Methods for finding faulty chip capacitors

[varunme]

what are the Methods for finding faulty chip capacitors other than visual inspecting the smd part in a buch of capacitors in a PCB?, the PCB indicates short circuits but its not easy to identify a faulty cap in the circuit..., Is there any trick to identify the part that is failed ?

 

[pinout]

Not sure what the 'up to date' methods are, but in days gone by I used an HP current sensing probe (like a ferrite bead/coil at the end of a pen) and this could be set a particular sensitivity and trace the current path. It could locate sc caps, solder bridges, PCB whiskers, was a great tool. From memory it had a generator that could safely energise the test tracks without applying power.

Today I guess IR cameras are a good faster alternative method.

Don't know if there isa modern version of the HP probe?

 

[varunme]

eagerly waiting for more valuable inputs ....

 

[betwixt]

Similar to Pinout's method is a low resistance measuring meter. They generally use a 4-wire measurement to allow very low resistances to be traced and a tone generator giving different pitch per milliohm measured. By probing along tracks it is possible to find the lowest resistance point which is of course where the short circuit is. They are accurate to about 5mm along 1mm wide tracks so isolating an individual shorted capacitor, even if many are in parallel such as supply decouplers is quite easy. They are of less use when a PCB has an internal power plane and ground plane because they inherently give less resistance beween probeable points.

Brian.

 

[pinout]

How about IR camera?
**broken link removed** is an interesting article that I might investigate.

Temp rise does not necessarily indicate the actual fault but the vicinity, consider a true dead short capacitor which hence has no power dissipation/ temp rise, but the local tracks will .

 

[varunme]

Bitwix, is that method is of use in a system like personal computer motherboard ?, as it is having ground plane and power plane ?

pinout, for that technique, we have to apply power to the pcb isnt it ?, if there is short, it will not damage the components ?

 

[betwixt]

Yes, like most PC motherboards. Because the ground and power layers cover most of the board area, they present very low resistance everywhere, it's good for design but makes fault finding more difficult.

I'm not sure the IR method would work. I agree that a hot component could be found easily using IR but in the case where power is dissipated along long traces it wil be hottest where the resistance is highest. Usually this would be anywhere the copper was slightly narrower or thinner or at connectors, not necessarily where the faulty component was.

Brian.