1
\$\begingroup\$

I recently started an advanced SMD soldering project. I hqve a couple smaller projects under my belt, so I'm raising the bar this time.

I was pre-tinning some 0603 resistor and capacitor pads and must have accidentally scratched the surface of the pcb between 2 neighbor components. The result is that solder flows between the pads of 2 different neighbor components (resistor and capacitor) and now there is continuity between both pads. When I look at the schematic, these 2 components are actually serially connected, but I'm not sure how to confirm that these are the 2 poles that are meant to be connected. So first question is how to go about making this diagnostic?

Second question is how to fix such a mistake? It's amazing sometimes how easy it is to scratch the surface of a pcb without even noticing. I have searched quite a lot online and haven't found a technique specifically for such a problem.

Thanks enter image description here

from KiCad: enter image description here

\$\endgroup\$
8
  • \$\begingroup\$ Maybe you should include a closed-range photo of of the troubled area of your PCB. So far your description doesn't make much sense. \$\endgroup\$
    – Ale..chenski
    Commented Sep 9, 2017 at 15:24
  • \$\begingroup\$ I just added a picture, you can see the bridge between R3 and C13 \$\endgroup\$
    – matjolic
    Commented Sep 9, 2017 at 15:29
  • \$\begingroup\$ Was there a trace there between R3 and C13 and the solder resist was scratched off? If so, then this isn't a problem. If there was no trace there and it is a legitimate solder bridge, then just touch it with some hot solder wick. I'm more concerned about what the copper-colored thing is between C106 - is that copper? Clean with 97% isopropyl alcohol or lacquer thinner. \$\endgroup\$
    – rdtsc
    Commented Sep 9, 2017 at 15:36
  • \$\begingroup\$ It is still unclear how any problem can occur as result of "scratching". There is no metal between correctly designed and placed pads, so there is nothing to stick on. If you mean solder bridges, then applying more flux and re-touching solves this usual issue. Can you circle the troubled area? If there is a trace between pads, then a bit of solder doesn't hurt. \$\endgroup\$
    – Ale..chenski
    Commented Sep 9, 2017 at 15:37
  • \$\begingroup\$ No, on the mint board, there was no trace between C13 and R3 on the lower left. The scratch on C106 is also my fault (the scratches occured while using solder wick...), so yes that is copper. But there is no continuity between the 2 pads of that component. \$\endgroup\$
    – matjolic
    Commented Sep 9, 2017 at 15:42

2 Answers 2

Reset to default
1
\$\begingroup\$

How to fix:

Clean the two pads with solder-wick. Then cut a small piece of self adhesive Kapton tape (polyimide) and put it over the scratched surface but not over the pads.

The tape will not melt while you solder and will prevent the solder from forming a bridge, so it works as a temporary solder resist.

Now carefully solder your components. Inspect your work using a microscope. If you're happy with the result remove the tape. You can paint a little bit of solder resist over the scratch if you want.

If you don't have a microscope maybe it's time to invest into one. There are affordable USB microscopes. They aren't perfect while you do the work because the picture is delayed, but they are pretty good for inspection.

Regarding your question how to know if the pads should be connected or not: If you don't have the board files and the traces are not visible you can measure a non scratched board to find out.

\$\endgroup\$
4
  • \$\begingroup\$ Thanks for your answer. This makes sense to me and I have some kapton tape lying around. How can I completely remove the existing solder with solderwick? It seems there is always some left on the board. I don't have a microscope yet, but I do have a magnifying glass with LED lighting that does help a lot with these smaller components. I didn't know about solder resist, I will look into that. \$\endgroup\$
    – matjolic
    Commented Sep 9, 2017 at 15:38
  • \$\begingroup\$ @matjolic Heat up your iron and apply a bit of flux onto the solder wick. Then put the wick over the pads and give it a good heating. Your pads will be really clean afterwards. Sure, the will still be a thin layer of solder on the copper, but that won't hurt. If you can still see a bridge between the pads after this treatment, then there is a trace that connects the pads. So no repair is required. \$\endgroup\$
    – Nils Pipenbrinck
    Commented Sep 9, 2017 at 15:44
  • \$\begingroup\$ Sounds good. I just loaded up the kicad for the board and it seems that those pads are meant to be connected inline, so I think I'm good, good riddance. \$\endgroup\$
    – matjolic
    Commented Sep 9, 2017 at 15:47
  • \$\begingroup\$ Note that in a pinch, if you need to cover a trace, a small swipe of the wifey's nail polish works wonders. :) \$\endgroup\$
    – rdtsc
    Commented Sep 9, 2017 at 19:15
1
\$\begingroup\$

As this is a connection which is supposed to exist no repair is really needed for ordinary low voltage benign environment purposes (and in the other cases, you'd need to worry about far more than just this).

The main benefit of having solder mask over a connection that is supposed to exist is to control the flow of solder during a "hands off" process like reflow or hot air rework, where surface tension could cause a small component to move out of position or even stand on end (ie "tombstone"). Especially when the initial placement is provided only by the green strength of unmelted solder paste. The soldermask would normally isolate these two pads, so they can't steal solder from each other and the components are less likely to pull one another out of position (though differing thermal masses at each end of a component can cause issues).

If you are soldering with an iron, this is less likely to occur, and even with hot air you can usually enforce placement with tweezers (often to "bump" as much as to "hold") and care in directing the air stream. Working under a binocular zoom microscope (even in this case under fairly low magnification) makes such a task far less frustrating.

\$\endgroup\$

Not the answer you're looking for? Browse other questions tagged or ask your own question.