I am trouble shooting less than good gas mileage (~15.5 mi/gal) in basic city driving under reasonable acceleration conditions.

I have replaced the ECU temp sensor in the back of the cyl head.

I have examined the O2 sensor output on an oscilloscope with a 10 MegOhm probe and the output looks reasonable with excursions between ~ 0.1V & 0.9V with a period of 1-2 sec.

One other symptom which seems anomalous is that when the engine is warm & shutoff and the O2 sensor is cooled, if I start the car, I see an uneven idle with the O2 sensor output at ~0.9V, which indicates to me that the mixture is rich. This uneven idle lasts for ~10 sec, with the O2 sensor remaining at 0.9V with a kind of surging idle from ~600-1000 rpm. After this ~10 sec period, the ECU will start O2 sensor control, and the idle stabilizes and behaves normal.

I assume that the ECU is not using the O2 sensor during the 10 sec as it is waiting for sensor to warmup, but should the sensor output look rich during this period?

Is there some basic mixture setting to be made independent of O2 sensor control?

One other bit of data on the O2 sensor is that with sensor warmed up and ignition on but motor off, output voltage is ~0.48V, whihc I think should be normal.