Homebrew PCBs

> Phil wrote:
> 
> 
>>Very nice site!  It would have been very helpfull when I was trying CuCl.

...
Adam Seychell wrote:
 > Thats interesting your experience were different than what I measured in
 > my test. The bubbles agitation for my test was very vigorous, so maybe
 > that had something to do with it. The FeCl3 was full strength as sold
 > industrially (sp.gr=1.47), the AP was 200g/l almost fresh (very faint
 > blue color due to previous tests). For AP I was getting about 9 minutes
 > with 1oz copper at 42\ufffdC. 5 minutes sounds fast! Two possible reasons:
 > you unknowingly have "1/2 oz" PCB material, or the ferric chloride you
 > were using was for some reason not very potent which made it slow in
 > comparison. I'm equally curious becuase it could mean my test results
 > are botched.
 >
 > If you have access to a micrometer caliper, you can peal copper away
 > from the board using a heat gun and pair of pliers, burn off the
 > remaining epoxy stuck to the back of the foil, and measure its thickness
 >
 > Adam

Ferric chloride works faster if it is diluted with water (to 1 part or 2
parts water). A properly cleaned 35um copper pcb should etch in 5 minutes
at room temperature with gentle/moderate bubble agitation.

Diluted ferric chloride:

Fe(3+) + 3.Cl(-) + 2.H(+) + O(-)   ions in solution

When in contact with Cu metal, the Cu goes into solution as Cu(2+),
so two electrons need to go somewhere. These are consumed to make
ferric into ferrous or Fe(3+) -> Fe(2+):

2.(Fe(3+) + 3.Cl(-) + 2.H(+) + O(-))   +   Cu(2+) + 2.e(-)

-> 2.Fe(2+) + 6.Cl(-) + 4.H(+) + 2.O(-) + Cu(2+)

Water is therefore *required* for fast etching.

http://www.artmondo.net/printworks/articles/ferric.htm
http://www.jtbaker.com/msds/englishhtml/f1060.htm