Today I decided to solder the external 32 MHz kHz crystal supplied with the LaunchPad. I quickly discovered that the LaunchPad board itself has both SMD and through hole solder pads, but the supplied crystal is surface mount. I'd never soldered surface mount before, and there are no instructions anywhere I can find on soldering this crystal to the board.
Google to the rescue! This post from Justin's Tech website very nicely describes the process, and includes pictures. Justin even wrote a program to test for proper operation of the external oscillator.
I was really glad for this, as I didn't know the crystal should be lying on its side and have the body soldered to a third solder pad on the board.
My first experience with SMD soldering was not too bad. I managed to solder bridge one of the two leads of the crystal to the body of the crystal, and had to employ my solder sucker to clean it up. I was also not happy with the size of the solder blobs I left on the two pads for the leads, so I sucked more solder off them and resoldered. The picture below shows my just-barely passible handiwork.
While my iron was hot and the MSP430 processor was removed from the board, I also soldered on the supplied header pins to the LaunchPad. I had decided to use the male-female headers, and connect them to the bottom of the board so it could insert on top of a to-be-fabricated project board. To keep everything aligned during soldering, I kept the male-female headers attached to the male-male headers and loosely inserted into an empty protoboard below the LaunchPad.
I accidently soldered one male-female header and one male-male header to the LaunchPad. After trying to desolder the row of 10 male-male pins, I decided to call it a serendipitous accident and just kept it as is. So now I'll have a keyed pair of headers on the board and will have no way to accidently connect the board backwards.
So there we go. Physical soldering is done for the LaunchPad board itself; since the pins for the chip are broken out to the headers, that should be the last I need to do to it.
As a final step, I uploaded Justin's non-GCC version of his test code to the LaunchPad. Everything ran fine and the crystal appears to be working.