Homebrew PCBs

Well, I can't say as I am thrilled or dissappointed.  I tried running 
a small PCB with fairly large traces.  There are a number of problems 
that I will need to resolve before expecting a good PCB from the 
machine.

1) I have to fix the problem with the Z axis saddle - this is 
something one of the newer list members has been helping me with 
(Ron), but I screwed up my end of the fix so I was running the test on 
the old hardware.  This is a mechanical problem that creates some 
inaccuracy in positioning in the Z axis.  With a samll 'V' cutter, you 
don't have much room for variations before your trace gets milled away.

2) A glitch somewhere caused the machine to lose position in X axis by 
about .200". This didn't happen on the first attempt, but did on the 
second attempt.  Not sure at all what the cause was or could be so 
this one will be a long term bug hunt I think.

As one would expect from cutting copper with a V shaped tool, the cuts 
are not the greatest.  Copper itself is not the best machining metal.  
Couple that with the reality that the cutting SFM is so low because of 
the small cutter diameter.  The most RPM I can currently get out of 
the machine is only 7500 RPM, ideally it should be running at 100K or 
even faster but that can't happen till I can figure out a new spindle 
and afford some new hardware to do it.

The traces that were not mucked up, were a tad smaller than they 
should have been.  They should have been .020" width but as best as I 
can measure it, they appear to be .012" wide.  I am using a small 
adapter chuck that is normally used as a quick change chuck for Dremel 
type tools. Having a few thousandths run-out on that is expected so I 
am not too concerned yet.

The cutter I have is a V cutter, 60 degrees "D" shaped or in other 
words a 1/2 round tool bit (or another way of putting it is a single 
flute cutter).  It is solid carbide and seems to have held up good 
during these cuts so far.

Had the X axis not lost position on the second attempt, I would 
probably have a good PCB to continue onto the second side machining.  

Also not overly thrilled with the Gerber tranlation software I have.  
It's a trial version so I can't really complain till I learn a bit 
more about it.  The isolation milling seems to work well, but I can't 
figure out what it needs/wants for the drilling operations. Luckily, 
that is pretty easy to program manually so if I can't get that to 
work, it still may be okay.

I can see I will have to get going on the closed loop concept of the 
software and hardware.  Even though I am running stepper motors, I do 
plan to have a "near closed loop" operation at some point along the 
way.

As for the speed of the process, I can see where it will be slow.  
However, I cannot yet determine how much longer it will take to do a 
PCB using isolation milling versus etching.  A higher speed spindle 
would certainly reduce the machining time so that is a primary 
drawback with my machine at present.  

But, looking at what is involved, it may not be much slower overall if 
you have a good spindle.  By comparison, the layout software I use can 
be used to print the layout for toner transfer then etching, or it can 
output the gerber files.  The translation software I am playing with 
takes about 1~2 minutes to translate the gerber into G-Code.  Once I 
have that, I load the G-code into the machine, set the origin on the 
workpiece and can (once I fix a few problems) hit start and let it 
run.  Once the first side is done, flip it over, re-set the origin, 
and let it run the second side.  Using stops on the table will reduce 
the amount of time there.  The machine has a 38 tool, automatic tool 
changer so once it get's going, there isn't much to do except watch it 
do the work.  It will do the isolation milling, drill the holes, and 
then mill the second side (although that requires manual intervention 
to flip the part).

Compared to etching, I would eliminate the toner transfer which is 
mostly "Hit" for me rather than "Miss" so that doesn't take too long.  
Still have to print the layout, cut away the excess, align the two 
sides, transfer it, soak off the paper, scrub off the residue, touch 
up any spots that are not good.  Then it's into the etchant for about 
20 ~ 30 minutes (it's cold in the shop so it etches pretty slow this 
time of year).  Once done there, it's manual drilling.  That goes 
pretty fast for me, but it is SOOOOOOOOOOO boring to me that the CNC 
machine is a welcome sight just for that alone.

So, that's my first attempts at isolation milling.  I have to fix some 
software and hardware bugs before I can really do much else with the 
machine.

Chris