I'm wanting to do some Piezo TDOA detection using a Teensy LC (if it matters). I'm planning up to about 5 piezos - 3 for TDOA and 2 others for external noise rejection. The 3 TDOA sensors are spaced 120 degrees around the perimiter of the detection zone (ie, all hits are inside). The 2 noise rejection sensors was an add on idea - if the timing resolution is feasible. I'm hoping for ~us level timing accuracy. Spacing sensors 4/5 outside the detection zone on the mounting - about 2" apart would give about a 10us delta between them - if sensor 5 fires first, then the impulse is external and I can mask anything incoming for the next 60us or so. Time difference between the 3 TDOA sensors would be between 0 (center hit) and 100us (hitting on of the sensors). I don't think I need super precise results because I have three pairs to work with (S1-2, S1-3, S2-3) so I can estimate the approximate radial distance from that.

Once I've detected an event from timing info, I need to ADC one of the sensors (pick one). To get the interrupt I was planning to amplify the snot out of the analog signal, precision rectify, clamp and filter as needed

IS there a minimum duration that a digital input pin must be held steady in order to reliably trigger a pin change interrupt? Must I sample/hold/peak detect/low pass/latch my processes analog signal feeding the digital pin in order to get reliable triggering?