17 Oct
2012
17 Oct
'12
5:38 p.m.
On Wed, 17 Oct 2012, David Riley wrote:
Once I get access to the boxes that I have the boards stored in I'll see
if I can snap some photos. It may be awhile, however.
On Oct 17, 2012, at 12:18 PM, Tothwolf wrote:
When you get a bridge, clean the iron's tip of excess solder, apply a drop
of liquid rosin flux to the joints, and drag the iron down and away from
the component towards the ends of the leads. A narrow chisel or heel type
soldering tip helps, but the real key is the liquid RMA flux.
Despite how they look, QFPs are among some of the
easiest surface mount
parts to fit. A 25W pencil iron with a small tip (~2mm) and some liquid
RMA flux will do the job. I've used irons with tips as large as 4mm for
this task when in a bind. Granted, I've been soldering for decades, but
still...SMT isn't /that/ hard...
I suppose. I've had terrible luck with bridging, but I never really
kept at it long enough to get much good at soldering fine-pitch parts.
My wife got me a pretty decent iron as a birthday present recently,
though (I picked a good one!), so I may be doing a bit more in the
not-too-distant future (more so if my QBUS board ever gets off the
ground, because I don't really feel like paying for assembly). It was also
extremely common to see AMD 386DX QFP chips fitted onto a
PGA carrier board that could be plugged into a standard i386DX socket.
I own a bunch of these (and motherboards for them) which I plan to use
for a future multi-processor project. [That is, IF I can ever get my
hands on enough 40MHz 387DX math co-processors -- one guy on eBay has
had them for more than 10 years but his per unit price makes my project
cost prohibitive since I need at least a dozen of them.]
I had never seen them, but then I realized that I didn't start building
machines until the 486 era. SMT-on-a-carrier is certainly a common
paradigm; after all, it's basically what flip-chip PGAs are on a
different scale.