NOTE: Your laser printer was designed to have laser printer paper run through it. A printer will have some latitude in the type of material that can be run through it. Your specific laser printer may not be able to handle any of the processes described here. You should understand that you very well might completely destroy your printer trying this.
Any circuit board design must take into consideration the method of circuit board manufacture. If you are going to use a resist pen, you are not going to want to design in a 1100 pin BGA. Of the direct resist methods, pcb toner transfer has the highest resolution, and is capable of producing very complex circuit boards. This page attempts to describe ways to take maximum advantage of the process.
PCB Toner Transfer Method
Each manufacturer and model of printer requires a specific toner recipe. Toner is an aggregate made from styrene or styrene and acrylic, with pigments for color. The toner is fused to the paper with heat, generally from an infrared source. The temperature at which the toner fuses to the paper varies with the toner formula, but generally speaking, at a temperature of approximately 110°C (230°F) the toner is melting, and by 130°C (266°F) the toner has become as liquid as it will ever be. Beyond that the viscosity increases and the toner becomes less liquid, eventually becoming a solid again.
The goal of PCB toner transfer in circuit board making is to convince the toner to "un-fuse" from the paper and fuse with the surface of the circuit board. Ideally, all of the toner would leave the paper and fuse to the circuit board, and would do so without running, smearing, or flaking. In practice it is more likely that most of the toner will leave the paper with an acceptable amount of smearing and flaking. Flaking can be compensated for by the use of an etch resist pen. Running and smearing are not practically fixable. Running typically manifests itself as hairs trailing off of the traces, while smearing appears as widened or otherwise distorted features.
Applying even heat and pressure through a heating and cooling cycle would be the best method of transferring the toner. The amount of heat and pressure required will depend on the toner, paper, and circuit board surface. With just enough pressure applied to put the toner in complete contact with the surface of the circuit board, the temperature should be increased to the point that the toner is able to fuse to the circuit board. The temperature should then be lowered to the point that the toner again solidifies. A clothes iron does not satisfy these requirements very well, yet it can work once you get skilled at using it. A note, just from observation - not from experimentation: The circuit board should be heated to the melting point of the toner. You want the toner to melt and then solidify in contact with the circuit board. It makes no difference what happens at the boundary between the toner and the transfer medium. The circuit board will fuse to one side of the toner, and the transfer medium will give up it's hold on the toner in it's own particular way. It will not be because it was melted - if it was, it would mean you have to pull the transfer medium off of the circuit board while the toner is still melted! That would not be very practical.
After the heat and pressure have been removed you have a sandwich of paper, toner, and circuit board. Which surface the toner is fused with more strongly depends on the transfer medium. If the solid toner has little affinity for the transfer paper, when the paper is removed the toner will remain attached to the circuit board, and we will have nearly perfect transfer. If the toner has a higher affinity for the paper than the circuit board, it might pull the toner off of the circuit board if we just pull it off. A means of removing the paper from the toner will be required. A good solution to this problem is to either have a surface that can be completely dissolved in water (gummed paper), or a paper surface that is easily removed from it's substrate (transfer film). Both are available from retailers on the web. Barring one of these, the paper itself must be soaked and removed, generally leaving the paper that is mixed with toner. The remaining paper has no detrimental effect on the process - it just looks bad. It may actually help by slowing the progress of the etchant through any pits or cracks in the toner.
Another alternative is a transfer paper to which toner can be fused, but which will release the toner with very little force (less than required to break the toner-circuit board bond). This would include some silicone release backing for adhesive labels (Avery laser printer labels). You can also make a release sheet from regular printer paper and a silicone windshield sealer like Permatex® Flowable Silicone Windshield and Glass Sealer, using a stiff squeegee to force the silicone into the paper's fibers, leaving the surface as clear of sealant as possible.
Heat and Pressure
There is an ideal temperature where the toner gets tacky enough to stick to the copper, and doesn't smear with pressure. Above that, the toner becomes liquid, and either runs or smears under pressure. Below that temperature the toner refuses to adhere to the circuit board, regardless of the pressure applied. And of course that temperature is likely different for every toner formulation.
It helps to remember that the side of the toner that you are trying to fuse is not the side with the transfer paper, but the side in contact with the circuit board. Heating the circuit board to the correct temperature and then pressing the transfer image onto it would be the ideal, especially if the act of pressing the image on reduced the temperature by a few degrees, fixing the toner onto the circuit board. Of course, if you are relying on the toner to liquefy in order to let go of the transfer paper, you will need to heat it as well.
The other day someone said something that reminded me of my Foreman grill, the one with the removable griddles, timer, and thermostat. Inside are two flat heating elements - one top and one bottom. You set the temperature and time, and it beeps when either one is met. It does not automatically shut off the heater when the timer goes off, but it does regulate the temperature.
Cleanliness
The circuit board must be clean. It should be clear that the tacky toner is not going to stick to a surface coated with hot oil from your fingerprints, so scrub the circuit board well. The copper oxide coating is porous, and easily absorbs oil, so make sure the oxide is entirely removed from the circuit board just before starting.
Example
In the image of the circuit board, below, you can see a little bit of smearing on the left side, where the blue outlined area is. I didn't tape or otherwise hold the paper on the circuit board. I set it on the board and began ironing. This is the first pass at this method (silicone impregnated paper), and I am pleased with how well it worked. Those are 10 mil traces, for reference.
There was one pinhole - smaller than the trace - where outlined in green above , that I corrected with a pen. There was one pinhole that I missed, outlined in green below, after etching. This is where pcb toner film comes in.
The amount of work required is proportional to the area of the image. Be prepared to lean on the iron for several minutes to get the toner to transfer. Also, don't expect large areas of toner to be without blemish. The transfer may be complete, but your printer probably isn't that great at making uniform large black areas.
A Little Research
I initially did the transfer from the silicone impregnated paper the same way I did with plain paper. Lots of heat, lots of pressure. The results were a little smeared, and I wanted to find out if there was a way to do it without smearing. To that end I came up with a little target.
The resolution of the target is 600 x 600 dpi. The maximum resolution of my laser printer is 1200 x 600 dpi. I printed it out and used the monitor as a light table.
The first test of reproducing this on copper was to use "normal" pressure and heat, by letting the iron heat until it reached temp while lying on the circuit board. After the iron was at temp, I put the design face down and put the iron on it, moving the iron while applying pressure. This is the result:
There is some smearing, outlined in red above, and a bubble during etching left a short circuit, outlined in white. Overall the resolution test showed that with no special steps taken the resolution of the circuit board could be as high as the CAD package would generate.
PCB Toner Transfer Conclusions
PCB toner transfer looks like a good way to go, if I can get the last few bugs worked out:
It smears due to lateral motion of the iron squezing out the molten toner.
It does not adhere uniformly to the PC board material, because the PC board temperature is not uniform.
The iron delivers unpredictable temperature and pressure.
Ideally the temperature of the PC board would be exactly and uniformly the melting point of the toner, the pressure would be applied directly down, also uniformly, on the entire surface of the toner transfer paper. These are not things that are challenges to someone who has mastered using an iron, or a laminator. They are problems a novice would have when starting out with this process.
