My first experiments were with hand agitation in a small plastic tray, using sprayed-on liquid photoresist from a local electronics supply store. I quickly found this was not satisfactory. The etching was very uneven, and by the time part outlines were etched fully through, there was severe undercutting of the resist in areas.
I then built, from acrylic plastic, a tank in which I could hang the brass sheet vertically, still submerged in the acid, with a pump to circulate the liquid quickly. (It was a centrifugal pump made for chemicals, with a magnetic coupling to the external motor.) I added an aquarium heater to keep the etchant at 40°C. This improved matters, but there was still undercutting of the resist pattern, and on relief etching, there was a very rough surface finish.
There were two basic problems. The first was in the design of the etching tank. I realized that the acid was not circulating quickly enough (even though the pump was about 4 liters/min and the tank only held about 1 liter of acid). With inadequate circulation, there is a thin surface layer of acid containing dissolved etching by-products. This "contaminated" liquid seems to be the main cause of the poor surface finish, which also slows down the etching process and requires the workpiece to be kept in the liquid longer, leading to the increased undercutting.
Another problem was with the resist itself. I was not able to get an even, thin coating, and it was even harder to get a coating without pinholes. From an article on commercial printed circuit board manufacture, I learned of dry- film resist. This is a partially-cured, but still tacky, plastic film which is laminated onto the brass using heat and pressure, and fully cures on exposure to ultraviolet light. I obtained a sample roll of this resist to try out.
I also made a tank with a higher-pressure pump, and plastic nozzles to spray the ferric chloride onto the sheet, with a clip to suspend the sheet vertically in the tank above the surface of the acid. This worked much better (about 4 times faster than the circulating tank!) and I even had to dilute the ferric chloride about 1:1 with distilled water to slow down the etching action. The surface finish was much smoother also. But, the etching was "directional" due to the acid running off the surface in one direction. I had to keep stopping the pump, and rotating the sheet 1/4 turn every so often to even out the directionality.
Which brings us to the present...
For applying the resist, I use a Canon PC Fuser I office laminating machine.
The fuser is, as far as I can tell, a fixed hollow roller containing a
halogen tube bulb (like in floor lamps) to generate the heat, and a motorized
pressure roller springloaded against it. Brass up to .020"(.5mm) is not a
problem with it, but when I tried running a .062"(1.6mm) PCB through it, it
creaked and groaned, but did successfully laminate the resist. I will not do
that again, though.
My exposure frame consists of two sheets of .06" (1.5mm) thick acrylic with
thick acrylic perimeter frames. The clear aperture is about 5" x 7"
(125 x 175 mm). Clearance holes through one perimeter frame, and tapped holes
through the other, allow me to clamp the two using thumbscrews. There is a
thin rubber gasket between the sheets along the inner edge of the perimeter
frame. Near one edge of the aperture, there is a vent hole which leads out
to one side of the perimeter frame on the "base" half. Airtight plastic
fittings and vinyl hose connect this vent to a "vacuum reservoir" made from
acrylic tube and sheet. This in turn is connected, through a stopcock, with
another fitting which allows attachment of a 60ml plastic syringe. An
expanded polystyrene "plug", covered with dark fabric, fits in the aperture
of each perimeter frame, and acts as a shutter.
My etching tank is made from a piece of 300mm diameter PVC plastic sewer pipe
(which I bought unused), and holds 4 liters of ferric chloride, diluted 1:1
with distilled water from the stock solution bought from the electronics
store. It stands about 1200mm tall, is closed permanently at the bottom, and
has a removable cover. The tank floor is clear acrylic, and is "V" shaped.
(I used clear acrylic on the theory that I could monitor the state of the
ferric chloride, which turns from a rusty brown color to greenish as it
accumulates dissolved brass. However, the liquid is too dense to pass
light...) A 150W aquarium heater near the apex of the "V" keeps the etchant
at 40°C.
There are two nozzles (equivalent to #H-83251-00 on pg. 1123 of the 1995 Cole-Parmer laboratory products catalog) in the wall of the tank directly across from one another, each fed by an oscillating pump. These pumps (equivalent to #H-07101-01 on pg. 1113 of the Cole-Parmer catalog) are made for laboratory use, and have all rubber/plastic internal parts. Unfortunately, the pumps do not put out enough pressure to completely atomize the etchant, but the spray is fine enough for the finest etching I do. I can turn each pump on individually, to fine-tune the double-sided etching. All plumbing is vinyl tubing with plastic fittings and hose clamps.
A shaft (titanium, to resist the acid, as even stainless steel will dissolve
in the ferric chloride eventually) comes in the side next to one nozzle, so I
can hold the sheet in the middle of the tank, and turn it continuously while
etching. To hold the brass while etching, I drill a ~2mm hole near each
corner. A four-armed titanium wire "spider" with hooked ends engages these
holes, and applies outward tension to keep the sheet from bowing. The
plastic hub of this spider presses onto the titanium shaft.
I use Laminar LM dry-film photoresist from Morton International. This material is a soft, .001" (.025mm) thick, light translucent blue partially- cured plastic film. It has transparent cover sheets on both sides to protect it before laminating. On exposure to ultraviolet light, it cures to a hard, dark blue coating. It is resistant to the etching acid, but the uncured resist is soluble in sodium carbonate (washing soda), a weak base, and the cured material is soluble in sodium hydroxide (lye), a much stronger base.
This avoids the mess associated with the generally recommended pumice powder, which is very hard to completely rinse off the brass.
I do all my drafting on a PC, using VersaCAD software. From the model
drawings, I develop flat patterns for the etched pieces, making allowances
for bends where required. I plot the output in Postscript. Areas to be
etched are black, while the areas to be protected are left clear.
For trial etching and non-critical pieces, I print the artwork directly onto NewsCraft LGN, a "high-tech translucent laser printer paper for negative images to use in creating printing plates", which is "specially coated to improve toner adhesion" and "dimensionally stable under heat and pressure of laser printing", using a 1200DPI (dot per inch) laser printer.
The laser printed artwork does have a slight drawback. Due to the way the paper is pulled through the printer, no two printouts are exactly alike. The misregistration between the artwork for the two sides might be .1mm over a 10cm span. I live with this for my test etchings.
For the final etches, I send the Postscript files out to be printed on film using a Linotronics phototypesetter at 2500DPI. The film artwork is always dead on.
I do all my etching from both sides, even on material as thin as .1mm. But most of my work so far has surface details. The part outlines, and window openings, are on both the front and back artworks, and detail to only be etched halfway through is on only one artwork. By the time the windows are etched fully through, the surface detail is etched halfway through.
For my etching, lining up the two artworks by eye, on a light table (or sunlit window) is sufficiently accurate. I leave a 1cm or so clear edge around the artwork, and tape the front and back pieces together along one edge, making a folder into which the brass can be placed.
I do my laminating at night, with all the lights turned out except for a yellow incandescent "bug light" in a lamp across the room as a safelight. This is probably a little darker than necessary, seeing that the commercial etcher I visited had fluorescent lights, and only a yellow-tinted plastic tent around their laminating area to filter the ultraviolet from the light. But I'd rather be safe than sorry.
The 60ml syringe is used as a vacuum pump, by opening the stopcock, drawing the syringe plunger back, closing the stopcock, detaching the syringe, pushing the plunger home, reattaching the syringe, and repeating. Six to seven cycles evacuate the exposure frame to ensure close contact between the artwork and the photoresist-coated brass sheet. The vacuum reservoir ensures that any small leaks in the gasket do not significantly affect the vacuum for at least several minutes.
I then take the exposure frame outside and use sunlight to expose the photoresist. I do all the exposing within an hour of midday. The resist has a fairly wide latitude of exposure time, but 60 seconds' exposure per side, assuming a clear day, is about right. This obviously has its limitations, requiring clear weather on a weekend, but I have yet to find an adequately intense artificial UV source that is not exorbitantly expensive. (Commercial photoetchers and PC board houses usually use arc lamps for their exposures.)
After exposing both sides, I bring the exposure frame inside, remove the brass, strip off the remaining protective clear sheet from each side, and soak the piece in a 1% solution of sodium carbonate. Within a couple of minutes, the unexposed photoresist is dissolved, and I use a soft nail brush to ensure it is all removed from the brass. I then rinse the piece with distilled water, dry and wrap in tissue paper.
I wear a Tyvek coverall, rubber gloves, and goggles when etching. The ferric chloride is not particularly dangerous, and will not harm skin if washed off promptly, but it will stain clothing fiercely, and I would not want to get any in my eye! In case of spills or splatters, a solution of oxalic acid (a mild organic acid) in water will soak out ferric chloride stains from both clothing and surfaces. It is mild enough to not harm even cotton shirts (much to the relief of my wife!) Maybe 5g acid mixed in 50ml of distilled water.
I monitor the actual etching progress through the clear acrylic cover of the tank. Etching double sided, .010" (.25mm) brass will etch through in 5-6 minutes. I turn the workpiece a quarter turn every quarter minute or so to ensure even etching, and turn off both pumps when the etching is complete.
I place the completed etching in a weak solution of sodium bicarbonate (baking soda) to neutralize any residual acid, while I etch the remaining workpieces (I usually do 5 to 10 pieces in a session).
To strip the resist, I place the workpiece in a 1-2% solution of sodium hydroxide. The resist will turn purple, swell and wrinkle, and eventually lift off the brass. The sodium hydroxide will also tarnish the brass, so I soak the now uncoated brass in vinegar with salt added (maybe a tablespoon per liter). This acts as a brightening solution, and within a minute or so, the brass is clean and shiny. A final distilled water rinse and drying, and the etching job is finished.
A second major project is a shell for Burlington's streamlined 4-6-4 Aeolus,
to go on a Kato mechanism. Other than that, I've made window frames, frames
and overlays for Commonwealth tender trucks (bogies!), footsteps, wagon-wheel
antennas, etc. In the works are a windmill, cattle pen fences and gates, and
some Burlington heavyweight passenger carsides, along with shells for a
Milwaukee F7 and NYC Commodore Vanderbilt, both to go on the Kato Hudson
chassis, for members of the Bay Area NTRAK, to which I belong. I'm sure there
are many other possibilities. I'm starting to model the late 1930's, so I
think in N scale, I could even etch cars and trucks (lorries!) to fold up,
since most road vehicles in those days were still very angular.
I have only recently rejoined the Association, and have corresponded with several members who sell etched kits. They all have been a great help in showing me the state of the art in etching (and caused me to go back and revise many of my in-process drawings!). I'd like to thank Andrew Cox, Nick Dearnaley, Ted Humber, and Tim Watson.
