The proper name for the laminator is Canon F21680 Fuser Unit. Following
Don Lancaster's "Hardware Hacker" column in the November 1988 "Radio-Electronics"
magazine, I added a temperature adjustment control to the unit. On an arbitrary 1
to 10 scale on the knob, "0" corresponds to a measured 75°C, "5" to 110°C,
and "10" to 130°C. After experimenting with different settings, I ended up leaving
the knob at about 4-1/2, which is a little over 100°C. Basically, adding the
adjustment wasn't necessary.
The window portions of the new exposure frame (which has a clear aperture of
about 10" square) are made from 1/8" UVT (ultraviolet transmissive) Plexiglas,
and the black acrylic frame pieces are hollow to act as the vacuum reservoir.
The frame and window are glued together. The recess in the frame where I
glued the window is not quite as deep as the window is thick, so that it won't
restrict the window seating down onto the gasket in the other half. The upper
and lower frame/window assemblies are the same, except for the gasket recess
in the lower half, which also has the vacuum inlet fitting.
I cut the gasket from a sheet of 1/8" thick sheet (I don't remember if it is silicone or neoprene, but it is pretty soft--check a good hardware store). Its recess is 1/16" deep, so 1/16" sticks out above the window surface. But it can compress down to accommodate the thickness of brass and film (which is never as thick as 1/16"). I didn't glue the gasket in place, but just lay it in its recess. That makes it easier to take out and clean when/if it gets dusty.
The vacuum inlet/valve is a brass quick-connect fitting that Small Parts Inc. sells. The socket part attaches in a 10-32 tapped hole in the frame wall. The male part has a hose barb that goes in the end of the vacuum hose. The quick-connect does a good job of sealing the vacuum in, and is easier to deal with than the plastic stopcock I originally used.
The current vacuum frame doesn't use any screws. After I turn on the vacuum pump, I just press down on the upper frame piece a little until the gasket makes contact all around, and then it self-seals and the vacuum sucks down tight.
The vacuum source is the inlet side of a Gast laboratory air pump that I bought for $25 at an electronics surplus store. It is rated for a vacuum of 25.5" of mercury. If the gasket is nice and clean, I can just run the pump to draw the initial vacuum and then switch it off, but for security I usually let the pump run for the whole exposure.
There are two main requirements for a light source for exposing the photoresist:
it must be collimated enough (i.e. the light rays must be parallel) to avoid
optically undercutting the artwork, and it must be intense enough to provide an
adequate exposure before the catalyzing reaction spreads on its own, "fuzzing"
the edges of the exposed photoresist.
Sunlight is admirably suited in both these criteria. It has a collimation angle of about 1/2 degree, and it is intense enough (at my latitude of 42°) to provide the proper exposure in roughly half a minute. It is also very affordable. On the other hand, sunlight has several disadvantages. It is of variable intensity due to sky conditions, season and time of day. It is also only available outdoors and generally only during working hours, effectively limiting exposures to the weekend.
I tried several alternate light sources, including incandescent, fluorescent and quartz halogen lamps. None had anywhwhere near the intensity or quality of UV light necessary for making proper exposures. Then I ran across Think & Tinker's website. This company deals in equipment and supplies for small-scale printed circuit board production. They offer a kit of UV fluorescent light bulbs with associated ballasts and starters, and materials to make reflectors and a grid to screen out the worst-collimated portion of the light given off by the tubes. I bought a kit, but it was several months before I had time to assemble it into a light source.
The frame of the unit is aluminum, with a 1/4" thick baseplate, 1" diameter struts, and 1/8" thick top piece, which supports the electrical components. The grid hangs below the lights on a PVC rod strut at each corner. I covered the baseplate with black velvet cloth to minimize light reflections.
I expose one side of the workpiece at a time, and use an exposure time of 2'20", which is very repeatable and reliable. I have even laminated, exposed, and etched in one marathon 4-hour session. (I don't do that too often; I usually clean and laminate one evening, and expose and etch the next.)
The new tank is a piece of 18" diameter, .2" wall PVC irrigation pipe I bought from a plastic pipe vendor. I had to call around and drive a bit to get it, since most plumbing stores only carry up to 4 or 6" diameter. PVC is a fairly soft plastic, and I used a woodworking handsaw and files for most of the fabrication. All gluing is done with PVC pipe cement and primer.
I cut the bottom end of the pipe to form a sump to hold two gallons of etchant, and closed it off with 1/4" PVC sheet, all glued together with PVC pipe cement. I cut the top end of the pipe at 25 degrees to slope the lid, which is 1/4" clear acrylic. To support the lid, I cut a 1" wide ring from the top edge of the tank, cut a small piece from this ring to reduce its circumference, and glued it about an inch below the top edge of the tank rim.
The reason for the large sump capacity is partly for a larger heated reservoir, but mostly to reduce the rate of contamination of the etchant. The ferric chloride seems to load up more quickly in etching brass vs. copper, and I wanted to be able to do a fair amount of etching before draining the tank, cleaning out and reloading.
For the new tank, I bought a diaphragm pump which put out 45 psi at almost
3 gallons/minute, and two plastic nozzles rated at .60 gal/min at 40 psi. At
that pressure, the pump had an excess of capacity, and would trip its
overpressure switch almost continuously. Luckily, another engineer with whom I
work has worked with pumps a lot and showed me how to plumb in a bypass valve
to let me adjust the flow through the nozzles. I could start the pump with the
bypass valve full open, which was useful for circulating the etchant as it was
heating up, and then close the valve until the overpressure switch started
tripping, then back off a bit.
I initially used 1/2" reinforced clear PVC tubing from the hardware store (it is clear, and you can see the white reinforcing threads in the tube wall), plastic barb fittings, plastic snap-on hose clamps, an all-PVC ball valve from the hardware store for the bypass valve. For the fittings in the tank wall, I cut 1" diameter pieces of acrylic rod, drilled through and tapped them for the appropriate pipe threads for the fittings (the nozzles are threaded 1/8NPT, and the fittings at the sump are pipe thread to hose barb elbows), and glued them in holes drilled through the tank wall.
I am using a 300W aquarium heater, built like a long test tube with the power cord and temperature adjustment knob coming out a plastic cap on the end. Although it is completely submersible, I did not want to take chances. I drilled a hole in the side of the tank (adjacent to the bypass outlet) and used silicone caulk to seal the heater in, with just the plastic head outside the tank. A small plastic bracket glued onto the inner wall of the sump supports the other end of the heater. It is just a 1/4" thick piece of acrylic with a hole in it.
The new tank worked very well, but I had two failures where the outlet tubing
slipped off the barbed fittings, even though I was using the nylon "snappy" hose clamps.
I replaced all the outlet plumbing with 1/2" PVC pipe, glued. I turned up an
adapter from PVC barstock that plugs into the pump outlet port (duplicating the
stock adapter) and has a spud to glue into the PVC fitting.
The new plumbing has been working without a failure, but I still keep the tank wrapped in a piece of plastic tarp "just in case", so I don't have another episode of ferric chloride spraying around the garage.
Another change is that I'm using larger nozzles that match the flow of the pump pretty well, so I don't use a bypass valve any more, and the etching time has reduced to about 40 seconds/thou of etching (10-thou brass will etch through in about 6 minutes, double-sided). The new nozzles are rated at 1.20 GPM at 40 psi. Since they are 1/4" pipe thread, I was able to use off-the-shelf PVC bushings to mount them.
This article is ©1998-2007 Randy Gordon-Gilmore. No part of this article may be republished without permission. Private copying for non-profit-making purposes is permitted, as long as credit is given. Last updated: 26 Feb 2007.