FWIW, I deal with bga parts a lot. I'v tried several methods including
hot air and all of them with the exception of one have been problematic.
With hot air, you run the risk that the part will be moved as the air
blows on the part and the solder paste begins to melt. Also, heating a
board with hot air can seperate the layers in the board because
typically, one one side of the board is heated and the other side is
not (And stays cooler).
Anyway, the method I use give about a 95% success rate on high pitch,
high ball count parts. The success rate is close to 100% with low ball
count parts.
The setup I use consists of two dark IR heating elements. One is placed
below the board and the other is placed above the board. The board lays
flat in a stand between the two elements. I measure the temperature of
the board using a thermocoupler. Using this method, I am able to solder
bga parts onto boards at a temperature of 325 F to 350F as opposed to
methods like hot air which require temps well about 450 F (The physics
of how dark IR work allow the use of lower temperatures. The metal
balls and/or pins absorb the heat while the other materials absorb
considerably less).
I've never had a board seperate on me when working on a board that has
been reworked multiple times. Parts do not move as the solder paste
melts. Both the part and board experience less heat related stress and
it shows statistically.
The cost of this setup was less than $150.00 (Excluding the power
supply for the dark IR heating elements. I already had a power supply
on hand that would work well) and took me about 10 hours to design and
setup.
I've been using it for about 2 - 1/2 years with good success.
Best regards
Marvin Dickens
> Egil Kvaleberg wrote:
Placing a BGA device is not a hand solder job. It is done in an oven
along with all the other devices or on a specific, fairly expensive,
piece of equipment.
I remember back in 1999, when visiting a cellular phone startup
company, "the guy" in their lab hand soldered >160 pin micro-BGA
chips with an astonishing success rate (comfortably over 90%). He
also replaced chips this way, as a matter of course.
Seeing that it was possible, made it possible for me to repeat it at
home. The success rate was less, but good enough for my purpose. I
never did this enough times that is became second nature.
The trick is to use a temperature controlled hot air blower with a
continuously selectable air flow. With a suitable, small, nozzle, the
$150 or so Steinels work just fine. You need to get the temperature
right, use a low air flow, make sure everything is clean, and use a
suitable flux. You direct the hot air flow at a 90 degree angle, so
that the chip is free to 'settle' correctly on the pads. You will
develop a 'feel' of what is enough heat. When the surrounding
decoupling caps start to blow away, you have more than enough.
If this will work with conventional BGA, I haven't tried. At the very
least, you need a larger nozzle and more heat.
But what I'm saying is that it *is* possible to hand solder.