool. My passion in amateur radio is building. Also I've recently obtained a copy of Experimental Methods in RF Design and I wanted to learn design by, well, experimenting. Starting from the beginning with schematic capture, PCB layout and manufacturing, building up a real circuit, then testing it out. There's no better way to learn electronics theory and fabrication than to build something with your own hands and see how it works with your own test equipment. It's even better when the circuit doesn't work the first time it's powered up! A broken circuit will require figuring out why it's broken and what it'll take to fix so it does work. No amount of book reading can compare to messing around and letting the smoke out of things! :)
I think the PCB Press is an excellent approach to making PCBs using the toner transfer method. It's capable of very consistent temperature, especially having the additional aluminum sheets firmly attached. It has very consistent pressure on the toner transfer paper and since the pressure is even across the entire surface of the board, the toner is pancake flat over the entire surface. There's no way to get bubbles under the transfer paper. The PCB is heated from both sides of the board so it stays at a constant temperature across the board. The board doesn't move so there's no chance of the transfer paper slipping. Theoretically, I could make 6"x6" double-sided PCB with this press, and up to about 8"x8" if I made the aluminum sheets as large as the griddle surface. I haven't worked on a process to make double-sided boards, that'll be a future update.
There you have it! Dig out that old Black & Decker waffle iron you haven't used in 10 years, drill a few holes, and give it a try!