In general, the car always sets the AF based on the MAF, TPS, RPM, etc. Call these the lookup table or LUT settings that determine the AF setting. In closed loop mode the ECU changes this based on the O2 sensor sensor. In open loop, it doesn't. Now imagine that the LUT settings are on the rich side. In closed loop the exhaust O2 sensor detects this and the ECU dials back fuel to lean it. In open loop, it uses those LUT settings as-is, no adjustment or feedback.
Those mixture settings came straight from the dyno, before and after runs. Installing the FMS headers leaned it just a bit. The car ran at 14.7 to 15:1 in closed loop mode (light throttle, low to mid RPM). When you opened the throttle the AF ratio immediately shifted to the richer setting. The dyno operator said the original 10:1 which is rich, was actually leaner than Mustang Cobras he tested having the same engine.
I'm very familiar with the Ford EEC programming (EEC-IV in particular) which is why I made the remarks I did.
Let me clarify further, we'll use two example scenarios:
1. Stock:
- 4,800RPM
- 100% TPS (WOT, so open loop)
- 485CFM
- 195F coolant temp
Based on the above, the ECM goes to say E19 which is where 4,800RPM at WOT and 485CFM align and injector pulse is 0.5ms. Timing is 26 degrees based on coolant temp.
2. Modified:
- 4,800RPM
- 100% TPS (WOT, open loop)
- 495CFM
- 195F coolant temp
Based on the above, the ECM goes to E20 which is where 4,800RPM at WOT and 495CFM align and injector pulse is 0.51ms. Timing is 26 degrees based on coolant temp.
In scenario #2, because we increased airflow through the engine, more air was registered by the MAF, which in turn resulted in an increase in injector pulse-width to compensate.
So in both cases, the ECM expects roughly the same A/F because it's supplying the appropriate amount of fuel for the metered amount of air.
When in closed-loop it is constantly searching to go lean so that it can go rich (switching) to try and keep A/F a bit above stoic for economy and will enrich as load increases unless throttle application drives it out of closed-loop and feedback adjustment is paused.
However, you can bugger up the actual A/F achieved while in open loop by changing how much air the engine thinks it is ingesting. A cone filter on the end of the MAF for example, eliminating the airbox, can have a significant impact. A CAI with an elbow is another scenario where "clocking" the MAF to get it back to where it should be by using a dyno is employed. The MAF is calibrated based on its stock location within the stock intake air tract plumbing, so any deviation from that can have an impact on what it registers.
10:1 is pig rich and I'd question the accuracy of the equipment if this is being registered by stock engines. Stock for a 5.0L was around 12:1, my H/C/I 302 was 10:1 with 30lb injectors and a boosted app Pro-M MAF and I was leaving about 20HP on the table running that rich. 24's and a Pro-M for a CAI made a big difference.
Boosted engines will of course be richer (like a Termi) to ward off detonation, but for a naturally aspirated mill, OEM target is usually around 12:1.