Collected works of damage
In this piece I am not going to detail the various repairs done to these
boards, rather I am going to focus only on the damage done. All of these
boards were repaired (as long as the MCPx survived) using techniques shown in
previous articles. I figured why not compress them all into one feature
and show where the original modder went wrong. Hopefully this saves some
of you from a similar fate ;)
Here we have a sample of why it's important to package your board
properly when shipping. Just because it's precious to you does not
mean that the guys loading the truck will handle it nicely. I've
even heard horror stories of boxes marked "fragile" being "tested" as it
were. |
|
While in the thumbnail, this looks bad, in all reality there is no
damage to the D0 point on this board. The issue was a lack of
adhesion, this is caused by not preparing the point to be soldered
correctly. If you are comfortable with using a knife, you can
scrape the copper ring and a bit of the trace to expose fresh metal.
If you are more timid I would recommend a fiberglass pen, designed for
removing soldermask. |
|
Admittedly a lousy picture, this one shows some relatively poor
soldering of the LPC. Some pins have far too little solder to be
considered a good joint, while others look like the joint was not
properly heated prior to application of the solder. |
|
This picture is proof of why I hate conductive ink. The
person who applied this did not seem to understand that the ink is
conductive all the time. Notice the splash around the VIAs
on the left and the bit running over to the next trace in the main
repair area? Both of these were causing problems. The splash
was not fully shorting anything out but was causing decreased resistance
between signals in the area. The over application of the ink in
the actual repair area led to a short between D0 and the trace above it
(A3 if I recall correctly). |
|
Here is a nice detail shot about the D0 A3 short mentioned above. |
|
|
|
Why someone was working on this cap I have no idea. By the
looks of the damage they tried to use a soldering gun or some other much
too hot device. The heavy trace on the left side of the component
is a power line and as such should be fairly difficult to burn away.
Well in this case it was melted into a little ball (top side of damaged
area). I initially thought this was solder, but when it refused to
melt it's true origin became clear. If you face damage such as
this, DO NOT use conductive ink. It can't handle the power
requirements of anything but simple signal lines. |
|
|
|
Here we have an attempted install onto a 1.6 board. The VIAs
were filled with solder mask and the user thought to clean it out with a
rotary tool (or something). Damage to this annular ring was fatal,
but fortunately bridging the bottom side resolved the problem rather
easily. |
|
Two more points from that failed 1.6 install. Both of these
points were still serviceable, though I wouldn't solder anything to
them. I finished this board by rebuilding the backside. |
|
|
|
Here is a wide shot to orient you as to where the following detail
shot comes from |
|
Here we have a small 50 ohm resistor nearly buried in solder.
As this component had been exposed to too much heat it was no longer
serviceable and had to be replaced. Surface mount components are more
easily damaged by heat than their through hole counterparts, this is
especially true for active components. It's a sad day when all the
work you may put into a board is for nothing because you baked a
component. Be wary of how long you apply heat. Heat needs to
be applied long enough to ensure a good joint, but no so long that you
damage the component. |
|
Where a previous photo demonstrated too little solder, this shot
easily demonstrates too much. In addition to the copious amount of
solder present, the flux appears to be of a fairly low quality, possibly
acid based. You should never use an acid based solder on
electronics . . . ever (the only exception deals with nichrome wire).
You can clean excess solder off with q-tips and rubbing alcohol .
Don't over do the pressure when scrubbing, nor should you be too timid
about the amount of alcohol used (avoid any big puddles of the stuff
though). |
|
Finally you need to pay attention to missing components. Here
we see that one of the 10K pull-up resistors for the LPC is missing.
Without these components the board will boot fine from a TSOP, but not
from an LPC modchip. |
|