another question:
does anyone have experiences in (hand-)soldering the tiny QFN packages?
best regards, stefan
It is possible, but a pain to do. You must have a microscope or some kind
of magnifier or really really really good eyes.
I use an Olympus SZ40. http://www.opelco.com/sz40_specs.htm. 0.7 to 4.0
with 10x eyepieces for a 7 to 40 overall zoom. At 40x only 10 pins fill
your view. Normally about 10x is fine. Depth of focus is a bit of a problem
because you have to refocus each time you raise the board to tilt it
sideways to look under stuff. (At 40x if you focus on the top of the chip,
the pads are just a blur.) You'll probably want a bright light source too.
It is amazingly dark where the trolls hide under the chip.
See http://www.labx.com/v2/adsearch/Detail3.CFM?adnumb=200978 for a used
SZ40 or perhaps something else like it. You're kidding yourself if you try
to do it without something like this.
I have put together several boards this way but it takes mucho patience.
The PCB proto places generally don't do solder mask between pins down at
those levels so bridging is common, but that is fairly easy to deal with.
The hardest part is getting good wetting directly under the pin. It will
look fine, but pop off when a little pressure is applied. When looking
through the eyepiece, the smallest tip you can get usually covers a pin on
either side of the one you are attacking, but a little tilt on the iron
will let you contact only the desired pin. It will just look like
everything is going to be one big blob when you pull it back.
As stated before, leaving extra length on the pads is a help. Do it on the
outside, not the inside though. :-)
Most of my 430 pad placements seem to end up with more pad under the chip
than I want. I think it is mostly due to the way the pins are spec'd.
Gary M
