Printed Circuit Boards Using Toner Transfer
Toner Transfer is a fairly easy way to make printed circuit boards. Existing art work from magazines, CAD programs (I use the freeware Lite version of EAGLE CAD), or even a graphics program such as MS Paint can be used. A new one that I like is PaintDotNet, freeware developed by Washington State University. A big advantage is its use of layers, like the high-end Adobe Photoshop program.
I typically develop a board design using EAGLE, and either print it directly or export it as a .bmp file for editing in a graphics program. By doing that, I add text, thicken lines when needed, and often I'll do some selective fill in ground plane areas. It is certainly not necessary to do this, but I just like to do a bit of final editing.
When I first started making PC boards using the toner transfer method (ironing an image onto a PC board, as opposed to using presensitized boards and photographic methods), I used a commercial product, Press N Peel Blue. Lately, I've been using an alternative media, which is free. More on that later. The artwork is used as a reversed image and either printed directly on the Press and Peel Blue sheets using a 300 DPI laser printer, or printed on paper using an inkject and then photocopied onto the material. Laser and copier toner is plastic based. The Press and Peel Blue sheet containing the image is then ironed onto a cleaned and prepped (steel wool or abrasive pads) blank printed circuit board. One thing I've found out through experience is that the steam holes in a typical iron tend to initially produce some wrinkling, especially around the edges of the board. A good solution would be to find an old-fashioned steamless iron at a GoodWill store. One member of the Yahoo Homebrew_PCBs group modified his iron by cutting an aluminum plate to size and attaching it to the bottom of the iron using JB Weld, a metallized epoxy. An advantage of Press N Peel Blue over some alternative media to be discussed later is that it deposits a hard layer of plastic over the toner image on the board, acting as additional resist and generally producing sharp lines.
With the toner transfer process, it is fairly common to produce boards with small patches where the material did not adhere, probably due to the board not being properly prepped in those areas, or localized uneven heat transfer while ironing. These areas can easily be touched up by going over them with a permanent marking pen, such as a Sharpie, or with lacquer-based fingernail polish. In the first photo, you can see areas where I've applied a contrasting color of fingernail polish to a couple of the boards. While making the board for a 40 meter transmitter recently, even after applying fingernail polish to a couple of areas where the blue material didn't stick, the polish came off during the etching process, indicating that the chief culprit of gaps is inadequate preparation of the blank PC board stock.
After the boards have been etched in ferric chloride (available at Fry's, and some Radio Shack stores, locally) or other circuit board etchant, the Press and Peel Blue has to be removed from the copper foil. During this step, it becomes evident just how tightly bonded the plastic fron the toner and Press and Peel Blue have become to the copper. Prepare for a few minutes of strenuous excercise with a Scotchbright ("greenie") pad. Using some acetone will hasten the process tremendously. Hint: try it first without the acetone, and you will discover just how tough the Press N Peel layer over the copper traces is.
I've been ordering Press and Peel Blue from Electronix Express (http://www.elexp.com/). Look under "Prototyping" and then "PC Board Development" on the site. They also sell other alternative products for PC board-making, as does Mouser and Digi-Key.
Some boards done using Press and Peel Blue are shown below.
Click on the photos for a higher
Several project boards
The board in the center of the group is the Belthorn SSB IF Module. You can see lighter areas around the pads; this is a double-sided board, the first one that I've done. The top side of the board is a ground plane, with etched circular areas around the component leads. Above and to the left is an un-etched version of the same board.
Closeup of the Belthorn board
If you look closely, you can see that the blue material is still on the printed circuit traces. It looks like the board traces are irregular, but they're not. I traced over the Press and Peel Blue material with a fine line permanent marker in most of the areas, which was probably not necessary. The irregular appearance is due to the contrast between the darker marked lines and the underlying blue material. After etching, the dark blue material fades to a lighter color, which can be seen in this photo and the first one. To avoid tarnishing of the copper traces prior to soldering the components on the board, I leave the Press and Peel Blue material on the board until ready for tin plating. Then, I remove the Press N Peel with acetone and the Scotchbright pad. Plating the copper preserves it and improves solderability.
Palm Serial Sender
This board is one I made using Eagle from an article and schematic by David Eck, NK0E. It is an interface between a common Palm PDA and a transceiver, such as my SW20+, that provides contest logging and transmitter keying features from the PDA. Mine is a PalmIIIXE, although I have one of the original Palm Pilots that I might dedicate to this use. The board is a good example of the fairly fine lines that can be produced using the Press and Peel Blue process. There is only one very small section that I'll have to bridge, in the upper right corner in this photograph, when I solder the components to the board.
Example of a Tin-plated Board
This board has been tin-plated using the Tin-It product, which is basically a stove-top method. You clean the board to remove any lacquer or other contaminants, and put it in the solution for a few minutes at 130 to 140 degrees F. The board is for a 30 meter superhet receiver based in part on at least three of Dave Benson, K1SWL's designs (note the credit etched into the board).
In the Spring of 2006, I bought a new laser printer specifically to eliminate the step of using a copier to make the toner transfer image. As it turned out, my Brother HL-2040 uses higher temperatures to fuse the carbon/plastic mixture onto the printed medium. I found that my initial results using Press N Peel blue were awful.
Below is a board I attempted to produce using Press N Peel imaged from my new laser printer. I tried to compensate by applying more pressure. As you can see, the image transferred, but it smeared the plastic/carbon toner mix such the board was unusable.
I then consulted with the Homebrew_PCBs group to find alternative media and/or methods. After trying several types of paper, including plain clay-coated inkjet paper, glossy inkjet paper, and several other types, I achieved my best results using paper cut from the mailed 8" x 10" version of the Harbor Freight tools catalog. Others have reported good results with Time magazine, etc, but the Harbor Freight catalog seems to have just the right amount of semi-gloss surface texture on the paper for my printer. Following are scans of some boards I've done recently (July/August 2006) using the new and very inexpensive method:
The image above (shot with a digital camera while the board was still wet) is an RF amplifier based on a design by Wayne McFee, NB6M. It was the first board I made using the catalog paper. There are a couple of defect areas that needed retouching. You will notice the white fuzziness over the black ground plane areas...this is residual fibers from the paper (one disadvantage over the plastic media of Press N Peel Blue). It is not necessary to remove all of this fuzzy stuff; only in the copper areas that will be etched away.
This board is a good clean board, the first one that I was really pleased with using the new method. You can see a few places where I had to rout out the traces, due to extraneous resist material still on the board while etching. I'll have to repair one of the traces. That notwithstanding, it is an acceptable board. It is a sidetone oscillator from EMRFD Chapter 1, combined with a relay to do some switching of the audio.
This board is a 4-band low pass filter board, switched by eight relays. It is to be used in a linear amplifier based on the WA2EBY design that was in QST awhile back. The filters for 80, 40, and 20 meters are borrowed from the CDG2000 transceiver design, and I synthesized a similar filter for 30 meters using the RFSIM99 freeware program.
It is a two-sided board, this being the bottom layer. With EAGLE, it is possible to flip a surface mount component to the bottom layer using the "mirror" icon. I used surface mount diodes across the eight relay coils to save some space; the pads are indicated by the annotation. Notice that the "+" and"-" legends I added in Photoshop to the image are clearly visible in copper. There are two areas where I made corrections after ironing the image, and these may need to be beefed up by adding a component lead shunt soldered along the tracks. I drew the top side traces by hand (not shown), as they consist only of the RF paths from the RCA jacks to one pad on each relay. One thing to keep in mind when etching a two-sided board is to mask the side not currently being etched. When I was etching the amplifier board, I suddenly remembered that point about 3 minutes into the etch!
AVR Flash Programmer; board designed by Steven Bolt. Transfer using catalog paper. This board was made in January 2007. The resolution, as can be seen by the printing, is very good. The board is a simple AVR flash programmer that interfaces to the PC via the paralllel port. I built it to program the AT90S2313 AVR for the ELSIE LCF meter with Morse code output. I made the photo with my Nikon N-990, without flash rather than scanning it, and the slight image distortion from the lens is evident.
These two boards (not the same scale) are based on Hans, G0UPL's LCD frequency counter, to be used with a VFO I am building. The counter, unlike most designs currently in use, does not use a microcontroller, only logic ICs and and 7-segment LCD display driver ICs. The counter will display either a 3-digit reading kHz reading or, switch selectable, 2 digits with decimal, giving 100 Hz resolution. The readout will be kHz from the bottom edge of the band. I CADded the counter board based on Hans' schematic, and I did the display board based on the Luminex display's data sheet. The total cost of parts for the counter, including the display, is just under $10. These boards are to be two-sided; first I will have to etch the side shown in the image, then drill registration alignment holes in a few key pads used by both the top and bottom layers. When aligning the top sides for toner transfer ironing, I will hold the board up to a strong light, and fix the position of the catalog paper image over the alignment holes. When light is shining through all of the holes, I will then place the board down and immediately iron the image. It is a bit tricky, so obviously I try to design single-sided boards whenever practical.
For the first board, I photographed it with some paper fibers still attached in areas of the copper traces. If theses are not carefully rubbed off (a wet finger works well), the individual fibers are likely to interfere with the etching process. The result will be wisps of copper in the areas between traces, possibly creating shorts between circuits. Always do a high-magnification final check of the board prior to etching. I use an inexpensive head-mounted flip-down magnifier, and sometimes even scan the board at 300 DPI and compare it to the artwork. The fibers are one disadvantage of paper-based media; another is that the traces (black plastic-carbon toner mix) are not as durable as those produced by Press N Peel Blue, with its plastic top layer deposited on top of the toner.
The second of the two boards is the display board; it shows the degree of fineness possible with toner transfer. The pads on the board are spaced .1 inch apart, and you can see that at the top of the board, I have up to six parallel traces of .014" width. This was my first experiment with such small traces and close spacing. The other board I've left for illustration partially cleaned. After an image is tranferred, the board is soaked in water. After awhile, the paper is soggy enough so that it can be carefully rolled off the board in sheets. This process is repeated until all of the paper stuck to bare copper is rubbed off. If rubbed too aggressively, the traces will also gradually rub away. This can be seen for examining the bottom row of pads on the display board. Those will have to be touched up by pen.
Above are closeups of sections of the two boards after retouching. Typically, this involves finding voids in the toner and filling it in using a permanent marker pen. If the pen colors too far outside the lines, so to speak, the excess marking can be simply scraped away. You may be able to see some scrape marks in the copper where I've done this. While prepping these these boards, I couldn't find my exacto knife, so I used the point of the smallest screwdriver in one of those ubiquitous 5-packs. As I mentioned before, I have one of those inexpensive head-mounted magnifiers, and I use that to get in really close. Also, the toner tranfer process may have merged some pads with tracks where it was not intended. Referring to the original artwork can clear up any ambiguities. You can also see that the pad holes are not necessarily perfectly round after retouching. These holes simply serve as a Dremel tool drill bit positioning point. By the way, I have an older Dremel tool mounted in an older model Dremel drill press. I put a softwood block on the drill table, and the PC board on top of that, and drill by raising and lowerering the table with its rack-and-pinion knob. The newer model Dremel drill press rig moves the tool up and down. For holes that have been filled in by copper after etching, I use a $2.25 spring-loaded punch I got from Harbor Freight on sale to make a good impression for the bit point.
Here are the two boards after etching. No surprises; however, I'll take a close look at them under magnification, and I'll flow solder on any tracks that are not very solid. An alternative is to apply self-stick copper circuit repair tape (soldering the ends), which is available in 1/16" and 1/8" widths locally from Fry's. Or, simply solder some approriately sized solid copper wire over the defect.
The above image was scanned at 600 DPI to see how the critical .014" paralled traces look after etching. There don't appear to be any shorts between the traces, but I'll check them with a DVM prior to assembly of the display board.
For comparison of the above pictures of the display board, here is the original art, flipped horizontally for normal reading.
The final picture is the pen I usually use for fine retouching.
The toner transfer method of making PC boards is simple and inexpensive. For more detail, refer to Jim Larsen, AL7FS's presentation on the subject, an amended copy of which I have here on my web site (I added a few pages at the end for a presentation at a NoGa club meeting). The presentation is a large file, several MegaBytes.