Homebrew PCBs

Hi, I have recently rescued a HP 7475A plotter and I am looking at 
converting it to handle pcb scratch etching (I think the mechanism is 
nowhere near strong for milling).
My question is do any members have any good links or files on scratch 
etching?
I am thinking either direct copper-scratching (styluses??) or 
scratching though an etch-resist and etching the board as normal.
Has anybody tried this? 
What type of results did they get on track resolution??

regards, PK

On Thu, 27 Jan 2005 15:07:12 -0000, onenastyviper  
<oneNastyViper@...> wrote:

>
> Hi, I have recently rescued a HP 7475A plotter and I am looking at
> converting it to handle pcb scratch etching (I think the mechanism is
> nowhere near strong for milling).
> My question is do any members have any good links or files on scratch
> etching?
> I am thinking either direct copper-scratching (styluses??) or
> scratching though an etch-resist and etching the board as normal.
> Has anybody tried this?
> What type of results did they get on track resolution??
> regards, PK


track resolution with a OHP pen (staedtler) was bad, here.
couldn't even get a trace between DIL pins, and the width is inconsistent  
because it depends on how fast the pen moves.

I never tried scratching a resist, but my suspicions are that it has  
similar drawbacks as milling in terms of speed and result.

As for tt we takled about, i think a old printer at ebay is cheap enough  
not to bother with the plotter method.
I'm not one of the usual nay sayers (hi mike ;-) ) but i have tried it and  
have had bad results. Problems seemed unresolveable so i abandoned the  
method.
Now the weakest link on bbc is about to begin, so i'll leave it at that.

ST

Hi PK,
	I've done scratch etching before, and it does work well, at least in theory. :-) There are refinements to my process that need to be made before it works as well as I'd like.

	I use a 7475A plotter. I've replaced the felt pen part of one of the pens with a carbide tipped machinists scribe. I clean the copper board real well and then apply a coating of Dykem blue (machinists marking fluid). I do all my layout in Eagle, then I use a custom ULP program in Eagle to generate HPGL that traces around all of the nets multiple times in increasingly larger distances from the net. Then I use a Liberty Basic program I wrote to take the resulting HPGL file and apply the finishing touches, like the drill file holes, centering the image on the page (for double sided registration), etc. Then my Liberty Basic program spools the HPGL file out to the plotter. Then I flip the board over and scratch the other side. Then etch and clean off the bluing.

	The good news from this (from my experimentation) is that the line that the scratch tip (if honed well) is less than .003 wide in the Dykem, and it is very consistent. No flaking of the bluing or anything. It generates very sharp lines in the copper when etched. Also, you run the plotter at it's fastest speed - it's not like milling where you have to go slow.

	The down sides of the process are that you have to go over, and over, and over since you move the scratch tip out from the net only .001 each pass (to make sure you catch everything). So if you want .020 then you have to make 20 passes around EACH net. It does take awhile. But it's something you can leave run on it's own. The other downside of the process as I'm running it right now is that there is a physical 'drag' or 'lag' between the tip of the scratcher and the center of the carriage that moves the scratcher. This manifests itself as a small 'kink' each time you change direction. It has not affected the quality of my boards yet (electrically) but it doesn't look quite as good. I believe that I can modify the HPGL file in software to compensate for the tip drag, but I haven't gotten far enough along yet to try that out.

	I've got the source code for the ULP that I use along with the Liberty Basic program too at http://greta.dhs.org/EagleScratch/index.html also, there are some examples of the scratched boards at the EagleScratch photo album at http://photos.groups.yahoo.com/group/Homebrew_PCBs/lst 

	If you have any questions, let me know.

	The bottom line is that it works, and it can work pretty well, but it takes more equipment, time, and putzing that TT.
	
*Brian

-----Original Message-----
From: onenastyviper [mailto:oneNastyViper@...]
Sent: Thursday, January 27, 2005 9:07 AM
To: Homebrew_PCBs@yahoogroups.com
Subject: [Homebrew_PCBs] RE: Scratch etching




Hi, I have recently rescued a HP 7475A plotter and I am looking at 
converting it to handle pcb scratch etching (I think the mechanism is 
nowhere near strong for milling).
My question is do any members have any good links or files on scratch 
etching?
I am thinking either direct copper-scratching (styluses??) or 
scratching though an etch-resist and etching the board as normal.
Has anybody tried this? 
What type of results did they get on track resolution??

regards, PK







Be sure to visit the group home and check for new Bookmarks and files:
http://groups.yahoo.com/group/Homebrew_PCBs 
Yahoo! Groups Links

Hi Stefan,

Then I will add that scratching method is not worth anybody's time 
unless they are imprisoned. Mike

--- In Homebrew_PCBs@yahoogroups.com, "Stefan Trethan" 
<stefan_trethan@g...> wrote:
> On Thu, 27 Jan 2005 15:07:12 -0000, onenastyviper  
> <oneNastyViper@h...> wrote:
> 
> >
> > Hi, I have recently rescued a HP 7475A plotter and I am looking at
> > converting it to handle pcb scratch etching (I think the 
mechanism is
> > nowhere near strong for milling).
> > My question is do any members have any good links or files on 
scratch
> > etching?
> > I am thinking either direct copper-scratching (styluses??) or
> > scratching though an etch-resist and etching the board as normal.
> > Has anybody tried this?
> > What type of results did they get on track resolution??
> > regards, PK
> 
> 
> track resolution with a OHP pen (staedtler) was bad, here.
> couldn't even get a trace between DIL pins, and the width is 
inconsistent  
> because it depends on how fast the pen moves.
> 
> I never tried scratching a resist, but my suspicions are that it 
has  
> similar drawbacks as milling in terms of speed and result.
> 
> As for tt we takled about, i think a old printer at ebay is cheap 
enough  
> not to bother with the plotter method.
> I'm not one of the usual nay sayers (hi mike ;-) ) but i have tried 
it and  
> have had bad results. Problems seemed unresolveable so i abandoned 
the  
> method.
> Now the weakest link on bbc is about to begin, so i'll leave it at 
that.
> 
> ST

On Thu, 27 Jan 2005 21:37:02 -0000, mikezcnc <eemikez@...> wrote:

>
> Hi Stefan,
> Then I will add that scratching method is not worth anybody's time
> unless they are imprisoned. Mike


lol, i expected you to ;-)
(I agree by the way, this time)

ST

--- In Homebrew_PCBs@yahoogroups.com, "Stefan Trethan" 
<stefan_trethan@g...> wrote:
> On Thu, 27 Jan 2005 21:37:02 -0000, mikezcnc <eemikez@c...> wrote:
> 
> >
> > Hi Stefan,
> > Then I will add that scratching method is not worth anybody's 
time
> > unless they are imprisoned. Mike
> 
> 
> lol, i expected you to ;-)
> (I agree by the way, this time)
> 
> ST

Since scratch-etching sounds like it can (mostly) run unattended, 
the time that *you* have to spend may not be all that great. Granted,
it may take a while to crank out a board. If it's a set-and-forget 
operation, though, I don't see the time being that big a problem 
when you're only doing a board or two.


Regards, Bob

On Fri, 28 Jan 2005 00:33:52 -0000, Bob_xyz <bob_barr@...> wrote:

>
> Since scratch-etching sounds like it can (mostly) run unattended,
> the time that *you* have to spend may not be all that great. Granted,
> it may take a while to crank out a board. If it's a set-and-forget
> operation, though, I don't see the time being that big a problem
> when you're only doing a board or two.
> Regards, Bob


Bob, i made a boad yesterday from 12 to 1 in the night, 2-sided  
smd/throughhole, both sides component legend, ground planes, about the  
hardest you can get with tt.

Even not couning all the wait time for the plotter i don't believe you can  
best the time or effort or lack of concentration tt requires once you are  
familiar with it.

I don't say it's impossible, just that it didn't work out for me.

By the way - pre-heating the paper worked like a charm to eliminate all  
visible distortion/shrinking problems. I just run the paper through the  
fuser first.


ST