virtual zero range should be 5v not 3.4v

for better AFR range we shall have wider virtual ground range

No need. The circuit was designed to use the full range of the sensor, which we do.

If someone can quantify an actual problem, I am happy to consider it.

is it about higher current in order to get quicker response time? @dron0gus did you say you understand the problem statement?

Matthew Kennedy 4:29 AM
there is a possible design in which the limited voltage range of a "3.3v native" design limits the available steady state sensing range

4:29
we do not have that problem
4:29
there is a possible sensor for which you could need to overdrive the pump cell briefly in order to "keep up with" a transient
4:30
LSU 4.9 is not that sensor
4:30
the chemistry (diffusion of ions in/out of the sensor's cells) is slow enough that we can stay in closed loop through any transient
4:31
the limiting factor is not an electronic one, it's limited by the tiny chemistry lab that bosch makes us

andrey 4:31 AM
i believe @Dron_Gus suggested logging used current to see how close if at all we get to 100% duty

logging used current

We measure that directly with the pump measurement resistor and amp - then convert to lambda.

We do also directly output the raw pump dac value:

wideband/firmware/can.cpp

Line 158 in 2167016

frame.get().PumpDuty = GetPumpOutputDuty() * 255;