I've made some initial checks, going back into 2023 to see when the discrepancy between the "Npot", "Ne", and "Ni" values started.
At first glance, it appears that the discrepancy started to appear around 12 May 2023, just after the IBIAS update I asked for a few days earlier, and right after the change was commanded on THB (late on 10 May 2023).
That IBAS update in May 2023 was meant to reduce the chances of sensor potential saturation that was happening on most of the EFI sensors across the constellation - the magnitude of IBIAS was much larger than it needed to be.
On TH-B the magnitude of IBIAS for V3 and V4, the two sensors whose potentials are used to compute SC_POT on board on all the SC, was reduced from around -100 nA/sensor to -15 nA/sensor. This certainly changes where on the I-V curve each of those sensors as well as the SC sit, the local value of the sheath resistance of the sensors and SC, and thus the conversion factor from SC_POT to ambient electron density, Npot.
In other words, it's probably not an issue with the EFI instrument performance per se, but rather a need to recalibrate the parameters that one uses to convert from the onboard SC_POT value to Ntot to reflect the different sheath resistances at the revised IBAS settings.
Some or all of these routines may need calibration updates:
From John Bonnell:
Some or all of these routines may need calibration updates:
thm_scpot_density.pro thm_scpot2dens_opt_n.pro thm_scpot2dens.pro
We should also keep Sasha in the loop for the Python side of things.