Stomp test reports 1444, no CEL. Operation at speed, part throttle seems to be ok, however long term records are showing a statistically-significant decrease in fuel mileage. Emissions were well within limits at the last test, about 7 months ago (but the weather was warm).

The problem is that the engine frequently stalls. I've checked and cleaned the ICV which I've mostly ruled out because when the engine doesn't stall it idles properly at 700RPM with varying loads. I noticed one thing that may be significant -- if I hold the RPM at 1100-1200 and close the thottle, the fuel consumption meter will spike to the left as if the fuel is being cut off (TPS is therefore working). When this happens, the engine idles too slowly or stalls. Now ordinarily the fuel cutoff is not operative unless the engine is running at ~1600RPM or above. It shouldn't cut off at throttle closure if the engine is only at 1200RPM. If the meter doesn't spike the engine returns to 700RPM as it should. Performing this exercise repeatedly shows that the problem occurs unpredictably and about half the time. It might work right 4 times in a row, and then act up the next 4 times.

Also, about 150 miles ago I dumped in a bottle of Techron with a full tank of gas. The engine has had no treatment like this for years, and there was no reason to think it needed any. It's been running mostly OK with good quality gas and Mobil 1.

So, given this behavior, I'm thinking AFM. This car has ~158K on it and a manual trans. That AFM has had a lot of activity and wear in its life. Coupled with the 15F air temperature, I'm wondering if it's not hanging open occasionally and/or is slow in closing.

I'd like to try an ECU reset, hopefully by pulling its fuse instead of disconnecting the battery. Anyone know if one or more fuses can cause a reset?

What do you guys think? AFM? Something else?