Thread: Home made PCB?
View Single Post
  #9   Report Post  
Old July 1st 08, 10:58 AM posted to rec.radio.amateur.homebrew
Leon Leon is offline
external usenet poster
 
First recorded activity by RadioBanter: Jan 2007
Posts: 24
Default Home made PCB?

On 1 Jul, 00:30, K7ITM wrote:
On Jun 30, 10:12 am, "numeric" wrote:



Hi,


I would like to build a PCB board with fine pitch (LQFP32) and very small
surface mount parts (1206 and smaller) parts. The board is a simple 24 bit
A/D and D/A converter with a full speed (12 Mbs) USB2 interface; although I
would prefer the USB2 high speed interface (480 Mbs). The temptation is too
hard to resist using today's microprocessors such as SiLabs C8051F350
(basically an 8051 cpu). I know that a commercial 4 layer PCB with a solder
mask would probably work for homebrew construction; but the board cost is
high.


So what is the chance of a double sided PCB without solder mask working with
very small parts? The process I would use in either case would be:


1. Place solder paste on all pads. Typically, when heated the solder will
vacate between pins and will suck up, like a magnet, under the SMD pin.


2. Place glue under parts that will possible move when the board is handled.


3. Heat the PCB in a toaster oven until soldered. Manually follow the heat
up and cool down time curves for type of solder paste used.


Comments/suggestions are appreciated.


I agree with Leon and Tim that making a board and mounting parts on it
shouldn't be a problem, though I do greatly appreciate being able to
do my soldering under a stereo microscope using a Metcal iron.

I commonly use boards coated with positive resist (from Circuit
Specialists in the Phoenix area), etched by floating them on top of
ferric chloride. *I find that (lacking a spray etcher) the float
process maintains fine line resolution--I have a theory that it has to
do with copper-laden etchant being heavier and falling away from the
etch site, but whatever, it works well for me. *Surface tension will
hold fairly large boards up, but you can also just put support at the
corners and let the board submerge. *Resolution for me is always
limited by the quality of the transparency. *Thanks to Leon for
posting a reference to another film material to try. *I've had the
best luck using an Epson inkjet printer: *my HP printer doesn't do as
well, and laser printers typically don't print a dense enough black.
Pinholes WILL print on the board! *I use a back-light film setting on
the Epson for highest density. *An alternative is to get a
photoplotting shop to do the film for you: *better quality but it
lacks the immediacy of being able go from layout to board in minutes.
Anyway, the best inkjet printed transparencies will allow me to do 5
mil lines with 5 mil spaces if I'm careful, and 8/8 pretty easily.
Most of my work is with 0805 as the standard passive, since that's
what I invested in an assortment of back a few years ago, but I also
do 0603 with no trouble, and lead pitches down to 0.5mm. *We have
fancy equipment to do BGA rework here at work, and the trained techs
have marginal success with it at best (for large parts), so I've
avoided BGAs in my home projects.

By the way, if you get boards from Circuit Specialists, be aware that
what they call 1 ounce copper is really 0.5 ounce...seldom a problem
and allows faster etching and it's easier to make fine lines, but be
aware of it for high current traces.

Cheers,
Tom


I use a Metcal and microscope, as well.

I use manual agitation in hot FeCl3. I don't have any problems getting
down to 8 mil tracks or even less.

Leon