What exactly is the circuit failure mechanism?

[mholin]

Hello All, first post here. Maybe it's because I never did much consumer electronics design, always commercial so if I needed a big power supply cap it would usually be tantalum or the larger value ceramics because they had superior filtering performance. And assuming these big electrolytics that fail are power supply filter caps, but from what I have read, the failed caps capacitance usually goes down while the ESR goes up. Now if it's a power supply, those caps are in parallel at the power supply output to reduce ripple voltage and supply local circulating currents to the load circuitry. So at the output of a power supply if the filter cap capacitance goes down and series resistance goes up, the effect would be for the capacitor to effectively slowly "disappear" (open) from the circuit. So the voltage ripple gets worse. If it's an analog circuit being supplied, the signals get noisier. Doesn't seem that should cause catastrophic failure. On a digital supply the ripple would have to get pretty bad to start affecting logic switching levels.

Have also heard the leakage current can go up in a failed electrolytic, so the load on the power supply goes up a few micro-amps even milli-amps maybe. These power supplies are usually designed for loads on the order of several hundred milli-amps I would think.

Now if the real failure mechanism is the electrolyte leaks out, the capacitor heats up and physically deforms such that the plates touch resulting in a short, that would make much more sense to me. Then the power supply goes into overload current shut down and you lose that voltage supply most certainly resulting in operational failure.

Idk, maybe it's obvious and I just don't see it, just curious what the actual circuit failure mechanism is.

Thanks All for a great board

[eccerr0r]

Re: What exactly is the circuit failure mechanism?

Having a capacitor short out is catastrophic failure, and usually due to improper use (reverse voltage, over voltage). More likely they fail to degradation of their characteristics.

It's hard to know what part of the circuit fails (first). For the most part, due to the ESR increase, peak voltages will get higher and may trigger some sort of overvoltage circuit to trigger and cause shutdown. It may also simply cause V^2/R heating causing meltdown of components. Noise is a significant problem causing erratic behavior in both analog and digital circuits. Both behave differently when voltage is higher and lower than the design condition especially digital (recall that digital circuits tend to be able to work faster when voltage goes up, and likewise slower when it goes down.)

The leakage current of capacitors is fairly inconsequential except for self heating which causes more liquid electrolyte to vaporize and thus cause runaway. However, some, specifically timing-type circuits require the leakage to be low and if not, will not work properly if the capacitor leaks excessively.

Hopefully this is at least first order explanation at what happens.