I recently read something posted on the web regarding Subaru engines and O2 sensors. Some of the statements contained are not correct or show a lack of understanding about closed loop operation, OE and aftermarket ECUs.

As a quick brush up on terminology here, closed loop mode is where the ECU is adjusting the injector pulse width based on the O2 sensor voltage reading. If the mixture is leaner than 14.7 to 1, the ECU adds pulse width, if richer, the ECU subtracts pulse width. As there is some time involved between sensing and correction, the mixture is never held exactly at the stoichiometric air/fuel ratio of about 14.7 for gasoline. In open loop mode, the ECU ignores the O2 sensor input and determines the pulse width by the other sensor inputs which are usually rpm, MAP or airflow, air temp, water temp, throttle position and sometimes fuel temperature. Contrary to this article, the ECU can quite accurately meter fuel based on these other sensor inputs.

All OE ECUs ignore the O2 sensor input at or near WOT and higher rpms because the engine will not make maximum power at 14.7 to 1 AFRs and EGTs will be highest here as well. All OE ECUs are in open loop under these conditions. It is dangerous for the engine to operate in closed loop at high power settings. This is the exact condition that aircraft installations most often operate under, therefore these engines are often not using the O2 sensor for mixture correction anyway. Idle quality is not dependent on the engine running in closed loop either. Most engines will idle smoother in open loop with slightly richer mixtures than in closed loop.

It should be noted, that contrary to the comments in the article, all modern aftermarket ECUs that I'm aware of have closed loop capability should you wish to use that function. Most also have knock sensing capability. We see some people running ECUs designed for atmo applications in supercharged or turbocharged applications. Obviously this is not the best idea. The biggest bugaboo is likely that no OE ECU was ever designed to run leaded fuel as it is detrimental to O2 sensor life. Fuel additives like TCP may make this viable however we have seen O2 sensor degradation at 25 hours even using TCP with 100LL.

In conclusion, OE ECUs were not designed for aircraft operation but I believe that they will function fine when operating on proper unleaded fuel and applied to the engine they were designed for. In fact many people are doing this successfully. Long term operation on leaded fuels is unproven in my view. Car limp home strategies with partial sensor failures, not only with the O2 sensor but others is also unknown in many cases. While 15 hp might be fine to trundle your car to the next service station, your aircraft will not fly on 15 hp. Programmable ECUs allow you to tailor maps to any application. All these things need to be considered when choosing an ECU for your auto based aircraft power plant.

Update 03/10/04

With an OBD II reader installed on an EJ22 and factory ECU, we looked at the conditions in which the ECU was operating in closed loop. Surprisingly, the system jumped into open loop at quite light throttle and relatively low rpm. We saw that at anything over 11 inches MAP and above 3750 rpm, the ECU was in open loop. Obviously, for aircraft useage, the ECU is rarely using the O2 sensor. This is perhaps good since if the lead fouls the sensor, it would make no difference to the running at normal power settings unless this triggers a limp home mode. The reader also saw 14 degress timing at idle and 34-36 degrees at WOT and high rpm. Timing would go to 44-50 degrees at part throttle and lower rpm. This is strictly for emissions as this has no other function. Airflow indicates the the AFR is around 13 to 1 at WOT.