The reason modern MLCC (Multi Layer Ceramic Capacitors) fail so often is due to excessive miniaturisation.
Each capacitor is composed of hundreds of microscopic layers of metalised ceramic, and the slightest stress through flexing, thermal effects from soldering and electrostatic discharge, can cause the internal layers to bridge.
That results in the capacitor, which should normally block current, actually conducting instead.
Each capacitor is composed of hundreds of microscopic layers of metalised ceramic, and the slightest stress through flexing, thermal effects from soldering and electrostatic discharge, can cause the internal layers to bridge.
That results in the capacitor, which should normally block current, actually conducting instead.
Have you ever wondered why modern ceramic pastors tend to feel so often destroying things like laptops and phones? Well, the answer is that because of miniaturization, they're cramming such a high capacitance into a small era. These little capacitors are actually made of lots of layers. Literally hundreds of layers of metallization and insulation. And when these are soldered, or when the circuit board is bent or anything is applied that causes stress, either it can chip or it can crack.
And if it cracks, these layers Bridge together and the thing shorts out and gets red hot.
had one fail on an old ms-7021 motherboard, it went red hot and even left scorch marks!
You also need to be mindful of this when soldering them, as the thermal stress from uneven heating can crack them!
Those things fail just by looking at them, thermal expansion is enough to kill them, you don't even need additional mechanichal stress. I can't even count the things that i saw failing thanks to one of those. But i gotta say they're easy money if you repair electronics.
Anyone remember the DPS5015 dc dc converters that could catch fire because of an capacitor soldered directly between the output screw terminals. The cap would crack when thightening the scews. The problen was fixed a couple years ago by removing the cap. The DPS5015 is in my opinion the best and most robust lab powersupply dc dc converter.
Ty
A common failure mechanism is electromigration at the end points. There is a high electric field near the ends, especially so on caps operated near their rated voltage. In Clive's diagram, one of the odd "fingers" from the left will grow towards, and eventually short to, the vertical termination at the right that is common to the even fingers.
This is a dead short and can often be found by using your finger tips….when you burn your finger, you can stop the search!
At first, I thought this was Grady's channel and we were discussing road and bridge design fails.
Thanks.
reheating solder joint on one end and not other, is also a really good method to fall them.
I've always wondered: Are we in for a new Capacitor Plague like all the devices from the 90s?
Except recapping 500 smd caps is gonna be a lot more work 😅
Shorts can be very bad for your equipment, and as you point out Clive they can be potentially dangerous. That's why I never wear them. 😉
Nice graphic, actually. It makes a lot of sense. thx
They can also go OC or intermittent. I invented a method to find a broken cap on a PSU line with hundreds of caps (due to failing ICT). Basically you measure the line capacitance on an LCR meter and heat each cap or area with a hot air pencil jet. The meter reading suddenly changes when the hot air hits the cracked cap.
Tower Bridge in the middle ?