The Saltwater Etch Process
Intro: The Saltwater Etch Process
I shall illustrate the process by etching and building a board for 18-pin PIC (for the PC16F54, but any 18 pin PIC will fit in it) in the figure. It has to plug into my breadboard and accept the programming signals from my PIC programmer (just go to http://geocities.com/it2n/circuits.html and look at it).
To avoid battling with signal conflicts, the two programming pins shall not be brought to the breadboard. To play around with the clock frequency, the crystal shall be made pluggable. The Master clear signal will not be brought out.
These decisions mean a board with two .1" pitch connectors, one with 13 connections and the other with five connections, one pin spaced apart from the rest.
This is a tutorial intended for the absolute beginner, and almost every step shall be illustrated. I've even included a video of the etching process.
STEP 1: Decide How Large the Board Has to Be
Say one and a half inches, a nice figure. Take a piece of copper clad board larger than 1.5 inches a side. Draw a line at one and a half inches.
STEP 2: Score a Line on the Copper
After some time, there will be a gouge on the copper, dividing it into two.
If you bear down with the knife, chances are that it might wander and cut the board deeply where you do not want it cut - and you will be looking ruefully down on your ruined PCB stock. Be patient. Being patient has its own virtues, as life will invariably teach you.
STEP 3: Make That Line a Deeper Groove
Then at each edge, mark the plain surface of the board and draw a line there, too, exactly on the other side.
STEP 4: Score the Other Side
You will have a board with grooves on both the sides, and bending it with the fingers will be sufficient to cause it to break neatly at this line.
This is the copper side, with a deep groove.
STEP 5: Groove on the Plain Side
STEP 6: Break It Apart
STEP 7: The Cut Piece of PCB Laminate
It will need to be sanded down to make those edges smooth and that will take away a bit of material.
STEP 8: Decide How Large It Has to Be
We need the two connectors, the PIC, the crystal, and some capacitors and one resistor.
Arranging them all on the board, it seems that about 2" will be sufficient.
STEP 9: Clean the Board
Remove rough edges of the board using a sandpaper (Or go out and rub it on a level rough cement surface).
Clean the copper surface using an abrasive cleaning pad - the one I use is intended for use in the kitchen, and copper is toxic so do not relocate it to the kitchen after you have used it - it would also be a good idea not to borrow the one in the kitchen for this purpose.
STEP 10: The Cleaned Board
The cleaned board will have a rough surface due to the scratching action of the abrasive pad, and this helps the board retain the etch resist better.
STEP 11: Apply Etch Resist
Now you paint the area with some etch resist.
It can be any paint - it has to hold together under water, that's all. Permanent marker comes in an easy to apply form, and that is what I use.
You can use nail polish, though fancy brands may be too expensive.
You need a thorough thin coat, one which can be scratched away in a thin line.
A thick coat is likely to pull away in flakes when you attempt to score lines on it.
STEP 12: Mark the Position of the Component Leads
I have clamped the 16F54 to the board with two clipodile crocs. Mark the position of every pin, and you can lift each alligator in turn to mark underneath it.
STEP 13: Draw Around the Pads With a Sharp Instrument
Repair them with a dab of the same stuff, and continue on to the next step: outlining the pads.
Use a clear transparent ruler and something with a sharp point to draw the outline of the pads. Refer to the layout you have prepared earlier.
You should think of separating point A from point B by drawing a line between them.
The conventional approach is to link point A to point B by drawing a line connecting them.
My approach helps to keep the maximum amount of copper on board, and minimises the amount of material to be removed.
This is important, otherwise the etching process will be too slow.
STEP 14: Draw the Rest of the Circuit In
I used a bit of veroboard as template for the .1" spacing, and placed the rest of the cuts as per the layout in the diagram.
STEP 15: Errors Can Be Corrected
Just paint over the scratches, and you have a fresh canvas to practise your art.
Now I decided that the board can be made more compact if the programming socket is relocated, so it was done.
One final necessary step is to place dots of paint so that all pads are connected together - this is necessary for electrolysis.
In my board, the pads are all connected along the left and bottom edges. The scratches in the paint stop short of the edge. They will be cut apart after the etching is completed.
Now the board is ready to be etched.
STEP 16: The Etching Tank
It is plastic, transparent, and has parallel sides, and is large enough to accommodate the board.
Also pictured is the negative electrode. A thick copper wire will do. I've used a straightened paper clip, but it tends to introduce rust into the solution. Almost any metal wire will work here.
STEP 17: The Etch Tank Setup
I should recommend that you do not set this up close to your keyboard as the salt solution is corrosive if it gets into electronic equipment.
It is also helpful to have a tweety bird watching over you. In fact, I shall disclaim all responsibility if you do not have a twee ...
STEP 18: Alternative Tank
This had the advantage of requiring even less of the etching solution.
The etching solution is made by dissolving as much salt as possible in water.
STEP 19: Etching the Board
A 12 volt supply capable of supplying about 500 milliamperes is sufficient. Wire it in series with a filament lamp, say 12V, 6W, for an indication of the current.
It would take about five minutes for the etching process to be completed. You will be able to watch its process by the light shining through the gaps that open up as the copper gets eaten away.
Bubbles of hydrogen will be seen rising from the negative electrode. If they are rising from the copper, you have connected up the supply backwards and your wire is being eaten up.
To make up the salt solution, you take some water and dissolve as much salt as possible, thus making up a saturated solution. You add a little salt, shake it and watch the salt disappear as it goes into solution. Then you add some more, and it too disappears.
After some of this the salt will stop dissolving however much you shake and mix it, and at this point you have the saturated salt solution you need.
This is a saturated solution of Sodium chloride in Dihydrogen Monoxide. See http://www.dhmo.org for more information in handling this potentially hazardous substance.
STEP 20: The Etched Board
Or you can use the abrasive pad to rub the paint off.
It might also be possible to leave the paint on the board, and remove it only where it is needed to solder. In this case it might produce toxic fumes while being soldered.
You will notice that all the traces are joined together, these will have to be cut apart before the board can be tested.
STEP 21: Testing the Board for Short Circuits
You may use a 12V lamp connected to the same supply as used for electrolysis for testing. Or the way I do it - use a 12V, 15A supply together with a car headlamp for indication. Small shorts that used to be there get vapourised, and if that lamp lights up, boy, that really IS a short circuit.
For clearing the short circuits just run the sharp point of a knife through the lines. Angle it one way in one pass, and the other way in the next pass, and any copper still in there will just lift and crumble away.
STEP 22: Drill Holes
STEP 23: The Sockets
We need to liberate the pins from the plastic. Heat the pin carefully until the plastic softens (but does not melt) and pull the pin free. We need seven of those pins.
STEP 24: Crystal Socket
I have a closeup of the crystal mounting so that you can see the details.
It is possible to grind away some of the rear end of those pins in order to reduce height. The part which actually makes contact is the spring inserted inside the hollow pin, and you can grind away quite a lot of material without damage to the internal working parts of the pin sockets.
STEP 25: Mount the Components
The seven pins making up crystal and programming sockets.
Two capacitors around the crystal
A 10K resistor pulling up the MCLR line to Vdd
A supply decoupling capacitor across Vss and Vdd
Two links have been soldered into place, and one more has to be fitted.
The lands for fitting the integrated circuit have been coated with solder prior to fitting it in position.
STEP 26: Mount the Integrated Circuit
It is first placed in proper position and one corner pin heated with the soldering iron. Since there is a small amount of solder on the board, this melts and holds the ic in position.
Then the diagonally opposite pin is soldered, after making any slight adjustments in position as needed.
With the two corners soldered, the ic will be held firmly in place and so the rest of the leads are soldered.
This completes assembly of the board, and it may be tested by loading a "LED Blink" program or something.
I have soldered a Microchip PIC16F54 on this board, but this board will work with any eighteen pin PIC as well.
Some of the more advanced chips permit using the MCLR pin as an input so this might also have to be brought out to the edge.
STEP 27: The Completed Board
It is compared with the original plan. I have made a few changes, mainly because it was simpler to route the traces that way.
It is essential to record changes because of the possibility of confusion later when using the board.
In this case, some traces pass under the chip and it is not easy to ascertain the order of the signals at the edge just by looking.
Documentation is essential and will take the form of signal names written right on top of the signal lines.
To make a second board you have to do all this all over again - This process is recommended only if you are making a single piece of a circuit, for which conventional prototyping methods are inconvenient.
I hope all this has been useful to someone out there.
I hope to be seeing a few projects on hand carved, salt water etched boards here in instructables in the near future.
Have Fun
89 Comments
rotor 17 years ago
Btw, this is my favorite part:
Or the way I do it - use a 12V, 15A supply together with a car headlamp for indication. Small shorts that used to be there get vapourised, and if that lamp lights up, boy, that really IS a short circuit.
I actually loled.
ArifSae 7 years ago
I have noticed , this is only good where the copper is single plane, what about parts separated. also some copper is not etched and left as a dust of board.
I experimented and found these steps to work for me:
I etched pcb in 2 steps,
Step 1: etch away all major copper via salt water electrolysis.
Step 2: etch remaining copper dust via hot Vinegar+hydrogen per oxide + salt water
greatnameless1 15 years ago
kyle.marsh 14 years ago
splud 9 years ago
You could of course draw the resist on in the pattern of the circuit, and touch up around fine connections (such as the IC pins) by scraping away excess mask with a pin tip (or the awl tool commonly part of soldering tool sets).
I'm all for doing things on the cheap (I etch my own PCBs because I don't want to wait for a service to do it, and don't want to pay for fast shipping), but at the point you're doing circuit design work and have ICs and components to solder, surely a secondhand laser printer isn't a big outlay. The saved time alone, not to mention savings in material use (copper PCB material for instance) will make it a quick payoff.
splud 9 years ago
I have a wife, but I don't use her nail polish - I have several bottles of clear acrylic for coating windings, and a few older bottles of garish nail polishes I picked up for tamper seals and makeshift threadlock.
mbainrot 12 years ago
It should be noted that you need a substantial load when using an switch mode power supply otherwise it may not switch on
mbluett 17 years ago
neelandan 17 years ago
Jimmy Proton 13 years ago
stayputnik 12 years ago
neelandan 17 years ago
macumbista 15 years ago
reyfari 15 years ago
This is really just a "redox" (oxidation/reduction) reaction.
What happends is, The Chlorine ions (Cl-) migrate to the cathode where they loose electrons and bind to make Cl2(g). This gas is toxic even in small amounts, you should read MSDS for Cl2(g). (Just type Cg2(g) into google). You should therefore perform this in a well ventilated area (or not at all).
The Copper at the cathode (the cathode is the PCB board) looses electrons and dissolves in the solution. If there is excess water a complex-ion will form Cu[(H2O)6]2+ (Octahedral aqueous copper complex with a charge of +2), but this complex will be in some equilibria with the Cl- ions. If The Cl- ions are in excess (which would probably not happen in a aqueous solution) a yellow complex Cu(Cl4)2- would form (in equilibrium with CuCl2 which is quite soluble in water).
Some Hydrogen gas could form at the anode due to the self-ionization of water H20 <---> H+ + OH- (this equilibrium lies far to the left). H+ ions would migrate to the anode, pick up electrons and leave the solution as H2(g).
When two H+ ions leave the solution as H2(g) after reduction, two water molecules will dissociate to form 2H+ + 2OH-. The OH- ions combine with Cu2+ ions to form Cu(OH)2(s) which is insoluble so it will precipitate as a green-looking solid.
So, What you have in your solution is
a) Cu(OH)2 (s) (amount is proportional to the amount of hydrogen displaced, which is proportional to the amount of Cu(s) etched away).
b) Cu[(H2O)6]2+ (amount depending on the amount of Cl- to cause...)
c) CuCl2 +2Cl- <--> CuCl3(-) <--> CuCl4(2-)
d) Some Na+ ions and water
What will leave the solution?
H2(g) and Cl2(g)
So, now you know. And what you should do if you want to know if you can flush the solution down the drain is read the MSDS for Cu(OH)2, and Copper-solutions. Although I can tell you, we don't like solutions of heavy-metals flushed down the drain. You should dispose of this properly.
I suggest you find another electrolyte to avoid the Cl2(g). I don't have many ideas tough since my access to acids and metal-nitrates and such is not limited :-).
Hope this helps
Greetings from Iceland.
Benedikt Ómarsson, B.Sc
citizenray 14 years ago
Jimmy Proton 13 years ago
DrNerdington 14 years ago
To the author: I have long considered using both this etch resist application process and reverse electroplating for PCB fabrication, but you actually did them... at the same time... and you wrote instructions for us! Good show!
I plan to adapt this process to the use of my CNC router with some kind of spring-loaded scratching implement and of Peroxide+Muriatic Acid as the etchant (assuming I can find an etch resist that is safe from its ravages... I'm thinking wax).
Prodigity 14 years ago
No biggie if you do this out in the open but if you put a lid on it to erm fend of the fumes or something you might be actually doing more bad than good.
Can't see any problems however if you just do this outside =)
chaydgb 13 years ago
soapmaker72 14 years ago
1. High salt concentration in solution
2. A pH of 7 or lower( acidity promotes this)
3. An anode(+) that is CHEMICALLY INERT to chloride or chlorine
Notice that NONE of these conditions are met in this case. The salt is in lower concentration. The pH will probably increase if anything(go more alkaline). And last but not least the copper is not inert and hence it is etched.
Secondly, Disposal of copper waste is not an issue. The oxides/hydroxides can be dissolved in diluted muriatic acid, vinegar, or battery acid and an aluminum wire is added until a clear solution is obtained. This can be decanted down the drain as it simular to deodorant (aluminum chloride) or baking alum(aluminum sulfate). The pinkish copper grains can be mixed with saw dust, oats, flour or any organic that chars black and borax and melted with a propane flame to get copper metal for later use.