ok did some autism and traced the hall effect sensors on my P1. see picrel
so basically what's going on is the power for each column sensors is connected in parallel and comes out from a shift register output. shift registers have 8 outputs, first shift register chip has last output feeding into the input of the next. so there's 15 total possible columns they can power, in turn.
each column has 6 hall effect sensors that output through a fucking diode, into opamps configured as buffers, which in turn go into 6 ADC pins on the micro.
so micro always reads the 6 ADCs with data from 6 hall effect sensors on each column, depending which one gets powered by the shift registers.
weird that the outputs of the hc164 powers the chips but whatever, if it works it works.
one thing I didn't yet test is the output of the hall effect sensors. diodes are regular 1n4148 with 0.6Vdrop so there probably is some modding room (?) by using schottky's? curious of the output of the sensors with unpressed switch.
maybe this helps anyone with designing their own HE shit.
ah ye have no clue what model hall effect switch but they have 03DYC on them. these have certain sensitivity and they're matched to magnet strength from switches, most likely. so can't use any switch, gotta be certain sensitivity and magnet certain power else it doesn't work properly. just a guess