LTC3588 energy harvesting regulator

[LukeW]

Consider the attached schematic.

It does work, but after prolonged operation it occasionally, intermittently, kills its LTC3588s.

Any idea why the chips are dying sometimes? And how can this failure be prevented?

Overvoltage from the pizeoelectric transducer is a possibility, but the chip is supposed to have protective internal clamps to protect against this.

The only goal of the circuit is to light up a LED, powered by mechanical energy harvesting. A large output capacitor is used, along with a very high brightness, high efficiency LED and a very small current through the LED, in order to allow the capacitor to carry the load during occasional intermittency in the mechanical excitation, and allow the LED to smoothly carry on shining for a couple of seconds without the capacitor being discharged too fast.

[David Hess]

The first thing that comes to mind is a common mode voltage spike, maybe caused by ESD, between the cantilever and the LTC3588 ground.  This could also enter from the LED side.