An essay in trauma.

An essay in trauma; this now easily is the worst, most badly damaged board I have ever seen . . . bar none.

This board has no business being in the repairs section other than I have no place else to put it for now; I see a hall of shame forthcoming though.

I have no words to fully describe the absolute trauma done to this board. I do try in the following article, but the pictures really tell the story, all I can do is narrate.



Attached, the figurine of a monkey in Persian robes, playing the cymbals. This item, discovered in the vault of the theatre, still in working order; shown here. . . . oh wait, wrong narration . . .

Oh here we go:

Holy Cow!
How on earth can so much damage be done to one PCB! I've never in my life seen a board so badly damaged!
Warning: The following pictures show graphic damage to an Xbox PCB and should not be viewed by people with stomachs too weak for such things. . . Oh and the pics are kind of large as well.
In this first photo I've highlighted three things across five zones. First we'll cover the most fatal part: The yellow portion. This highlights a trace torn back nearly under the MCPx. Even on a good day I'm not sure I could fix this. The area in red near the MCPx should show why: There is so much general trauma to the solder mask that any attempt at soldering would be nearly guaranteed to cause a bridge. While I could throw this under the microscope and try to deposit a new trace and weld it to the old one, I think that would still be marginal at best. The other red highlight illustrates some general damage to D0. While I have not even tried to check continuity to the TSOP, based on a close examination, I'd bet that the connection is toast. Finally on to the two portions highlighted in cyan. In both these cases both pads for the components have been physically torn from the board. It's hard to tell if the pads were heated until the epoxy fell apart or if the damage was purely mechanical aside from heat. My best guess is that R7D3 is heat related failure while C7D3 was predominately physical (non-heat) damage.
This is just a closer shot of D0 so you can see what I'm talking about in the above section as in that shot D0 was outside the plane of focus.
Hopefully this pic is obvious as to what I'm showing. How do you damage something so robust as a eyelet for mechanical retention? (Photo scaled to 50%)
While not impressive, this shot shows trouble waiting to happen. Note all the detritus on the PCB (MCPx is in the photo on the right). All this buildup is on relatively critical signals (based on proximity to the MCPx) and under the right conditions can be at least partially conductive. This low level conductivity can cause erratic behavior which would be quite difficult to troubleshoot.
Here we have a view of a cap near the memory that has clearly been through some sort of re-work. Not sure if I would trust it anymore based on previous slides :-) (inset photo is just another view of the same component)
UPDATE 13OCT2004 0920:
Hmmmm. I seem to have missed a missing component; not that it would be too hard to do on a board this bad. The other memory pads on the topside of the board (where one would solder additional memory), is also missing the larger of the bypass caps. In this case I bet that wouldn't matter as the memory is not installed anyway. I wonder if there is anything else I might have missed?
EOM		 no pic available ::sry::
Now we flip the board over. . . . Hmmm surprisingly little damage. In the Cyan highlight we have a little solder splash that currently would not cause problems, but were it to shift when bounced around in a box being carried to a LAN party, could have disastrous results. In the Yellow highlight we have a component with one pad torn away, this would be fixable with a little effort. Finally in the red highlight we have a component where the solder is still there but the component is not. This is indicative of rough handling and the component leadframe was not as strong as the PCB pads or solder joint.
UPDATE 14OCT2004 0824:
The portion in yellow is from a botched TSOP flash. Normally there is no component on these pads. They are bridged as part of enabling write access to the TSOP. Well now, that still doesn't mean it's not fixable, in fact it's even easier to fix compared to if there was a component missing. (if you wanted to fix it at all). Thanks to A@ron and lordvader129 at Xbox-scene for pointing out my brain fart :o -nB
EOM
While I can often fix anything handed to me this one is an exception. I knew, even before it arrived, that I was going to have a good time laughing when I saw it. There is so much damage I may just try to fix it to see if it is even possible . . . if I ever have a spare couple of hours.

So in keeping with my new tradition, the key takeaways from this are:

  • If you're like the guy who did this: Drop the soldering iron and step away (may want to unplug it first though)
  • If you see a trace starting to lift up from the PCB . . . STOP and send the board to someone who can fix such things. That you got in the predicament of the trace lifting is good enough to make it highly unlikely you can fix it.
  • If you blow a component off the board, no big deal, just replace it. If, on the other hand, you blow away the pads for said component, then the same rules apply as for a lifted trace.
  • If you can top this for damage to a board send it to me and I'll do a write-up on it too. :-)

-nB