[sdiy] PCBs

[The Old Crow]

On Thu, 3 Jan 2002, Scott Gravenhorst wrote:

> I always read and practiced that gentle agitation and warming 
> the FeCl3 solution speeds the reaction.  I don't see a real 
> good reason to rush etching anyway, especially for home DIY.
> AKAIK, a reaction that is even across the surface is the 
> ultimate desire.  A slower reaction will be easier to control.
> Once the reaction has gone "too far", you start over and the
> board, the etchant, your money and your time are all wasted.

  In the 1980s, the industrial controls manufacturer I worked for (and on 
occasion, still consult for) had an in-house PCB lab.  It was by no means 
state-of-the-art (no hole plating or silkscreening), but it was certainly 
good enough for me to make just about any synth module PCB.  We had a pair 
of 25-gallon etchant tanks that were heated by four immersion heaters such 
as used in aquariums.  Etch time largely depended on the age of the ferric 
chloride, but typically a board could be done in about 40 to 45 minutes.

  The trick we used then was to immerse the exposed & developed boards for 
about 25 minutes, then remove them and rinse under hot water.  It was 
possible to tell the estimated completion time by how much copper had 
already been dissolved.  For boards that showed copper etched away in 
places, another 15 minutes was all that was required--though of course the 
final minutes of a board-etch are closely observed.  The only problem with 
fresh etchant was the tendency to etch under the photoresist before all 
the bulk copper was dissolved.  To remedy this, we decided to stagger the 
replenishing of etchant between the two tanks, replacing the solution in 
each tank every six months, but offset by three months for tank 1 vs. 
tank 2.  This way, if we discovered boards etching 'too quickly', we would 
rinse the boards then place them in the tank with the older, slower-acting 
etchant.  This allowed the boards to finish without etch-under problems.

Crow
/**/