[moved] capacitors shorting in our 4 layer pcb

[sirani]

Hello

I am facing problem in my design which operates on 12 volts, the product is coming back from market after installation with the complaint that capacitors which are mainly on power output line of 3.3v are shorting. We have used 16v smd capacitors in package like 0402 and the pcb is 4 layers. We see that capacitors get shorted and when we replace them they start working. Can anyone suggest possible reason, we are not able to figure out the issue. Thanks in anticipation.

 

[andre_luis]

Unless the problem is due to poor quality of the capacitor, you need to provide more details about what is connected at the 3.3v bus.

 

[KlausST]

Hi,

Capacitor vendor and exact type.
Temperature
Frequency
Current ripple
Voltage, max incl. Spikes

Klaus

 

[Dan Mills]

Mechanical environment?
Small smt parts do not like the board flexing (The automotive guys actually have 'soft termination' caps to reduce this problem).

Regards, Dan.

 

[KlausST]

Hi,

Small smt parts do not like the board flexing

Oh, yes. Near a V-cut for example...

Klaus

 

[marce]

I would also go with mechanical, seen this happen after testing, board getting stressed during installation.
Record the position of all shorted capacitors as well as any other factors....

 

[andre_luis]

Small smt parts do not like the board flexing

Indeed, but this would open the capacitor from the circuit instead of short circuit it. The reasons pointed by Klaus seems likely to be the cause.

 

[Mattylad]

Someone fitted diodes instead of capacitors

 

[marce]

Indeed, but this would open the capacitor from the circuit instead of short circuit it. The reasons pointed by Klaus seems likely to be the cause.

There is no guarantee the cap will fail open circuit some do fail short circuit... For realy critical designs MLCC caps are placed in pairs 90 degrees to each other daisy chained electronically so that if one cracks due to board flexing you don't get a short across a power line (often where these are used) as the other in a different plane is less likely to *****. You cannot assume an open circuit failure with MLCC capacitors you must always assume worse case a short circuit. I have also seen this problem in real life with v-scored boards (one of the most violent de-panelising processes).

**broken link removed**

- - - Updated - - -

Moderators its a bit of a joke when we cant discuss failure modes in MLCC caps because the word c r a c k is censored!

 

[binlch.lv]

do your cap is neer board edge or is Very heavy forced .