0
\$\begingroup\$

I’ve been soldering components to my PCB; all has been good, but recently, after soldering a big electrolytic through hole cap, the power and gnd plane were shorted. This persisted even after removing the cap.

Is it possible that the plated throughole has a gap in between layers where excess soldered flowed through and connected to GND in one of the internal layers? This board was made through OSH Park (it’s a 4 layer board).

EDIT: Solved the problem by drilling through the center of the throughole with an oversized drill. I got to the assumption that the throughole was causing the short by using the suggested method of applying current to go through the power plane/GND and seeing that the lowest voltage was in the vicinity of the hole.

\$\endgroup\$
  • 1
    \$\begingroup\$ Are you sure the short wasn't there before? Test a spare board and carefully examine the Gerbers. \$\endgroup\$ – Brian Drummond May 12 '18 at 10:51
  • 1
    \$\begingroup\$ When you used your test meter, presumably in continuity mode, did you wait a while? If you put a high capacitance across a continuity tester, it will read a short circuit for a while until the capacitor has charged up. \$\endgroup\$ – Tom Carpenter May 12 '18 at 17:23
  • \$\begingroup\$ This doesn't directly answer your question, but if the short is relatively high resistance (i.e. a sliver of copper), you can put some high current through it (provided the PCB is not populated or the components on it can take the voltage) to look for hot spots. I've done this on PCBs and found hotspots using standard issue fingertips MK1. If you have a thermal camera I bet its easier. \$\endgroup\$ – Wesley Lee May 12 '18 at 17:37
  • \$\begingroup\$ What are your power plane clearance rules? \$\endgroup\$ – Spehro Pefhany May 12 '18 at 18:10
  • \$\begingroup\$ The short was not there in the original boards. \$\endgroup\$ – Michael E May 14 '18 at 22:19
1
\$\begingroup\$

The answer to your question is no, that is not normal or possible with a board that is well designed and well-made. The plated through hole would not be permeable and there would be no possibility of solder flowing through to a plane layer. It is possible for this to happen if the board is not well made or not well designed with adequate clearance around plated through-holes. If the plated through hole really did short to the plane, then you can try drilling out the plated through hole with a small drill. If that makes the short go away, then it was most likely a plated through-hole problem.

In general, if you want to get to the bottom of this, you will have to be systematic about it. If you haven't already, do a careful visual inspection of the whole board looking for solder blobs.

A careful inspection of the board layout might be a good idea also, to look for places where power and ground are, perhaps, closer than they should be.

If there are other components on the board, one of them could be responsible for the short. Removing all the components might be a good idea. Check the short after each removal. If the short goes away as you remove components, then it could be a solder bridge hidden under a component or a bad component. Or maybe a backwards diode.

I don't know what equipment you have, but you could also try applying a high current to the short and use a thermal imaging camera to find a hot-spot on the board. Most likely the short area will appear as a hot spot because the current density will be much higher in that area. It is a current bottleneck. It might only be a few degrees hotter than the rest of the board, or it might actually get quite hot.

If the short is definitely in the board, and the board is large, you can repeatedly cut the board until you are left with only a small segment that has the short in it. Then this area can be examined very carefully, and hopefully the short can be discovered.

Hopefully one or more of these ideas will help you.

\$\endgroup\$
0
\$\begingroup\$

I doubt it, it's more likely that there's a solder blob somewhere shorting earth to ground on the exterior of the PCB.

The reason for the design rules (keep-out zones around unconnected vias, inter-track spacing etc, is to allow some inaccuracy during manufacture without it causing shorts or opens where they don't belong.

\$\endgroup\$
  • \$\begingroup\$ I’m aware of the importance of DRCs, but this has actually happened before in an OSH park board (shorting through a plated hole) so wanted to see if this was a common manufacturing failure on their part. \$\endgroup\$ – Michael E May 12 '18 at 14:06
0
\$\begingroup\$

This persisted even after removing the cap.

How does the board look like? It is possible to overheat the PCB material to the point where it burns away and leave shorted copper layers.

Beefy electrolytoc cap can cause this if the soldering iron cannot put enough heat into the solder joint fast enough due to high heat transfer through GND/power planes. Note that there should be visible damage in this case.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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