Homebrew PCBs

Jeremy Taylor wrote:
> 
> To note however, I dont use toner transfer. Any images posetd by me, (as in the photo of the traces between an so-70 nanostar package, was done with photoresist imaged in a vacuume uv box. with a transparency printed at 1200 dpi, with selective calculated undercutting during etch that gave me sub 5 mil traces.
> 
> JT


Jeremy,
Can you describe your inkjet setup, what printer brand/model ?, 
is it OEM ink or refill, and what transparency brand ? All these 
factors effect quality of photomasks made from inkjet printer. I 
have tested a few transparency and inkjet combinations and found 
huge variations in quality. In general, the manufactures 
recommendations are the best solutions. There are other 
combinations of products that will save you money but require 
lots of testing.

As for printers that claim "1200 dpi", I am wary that this is a 
useful specification. All the printers I've tested show the ink 
droplets to spread or land in random location with error of 1 or 
2 mils. For printing color photos it doesn't matter.

Here is a picture from my Epson stylus 660/Epson 
transparency/Epson ink, I haven't seen any inkjet prints with 
line edge jaggedness much better than in this image. My cheap 
Cannon S330 inkjet is only slightly worse, but still very usable.

Notice the vertical edges are jagged, while the horizontal is 
smooth because the head direction is in this axis.

http://members.optusnet.com.au/~eseychell/pictures/Epson_inkjet_film.jpg

Here is the same image in the photoresists on the copper.
http://members.optusnet.com.au/~eseychell/pictures/resist_image_on_copper.jpg

Adam