SMD Adventures

Moving on to new adventures: SMD desoldering of big fat chips. Getting a hot air rework station is too expensive now, so I'll try this first. The first victim will be an ATmega64 sitting on my Jyetech oscilloscope DIY KIT. www.chipquik.com

You don't need much of the alloy. It's also very brittle. Bending it like normal solder doesn't work. I just cut it to size.

The only winning move is not to play. This time I put the LEDs into my freezer, and I only wrecked 1 out of 40! Considering the small mass of the package and that hand soldering each of them takes me at least 30 seconds, I doubt that it had any effect at all.

I was planning to use 0402 components in my next project, as well as AVRs in MLF20. Saving space is good I thought. Now that I have seen the size of these LEDs, or more precisely the LACK thereof, I'm not so sure anymore. 1206 and 0805 are not critical and I have dealt with these many times before, but 0402... holy sh... for these I'd need a new pair of glasses + a good magnifying monocle.

Now this stuff is GOOD. No smoke (at 285°C), stays on the PCB for quite a while, almost no smell at all. And it already comes with a nozzle. No need to buy another one as I did. The nozzle was horrendously overpriced at close to 2$. I totally agree with the verdict of the guys at dangerousprototypes. I did a quick test job (TQFP-32 7x7 package) and it just worked nicely. One 'helping' of flux was enough. I didn't have to constantly re-add flux like in case of the FL-88 dispenser pen.

LT1618 boost converter finally mounted on a DIP adapter. Prototyping can begin.

I shouldn't be able to say this, but I AM surprised how small these traces actually are. I will need to use a magnifier to be able to cut them - if the blade is narrow enough...

I hope this thing is air-tight enough. Edit: I won't use this to store solder paste, I've found something better and cheaper ;-)

I was adding a bit too much paste, but it worked out fine.

for the first time ever.

I have ceran hot plates, which essentially are strong IR heaters, so I just placed the board over it with some distance and played with the dial to simulate the temperature profile for SMD reflow. In the end I just set it to max and turned it off after all parts were done.

I was terribly afraid of frying the bottom of the board, but it survived unscathed. The ground-plane conducted the heat nicely.

No major goof-ups.

DOH! I don't remember having dropped or hit it hard. Let's blame it on material fatigue. I should've ordered 20. Now I'm one short ;-(

Having a full reel of 1k resistors makes you want to use them. Sometimes this makes you blind to the obvious. Soldering 2 larger parts vs. 8 quite small ones is so much nicer and more time efficient.