One way to make etched data plates for equipment.

This is a basic outline on how I make new etched data plates for things like stationary engines or machine tools like lathes. This lets you make "Older" style data plates where there is a noticeable removal of material and typically a paint backfill of the relieved areas to give contrast. Newer data plates are made with a different method from anodized aluminium with a paint overlay and are flat with no etched features. The two methods give different results.

So in a nutshell we want to mask off areas/features we want, like letters and graphics then remove the unmasked areas with an etchant of some kind. Etchants can be a whole range of chemicals so long as they attack the base material of the data plate and not the masking artwork. Etchants have been around a long time, so there's a massive range of recipes out there, some are very toxic. However it's possible to pass a DC current through the metal to be etched to drastically speed up the etch rate without using much more dangerous acid cocktails to do this.

Cuprous chloride is a slightly weaker and slower etchant than ferric chloride. The printed circuit board industry uses it almost exclusively as it's copper based, the released copper from boards doesn't contaminate the etch solution. In fact they can add extra hydrochloric acid, aerate the solution to dissolve oxygen into it and convert the extra copper into more etchant. This is handy, no exhausted etchant to dispose of, they just make more etchant. Cuprous chloride can be neutralized with sodium hydroxide which makes it precipitate out as a solid copper II hydroxide particles while the leftover liquid is just salty water and can be safely tipped down the drain. The copper hydroxide slowly turns into black copper oxide over time which is saved to start a new batch of cuprous chloride down the track when you next want to etch something.

Cuprous chloride is a bright green liquid when it's active and ready to etch and a dark olive brown colour once it's exhausted and no longer active. You can recharge the exhausted solution by adding a small amount of hydrochloric acid and bubbling air with a fish tank aerator through the solution for a few hours until it goes bright green again, usually overnight.

Cuprous chloride etching with a DC assist doesn't produce any gas bubbles on the surface of the metals that can mask off areas and produce a stippled, pitted surface, you really have to turn up the current way too high to get gassing which will be apparent by the small yellow gas bubbles of toxic chlorine. You should see no bubbles at all unless the electric current is set way too high.

Metals that don't need DC current are:

Copper (slowly)
Brass (slowly)
Aluminium (very fast, 120 seconds for a deep etch)

Metals that won't etch without DC assist:
Stainless steel (30-60 minutes)
Carbon Steel (30-60 minutes)
Brass (quite fast, 30 minutes)

Here's a link to the website where I got up to speed on making and using cuprous chloride: http://home.exetel.com.au/adam.seychell/PCB/etching_CuCl/index.html
To make cuprous chloride, I just dilute some hydrochloric acid 50-50 with demineralized water/rain water and then throw some scrap copper in. Not a lot will happen at this point, so you need dissolved oxygen in the solution, this can be from a fish tank air bubbler or some hydrogen peroxide solution. This will have to sit for days to fully react: it starts slowly but as the cuprous chloride forms the reaction speeds up and it turns muddy green as the chlorides form then emerald green after sitting for a few days. After a while either the hydrochloric is consumed or the copper is and the reaction halts. You can remove any remaining copper and add a splash of hydrochloric acid and it will go the characteristic emerald green colour in a day or so.

I'll add a part two later on today covering artwork/masking

 

Basic safety:
If you're beginning to experiment with etching, keep in mind the safety aspect of using corrosive liquids, so at a minimum you want to be in a well ventilated area, wear gloves and safety glasses. A bucket of water with a teaspoon of washing soda/ sodium carbonate will neutralise small spills and hopefully not be too alkaline to splash in your face in emergency. I have personally been blowing pale pink snot due to inhaling several hours of the hydrochloric acid vapours from the free acid mixed in with the cuprous chloride. I now keep a cheap pedestal fan blowing fresh air a few inches above the tank and away from me. You will know if you have any droplets on your skin from the irritation that occurs in seconds. I have about 8 litres of the solution in a plastic tub which lives inside another UV resistant plastic tub outside the workshop away from sunlight and where fumes won't cause flash rusting of workshop tools. Every time I open the inner container, there are condensed green droplets of cuprous chloride on the lid from evaporation/condensation.

Etching with hydrochloric acid and cuprous chlorides will only form at most trivalent chloride compounds: chrome chloride, nickel chloride, aluminium chloride (anti-perspirant ingredient), ferric chloride (used for water purification/flocculant) and whatever. Some of these are considered carcinogenic but then so is motor oil and drinking ethanol (class 1 carcinogen). This is to ward off the Chicken Little doom predictors who think I'm making hexavalent chrome compounds (it's trivalent just like the hexavalent chrome replacement chemicals). Unless you do a lot of stainless steel, there won't be any chrome chloride in your etch tank anyway.....you know, the etch tank where you precipitate out the metals and only tip salty water down the sink.

 

Preparing artwork:

So I'm using printed circuit board negative photoresist film, this is a light sensitive polymer that gets hardened/cured by UV light exposure. You prepare a negative artwork transparency of what you want to print. I found a cheapo inkjet with cheapo refills gives a good solid black artwork on cheapo inkjet transparency film. Because it's negative film (positive film can be found) your lines and text will need to be transparent and the areas that get etched away need to be a solid black to mask light from the photoresist film.


These are original data plates for my lathe, they were loaded into my scanner and scanned, then imported into a drawing program like Corel Draw or Inkscape. The program then lets you scale the images in millimetres so that the drawing is now 1:1 scale right off the bat.





You can then draw your artwork over the original image, there are tons of free .ttf truetype fonts out there and with a lot of a searching, very close matches to the originals can be found. The originals are almost certainly drafted by hand so it's hard to just type in the words and have them line up perfectly with the scan, most need individual letter placement and scaling to match. Once you're happy with your art, the layer with the scanned originals can be deleted. At this stage we can check that it prints to scale correctly and in high resolution, most printers will fight you on this and nearly no modern printers have an "Inkjet transparency" setting so you'll have to try something like a "high resolution glossy photo paper" setting and turn up the ink density setting to compensate. PNG image format gives good crisp lines/edges and seems to have the best scaling results when set up right.



Next we create a negative artwork image and print it on the inkjet printer




At this point you're ready to coat your metal label plate with the photosensitive film.

 

Coating your label plate:

So anyone who has coated a school book with protective plastic film, knows how hard it is to get a completely dust and bubble free covering. Photoresist film is exactly the same, but there are few methods to get it on without too much stress. Photosensitive film can be bought cheaply online, it's a clear pale blue plastic film which darkens on exposure to light, so you'll be working in dim, but not pitch black lighting conditions when applying the film. It's designed to be applied to freshly cleaned copper coated fibreglass sheet and then laminated on with a hot roll laminator just like you'd use for documents. Photoresist film is a sandwich of a fronting film, the photoresist layer, a thin clear hot melt glue film and a backing film. You peel the backing off first to apply it to the metal surface, the backing layer sheet is matte in finish and the fronting sheet is glossy in appearance, a piece of masking tape helps pull the matte clear backing off at this point.

Your metal data plate blank must be a completely fresh, clean, metal surface, a scotchbrite pad scrub and laundry soap works great. Aluminium should be coated with photoresist immediately after cleaning to minimise oxide forming on the surface to ensure best adhesion of the film.

Now the non manufacturer recommended method to coat an object that can't be fed into a hot laminator is by floating the film on the surface of a container of water and bringing the object up from below into contact with the film. Some water is trapped between your metal surface and the photoresist film sheet. The excess water can be pushed out with a clean cloth to the edges of the film-sheet metal sandwich until it's almost completely water free. It then dries in a dark cupboard for a few days, even small drops of water stuck under the film as bubbles will diffuse through the film and disappear after three days or so.

After lots of experimentation, I use a primary water container to rinse the metal object to be etched and then into the second water container with the peeled photoresist film floating on top ready for the metal to be brought up from below. There are unavoidable particles floating on the surface that can be coralled at one end by dragging a sheet of sandwich wrap plastic (glad wrap, saran wrap etc.) across the surface to one end of the container and leaving the sandwich plastic resting in the water, dirt trapped underneath. Each use of the water container seems to add dirt, so you are constantly dragging the film across to clean it ready for the next attempt. Even if a particle of dirt gets trapped under the film, it's worth exposing and developing the plate as any pinholes in the coating can be patched with bees wax or blu-tack and etched perfectly well.

I have a very boring, long winded, poorly edited video here showing a lot of the process:

 

Everything from the etching solution to the film exposure times to the developing liquid is sensitive to temperature changes too. My winter etch time for aluminium is 240 seconds and it's 120 seconds in summer. It pays to keep a notebook to get you up to speed with what works for your particular etching setup. Having a few test sample metal coupons to run a test exposure, develop and etch will let you get a good etch at different ambient temperatures. Beeswax melted on small sheets of stainless steel will coat the surface with a very acid resistant coating that can be written through with a ball point pen and etched. Beeswax coatings let you experiment with your newly made etching solution while you wait for the photoresist film to arrive in the mail.

 

Nice high quality stuff there!

 

Masking off your etchings:

Cuprous chloride etchant seems to have a very low surface tension and will creep under the tiny gap formed by one layer of duct tape overlapping another to get to bare metal and etch it. A lot of the regular etch jobs, I do use that cheap 2" wide silver duct tape for masking, not the good, very sticky stuff with embedded cloth. It probably take a bit of direct experience to get a great tape masking and it involves either window silicone dabbed on every tape overlap to block the tiny gap or alternatively good old blu tack pressed over the gaps.

Bees wax is a pretty good masking agent, unlike paraffin candle wax, it's flexible and does not crack off when disturbed. You simply apply your photoresist film and ensure the film is larger than the data plate to be etched, then lay the data plate with photoresist face down, duct tape the bared end of an insulated copper wire to the metal and pour molten wax over the back of the data plate. It may take a few layers of wax to get it nice and thick and try to attach the wire so the data plate hangs vertically in the etch tank. Cracks in the wax, especially around the insulated wire can be fixed with silicone sealant if needed.

I have some old optometry lens blocking wax which is some synthetic formulation that is flexible at room temperature but falls off aluminium in sheets after cooling in a bucket of water. I dread the day it's no longer usable for me as it masks off my gear shaped objects for etching beautifully. I apply the masking wax before even exposing the photoresist as any resist film that has been exposed to light will be hard and fragile, so a fairly rigid wax backing prevents a crack in the film from letting etchant in around the edges of the data plate.

Exposing your Data Plate

To begin with, the best intense, collimated source of UV light for occasional use is going to be direct sunlight. You need to place your artwork in intimate contact with your photoresist and ensure it doesn't move during exposure. A flexible stretchy tape placed on each four corners of the artwork to the resist coated metal sheet, in such a way to keep the artwork under tension and flat will help a lot. A pair of glass sheets hinged with duct tape will let you sandwich the art in between the glass and keep things fairly flat. A few test resist coupons and timer will let you get a consistent exposure of the resist. Everyone's light box and intensity of the sun at their location is different, so some experimentation is in order. If you compare a piece of resist that has been left in the sun all day with an unexposed piece you will see it's a much darker blue and will also be fragile and crumbly when peeled. You want a shade of blue about halfway to that of an overexposed resist. My 400W mercury vapour streetlight discharge lamp takes 2.5 minutes from 2 feet away for comparison. All you can do is process several samples to bracket times for your setup and physically see how dark the resist goes.


Developing your exposed photoresist.

Negative photoresist is cured, crosslinked, hardened or whatever by exposure to UV light. This means the areas under the black masking are soft and can be washed away with a weak alkaline solution a bit like paint stripper does. Washing soda or sodium carbonate in water works very well to develop the photoresist. I don't have a specific strength, I put about a level tablespoon full of sodium carbonate into about 2 litres of water to start with and strengthen if necessary.

Use a bit of tape on the front of the photoresist film to strip the shiny front layer of mylar film off the photoresist and then immerse it in the developer solution. After about 30 seconds of soaking, a soft 3/4" house paintbrush can be used to remove the softened photoresist. Using a brush is far more positive in removing the photoresist than just soaking and lets you reduce the time in the alkaline bath, much more time than 30 minutes in the bath will damage even the exposed, hardened photoresist and begin to etch aluminium and stain it. After about 5 minutes in the bath should have all the unexposed areas stripped and it's a good idea to rinse it off, dry it and examine the stripped areas to see if there's any of the clear glue layer left behind from under the photoresist layer. If there is then continue to strip bath for a few more minutes and check again.

Etching your data plate:

So at this point you've patiently laboured through my tedious descriptions and have a successfully exposed and developed photoresist film over some sheet metal ready for etching. You also have some newly made cuprous chloride etchant. Wearing some gloves and eye protection you can dip aluminium plates straight into the cuprous chloride with no need for DC electrical assistance. Aluminium is pretty reactive and begins to go dark and fizz from hydrogen gas bubbles immediately. Copper sponge forms around the immediate etched area and may be from an electrolysis/battery effect or aluminium robbing chlorine atoms from the cuprous chloride and leaving elemental copper sponge. You can gently brush off the copper into the etchant where it will eventually react with the hydrochloric acid. Gentle brushing prevents you dislodging the resist film and knocking loose small bits, like the inside parts of 0's, 8's, 4's, 9's etc.. Large areas of bare aluminium will heat the etchant solution speeding it up, with potentially a thermal runaway. For that reason diluting the etchant is an option with aluminium.

Etching brass with a DC electrical assist will require a power supply capable of about 5 Amps at no more than 3 volts when in use, so in practice, a lab power supply of 0-15 Volts and 0-5 Amps will be a commonly available supply. Electroplating power supplies are typically 6 Volts maximum and 20 Amps or more and would be suitable. You do want some smoothing to the power supply unlike an old fashioned transformer rectifier battery charger that has full wave rectified output which drops to zero at twice the mains frequency.

To ensure the long life of your etchant solution, you'll want carbon electrodes for the negative terminal in the solution as it won't dissolve and contaminate your bath. I use some pliers to pull the rods out of lantern battery cells and wrap some copper wire around one end before wrapping some insulating tape round the wire to protect it a bit. The positive terminal of your power supply connects to the data plate to be etched. Dial the power supply down to a volt or so and stick the plate in the etchant solution while taking note of the current flow. It should only be 1-4 amps depending on the size of the plate, if the current is too high there will be yellow chlorine bubbles round the carbon electrode so dial back the voltage a bit. Rarely have I seen etching voltages more than two volts. As the etch progresses there'll be a thick soft layer of copper sponge forming round the carbon rod. If you're etching brass you'll see a high speed flickering of patterns of flecks of silver zinc on top of the gold of the brass as the metal is removed. The pattern looks like an animation of the pattern of galvanised iron flickering into and out of existence and is quite hypnotic.

The entire time the metals are in the etchant, you must keep the etchant agitated by gently moving the metal plate in the etchant while taking care not to touch the carbon rod if it's being used. Without agitation you will see weird uneven etching, for example I was dipping one brass plate in and out of the etchant while holding it almost vertically: the flow of the etchant across the brass was etching the top of the letter more than the bottom as the etchant flowed from top to bottom. Holding the brass horizontally and agitating below the surface of the etchant gave a much more even result.

It's difficult to do fine detail and a deep etch as the edges of your letters will be undercut and also rounded off to some degree. The photo resist edges will be hanging in free space with no metal as the etch works under the surface of the resist layer.


This is a 1/32" deep etch in 316 stainless steel, it's interesting as it shows the peculiar way stainless steel etches: note the corners have a round radiused undercut despite square sharp corners on the mask. Eventually the bottom of the text will have a cylindrical cross section. The dash after the '7' had a rectangular mask but the metal has circles cutting under the film faster from the higher electric field due to corner geometry.




This is the stippled effect of using hydrochloric acid and hydrogen peroxide etchant which allows small bubbles of gas to form and mask the surface. Note the much deeper undercut of the photoresist which shared the artwork of the first example.


Duct tape masked 316 stainless steel with silicone on the tape corners and other parts to prevent leakage. There's a weird optical effect from rinse water trapped in the letters making the etch look weird. The photoresist is visibly undercut as the number one vertical edges show a frilly appearance from the photoresist in free space round the edges.

 

Very interesting write-up, and professional results Mark. I'm going to have to find a local supplier of hydrochloric acid. One thing I have noticed is much better results with plates to be etched laying flat rather than hanging vertically.

What medium have you found works best as the photo-mask? I have tried various OHP laminates and usually end up with 2-3 layers to get sufficient opacity.

 

Hardware stores and swimming pool shops have hydrochloric acid for sale.

Without some agitation or flow, the etch will cause localised depletion of the etchant, just like when electroplating. With an aquarium pump you can get good results directing fluid flow over the surface but you have to monitor it constantly. Maybe a spray system with a horizontal plate to be etched just below the surface of the etch tank and the etchant sprayed onto it. Unless the flow is even, there will be uneven etching.

I'm using a gel coated inkjet film as opposed to a laser film which has to be stacked to get a similar opaque black to one layer of inkjet. Inkjet art tends to bleed after a few months in this humid climate. Laser film can unevenly contract when it get heated by the fuser rollers so dimensions on fine details can make stacking a problem too. I was trying to etch 0.5mm pitch QFN packages and could not stack it accurately enough with laser printer film. I had an old HP inkjet with Calidad black refills that were solid black that worked great.

 

Hi Kelly, 1: I have skipped a step which is the "Development" of the exposed photoresist: as in the Youtube video linked above, a solution of sodium carbonate in water, about ten grams per litre will strip the unexposed film once the front protective mylar film is peeled off the photoresist. It takes a few minutes of soaking and light brushing with a paintbrush to work and has to be dried and examined with a decent loupe to ensure the thin clear layer of adhesive under the photoresist is also stripped at the same time as the unexposed resist or it will prevent etching.

2: If it adheres well enough to act as a resist then it's legitimate in my book, some experimentation may be needed as the damndest things will fall apart in etchants.

3: I used to have a business card made from etched stainless shim with some very fine detail. Under a microscope the matrix of fine round holes were actually conical and must have been etched from both sides to meet in the middle. The guy whose business card it was, was over the moon to have someone to talk to about the topic until his wife caught him and elbowed him in the side to shut him up. The closest thing these days would be the stainless solder paste masks available from the printed circuit board industry. I have yet to etch those lathe plates shown above in stainless but the border is wider to align with the etched lines on the other side to hopefully cut through 1.6 mm stainless sheet.

 

Very informative.

Thanks for sharing.

 

Hi Kelley, I've re-edited the posts to add a description of the development of the photoresist in the correct sequence.

-Mark

 

Tried my hand at etching brass with Ferric Chloride back in September. I found that etching vertically with no agitation gives poor results.
There are "streams" of etching above the writing which I think will not be there if etched flat.



Left it in the ferric too long as you can see severe undercut as advised by Mark. The photo mask tail of the S must have fallen off early during etching. Measures 3-1/2" x 2".

 

I have experienced the grooves cut by the flow of the etchant across the vertical brass that you describe when I dip the plate in and out like a tea bag and for that reason keep the plate under the surface of the etchant. I'm not sure of the reason for it but the more etchant flows over the metal, the faster it removes the metal and any uneven flow becomes readily apparent. An electric assist helps but doesn't eliminate the effect as well as having a decent volume of etchant reduces the turbulent flow over the plate. If you're making your own etchant, it becomes a lot easier to have enough volume of liquid. It also lets you have a bigger tank with a flow set up from a small fountain pump to replenish the etchant at the surface being etched.

Undercutting is always the limiting factor with etched plates, you can compensate at the CAD stage: Corel Draw! and Inkscape let you draw a border around letters and objects and vary the width of that border. This way you can compensate right up to the point where CAD features touch each other. The few commercial examples of etched plates I have seem to have an etch just deep enough to successfully paint, I guess commercial makers want high speed so they etch the bare minimum depth to make it work.

Photoresist film does have different levels of adhesion to metals: copper sticks best, which is what it's designed for, stainless also adheres well. Aluminium and brass stick well enough to work but when doing a regular job on aluminium I keep an eye on fine detail bits like the little triangle of resist in the number '4' which is the first bit to break off due to it's tiny size. If it does come off, I just cut the etch short for that piece and save it.

That's a very good approach to making a one off plate and a good use for a laser machine. A big laser table would let you queue up a lot of plates for production. For comparison I take about three minutes to handle and expose the photoresist for a plate. My light box design makes it difficult to handle multiples although a vacuum setup would fix that. Developing takes about 3-4 minutes and I pipeline the work with plates developing while the next one is exposing to get a good production rate.

 

This sounds as if the zinc is being separated from the copper. Is it?

The reason I asked is that I've been looking for a way to get copper out of, say, brass.

 

If I had to guess, I'd say the sheet rolling process creates a layered structure and the etch attacks both evenly.