The air/fuel ratio is very unstable so I took look at the in and output on the LM324, with ~3,2V on EconoSealIII/18 - pin 13 (WBo2 #1 Nernst Cell).
With a 100nF capacitor (where exactly did you add this?)- after R98, I could all most "level out" the output signal from the LM324 but not enough (my mistake).
Q: It there suppose to bee a square pulse-signal on the input of the LM324 that just "adds on" to the DC-voltage from the O2-sensor?
- 256usec period comes from the nernst pulsing, that's good
- the 3.2V means that there is effectively no (resistive) load on the nernst input. When the sensor is heated, nernst to pump- resistance falls drastically to around 70..120 Ohm which (forming a voltage divider with onboard R144=10k series resistor) drops the nernst amplitude to 30..80mV peak-peak range . (you can try to use a 100 Ohm resistor between nernst and pump- as a table-test). Check that your sensor is heated sufficiently and pump- voltage measures good on the sensor.
Re: Yes I see, that will minimize the amplitude. Is the 1M resistor between WBo2 #1 Nernst Cell and WBo2 Pump- what chances when air/fuel ration chances? ..resulting in higher or lower DC voltage on ADC1?
No! That's just a load so input is not open when sensor is not connected or too cold.
wbo2_ri_target sets the target AC (controlled by heater) and wbo2_nernstdc_target sets the target DC (controlled by pump+). The processor does not actually care about the amplification, but we must set
- wbo2_nernstdc_target so nernst can be maintained as appr. 0.45V above pump-
- and wbo2_ri_target so the AC amplitude is appr. 30..60mV peak to peak
The best values for wbo2_ri_target and wbo2_nernstdc_target (where above conditions are met) depends on nernst amplifier HW, of course.
I suggest start over with the steps of GenBoard/Manual/WBSensorControllerTesting.
Input:
Output: