Argh, I lifted a solder pad

[Grinder]

While trying to recap an SE/30, I managed to lift a solder pad off of
the motherboard. It has broken right at the point where the pad becomes
a trace, and I can see a little bit of copper with my 30x magnifier.
The trace runs under a chip a couple of mm later, so I can't see any
other pad or connection that I know attaches to that trace.

A surface-mounted tantalum capacitor is supposed to lay on that pad, but
I have my doubts that the tiny bit of copper I see can make a
connection. Also, I'm wary of try to scrape what little trace I see to
get more exposed copper.

Any ideas how I can fix this mess?

 

[Grinder]

I suppose you could try to find where it ends up and "green wire" it...

but it's probably time for a new mobo.

I fooled around and got one re-capped a board and replaced all the
obviously bad caps...
but it was still flaky so to fix it properly I would have had to replace all
the caps...
that would have cost more than a new board.

Within the last year I picked up a few new boards from NewEgg for just $50

This isn't a very practical project. It's the PC I had during my
undergrad, I've been slowly restoring it. Sadly, it looks like I may
have just euthanized it.

 

[Paul]

This isn't a very practical project. It's the PC I had during my
undergrad, I've been slowly restoring it. Sadly, it looks like I may
have just euthanized it.

Can you epoxy the cap to the motherboard first, then wire it up ?

The purpose of the epoxy, would be for mechanical strength. Then
the wiring job you do, from the lifted track to the tantalum pad,
won't have to provide mechanical support.

This wouldn't be a viable solution, for a cap in a high frequency
circuit (such as PCI express differential, SATA or the like). There
are some ceramic coupling caps in circuits like that, where fooling
with the trace would ruin the signal quality. But for at least some
of the bypass caps on the board, high frequency characteristics of
the wiring aren't quite as important. If you feel simple wire is
not good enough, you could also solder a piece of copper braid,
from the cap to the motherboard (a wider conductor).

In casual conversation at work, someone mentioned it is possible to
repair copper lands on a PCB, but I've never seen that in practice.

Paul

 

[Grinder]

Can you epoxy the cap to the motherboard first, then wire it up ?

The purpose of the epoxy, would be for mechanical strength. Then
the wiring job you do, from the lifted track to the tantalum pad,
won't have to provide mechanical support.

This wouldn't be a viable solution, for a cap in a high frequency
circuit (such as PCI express differential, SATA or the like). There
are some ceramic coupling caps in circuits like that, where fooling
with the trace would ruin the signal quality. But for at least some
of the bypass caps on the board, high frequency characteristics of
the wiring aren't quite as important. If you feel simple wire is
not good enough, you could also solder a piece of copper braid,
from the cap to the motherboard (a wider conductor).

In casual conversation at work, someone mentioned it is possible to
repair copper lands on a PCB, but I've never seen that in practice.

I don't see how I can even wire it up. There's only about 2-3 mm of
trace before it runs under a chip, so I don't know where there's a pin
or another pad I can solder a lead to.

 

[Paul]

I don't see how I can even wire it up. There's only about 2-3 mm of
trace before it runs under a chip, so I don't know where there's a pin
or another pad I can solder a lead to.

Peel up the tip of the end of the foil, tin it, and reflow
your wire to it.

Paul

 

[VanguardLH]

While trying to recap an SE/30, I managed to lift a solder pad off of
the motherboard. It has broken right at the point where the pad becomes
a trace, and I can see a little bit of copper with my 30x magnifier.
The trace runs under a chip a couple of mm later, so I can't see any
other pad or connection that I know attaches to that trace.

A surface-mounted tantalum capacitor is supposed to lay on that pad, but
I have my doubts that the tiny bit of copper I see can make a
connection. Also, I'm wary of try to scrape what little trace I see to
get more exposed copper.

Any ideas how I can fix this mess?

Use a continuity test to see if the foil trace goes to one of the nearby
chip's pins. Make sure to reverse the continuity tester to make sure
there isn't a diode somewhere and you're testing on the other side of
it. If the trace goes to a chip pin (and it isn't a surface mount chip
but has legs for the pins), there's your solder point.

Otherwise, look on the backside of the chip area to see if there was a
bleedthrough to being the trace to the other side. Again check for
continuity to see if any of those solitary blivets (for the feedthrough)
can be used. Problem is that you'll need a short path through the mobo
to bring a wire through to the other side.

One time, the trace foil was all I had to use but the chip had legs (not
surface mounted). I used a needle-pointed wire snip to cut the legs on
each side of the foil to make then thinner and open a larger gap between
them. I then solder a short bared end of wire-wrap wire to the foil and
then soldered the other end of it to the capacitor. I then used hotglue
to hold the cap in place since now it only had one leg for support. I
don't remember what I did to remove the varnish/plastic coating over the
trace so I could solder to it. I don't remember doing any scraping,
like with a Xacto blade, as I recall thinking that was too rough for
such a little trace foil. It's been too long since parts were more
expensive than my hourly rate. It's often cheaper just to replace.

Maybe taking a picture and uploading it to online storage to provide a
link to it here would help others in seeing what happened and what that
remnant trace looks like along with the nearby chip.

 

[Grinder]

Use a continuity test to see if the foil trace goes to one of the nearby
chip's pins. Make sure to reverse the continuity tester to make sure
there isn't a diode somewhere and you're testing on the other side of
it. If the trace goes to a chip pin (and it isn't a surface mount chip
but has legs for the pins), there's your solder point.

Otherwise, look on the backside of the chip area to see if there was a
bleedthrough to being the trace to the other side. Again check for
continuity to see if any of those solitary blivets (for the feedthrough)
can be used. Problem is that you'll need a short path through the mobo
to bring a wire through to the other side.

One time, the trace foil was all I had to use but the chip had legs (not
surface mounted). I used a needle-pointed wire snip to cut the legs on
each side of the foil to make then thinner and open a larger gap between
them. I then solder a short bared end of wire-wrap wire to the foil and
then soldered the other end of it to the capacitor. I then used hotglue
to hold the cap in place since now it only had one leg for support. I
don't remember what I did to remove the varnish/plastic coating over the
trace so I could solder to it. I don't remember doing any scraping,
like with a Xacto blade, as I recall thinking that was too rough for
such a little trace foil. It's been too long since parts were more
expensive than my hourly rate. It's often cheaper just to replace.

Maybe taking a picture and uploading it to online storage to provide a
link to it here would help others in seeing what happened and what that
remnant trace looks like along with the nearby chip.

For better or worse, I took my shot. I was able to scrape the lacquer
of the top of the trace and get it tinned. The replacement caps are a
bit longer than the originals, so was has able to reach to that trace.

It's had for me to tell if I have good solder flow onto the trace, but
it looks ok. I'll reassemble tomorrow and see if it's a go.

Your advice is good. If I can't get this present plan to work, I'll
like try what you advise.

 

[Steve]

For better or worse, I took my shot. I was able to scrape the lacquer of
the top of the trace and get it tinned. The replacement caps are a bit
longer than the originals, so was has able to reach to that trace.

It's had for me to tell if I have good solder flow onto the trace, but it
looks ok. I'll reassemble tomorrow and see if it's a go.

Your advice is good. If I can't get this present plan to work, I'll like
try what you advise.

The board may not miss one cap, a lot are put there 'just in case'.