Imagine I have a Pmos, witch runs a high current. I want to know if there is any leakage between the bulk of my Pmos (the Nwell) and the p-subsrate?
All I know now is that a high energy holes in the channel can break a bond in the N-well, and the resulting pair is evacuated by the bulk contact and the channel. But is there also any injection from the Nwell into the substrate??
The NWell should be reverse biased and the normal junction
leakage will always be present. However if you manage to
take the PMOS source above the well potential the vertical
PNP may turn on and multiply the S-B injected current with
some S-sub current @ gain (probably not much gain, but still).
Drain hot carrier / impact ionization might also increase the
well-sub current although these are usually pretty small
currents. You want to do a good job with body ties to
make these mechanisms stay quiet. Refer to latchup rules.