EEVblog Electronics Community Forum
Electronics => Repair => Topic started by: palpurul on March 24, 2022, 06:28:36 am
-
I discovered an unexpected short when I plugged the USB to my board that I developed, there was dead short from 5V to ground. This was totally unexpected because I used this board many times without a single issue.
As it turns out short was caused by this 0805 capacitor shown in the pics. Notice the solder bridge between terminals, that was the cause of short circuit.
Over time a solder bridge is formed between capacitors terminal causing short between 5V and ground as you see in the first picture. How is that even possible?
Has anyone experienced this before?
Short between capacitor terminals:
(https://i.ibb.co/C8tv1pc/IMG-20220319-175919-551.jpg)
The actual place of the capacitor in the board.
(https://i.ibb.co/L9znVHH/Inked150743679-856e5686-c5f7-48ab-8af0-7fbeebe9454c-LI.jpg)
-
Unless overvolted, MLCC usually fail because of cracking due to mechanical stress. What you likely see is that cap formed a short circuit inside and then started cooking as high current started flowing through it. Often it's not a 0 Ohm short, but something like a few 100 Ohms which then drops over time.
-
Unless overvolted, MLCC usually fail because of cracking due to mechanical stress. What you likely see is that cap formed a short circuit inside and then started cooking as high current started flowing through it. Often it's not a 0 Ohm short, but something like a few 100 Ohms which then drops over time.
It was a dead short (was measuring sub 1-ohm) caused by that solder bridge somehow formed over time. The capacitor was fine, after cleaning the solder bridge under it, it measured open circuit.
-
Then I guess it was a soldering defect which was on a verge of shorting.
-
Then I guess it was a soldering defect which was on a verge of shorting.
Probably. But what would cause it to complete the bridge? It was just sitting in the shelf :-//
-
Might be moisture, or PCB was bent a little bit, just enough to complete the circuit.
-
Well, it makes sense.
Probably a small bend or change in temperature caused it finish the solder bridge.
-
Maybe solder whiskers
-
If I remember that NASA report on tin whiskers correctly, the whiskers are too thin to carry a decent amount of current.
They are able to compromise signals, but will burn off under load.
-
Maybe solder whiskers
A key thing: they’re tin whiskers. (And by “tin”, we mean the element, not “tin” as a synonym for “solder”.) IIRC, practically all tin whiskers form from pure tin plating. It doesn’t take much of something else to prevent whisker formation. Consequently, as best I can tell, tin whiskers forming from lead-free solder has not turned out to be the problem it was predicted to be.
-
Silver also forms whiskers, like tin, just not as common.
-
The photo looks like the solder spattered or dripped, rather than grew as a whisker.
Either a manufacturing fault (that took a while to make contact across the cap), or something else failed and spattered the solder which then stuck to the capacitor terminations.
-
I watch lots of youtube pc fix video's and capacitor failure is common along with mosfet failures.
I havent seen one fail like yours before.
Sometimes if the blob was already there and not quite bridging then unleaded solder can whisker over time.
-
I agree with two last posts.
-
Are you sure it looked like that before you removed it?
Is this board near anywhere where you might be soldering and flicking around hot molten solder?
-
Are you sure it looked like that before you removed it?
Is this board near anywhere where you might be soldering and flicking around hot molten solder?
Actually this could be the case. It could be the solder which dropped from a soldering iron tip while soldering something nearby.
-
I had a stereo with SMD parts, and I guess for ground plane reasons, they had ~0.2mm traces of solder+mask, passing under the middle of all SMD parts. And during soldering/de-soldering for troubleshooting the stereo, and bad flux, I ended up with a bunch of unmasked edges, right under the RC parts and fine pitch quad op-amps. What a nightmare.
I wish I had a good way to test SMD caps, in circuit. Or for some stuff, I should pre-heat the PCB in an oven, then use air and just mass decap a few things I have. I don't have anything I suspect bad SMD caps on really, but in general, if WW3 starts, it will be worth the time to check all SMD caps.
Probably the most import thing's I've SMD de-capped, was the circuits for the controller chips, of computer SMPS. Then I got all the sizes, to plug into LTspice and try and make ideal models of the controller chips.
-
this smd capacitor is already unsoldered from the PCB - so all assumptions are nonsense..
-
Probably. But what would cause it to complete the bridge? It was just sitting in the shelf :-//
Its possibly a solder splash from manufacture.
Unleaded solder can whisker over time.
Satellite engineers found to their horror circuits were failing after a few years.
-
Probably. But what would cause it to complete the bridge? It was just sitting in the shelf :-//
Its possibly a solder splash from manufacture.
Unleaded solder can whisker over time.
Satellite engineers found to their horror circuits were failing after a few years.
Very interesting.
I use leaded solder, but still an interesting information.
-
Looks like someone put a blob of solder on the cap while removing it. Original board looks like a stock photo. Whatever information that was on the board to diagnose the failure has been destroyed. Put on another cap and hope for the best?
-
Well I recently fixed a attenuator which had a 1pF ceramic C0G through hole capacitor that was reading 500 ohms DC resistance. Not quite a short but I wonder if it was getting there.. (this had nothing to do with the solder).
-
Although I never verified it we had a few failures in one PWM amplifier due to a grouth
inside the end caps of a ceramic capacitor which caused the cap to go leaky 500 ohms
to shorted.
Resoldering would temporarly fix it.... good for troubleshooting.
Jeff
-
Unleaded solder can whisker over time.
Satellite engineers found to their horror circuits were failing after a few years.
No, it doesn’t. The whiskers come from pure tin plating on component leads. Lead-free solder is never just 100% tin, and even small amounts of alloying eliminates whisker formation.
https://www.indium.com/blog/pb-free-solders-are-not-the-cause-of-tin-whiskers.php (https://www.indium.com/blog/pb-free-solders-are-not-the-cause-of-tin-whiskers.php)
-
I sometimes get solder splashes on my pcb's but I make hundreds.
Probably from not wiping solder bit often enough.
Once there is a solder splash with unleaded solder it can start to whisker and cause a short.