The data contains measurements from DUT(s) 642 and the following variants:
## [1] "sap" "matrix_wp_off" "matrix_z"
## [4] "matrix_gnd" "matrix_wp_offset" "matrix_wp_single"
## [7] "matrix_wp_30" "matrix_wp_31"
There is quite some variation between SAPs and dies as show in the following plot (gain = 0,full SAPs active):
This variation is deterministic, i.e. each measurement shows the exactly same pattern, but it varies randomly from die to die and SAP to SAP; also for different ASICs, i.e. it is not (only) dependent on the on-board termination of the I/Os.
The following plot show the same SAP for two different gain settings (max and min gain):