Wrong MAF CO settings on Motronic 2.5 Mondial T cars can cause fires...

My project to add O2's to my '89 Euro stock non cat Mondial T finished today with confirmation of a number of things.
1) I never had a confirmation of the 383 OHM setting for the MAF
2) The MAFs are not connected in series with the injectors
3) The MAF CO screw value is not used by the ECU unless the ECU goes to limp-home mode when one of the key sensor values deviates more than 20% (on the motronic 2.5) from the table values OR the ECU pin 31 is open (not grounded) indicating to the ECU to ONLY use the table values (eg. for non cat stock euro cars)

You can follow the below links to see how it was all done:
http://www.ferrarichat.com/forum/sho...d.php?t=202434

To find the ideal CO screw value for each bank of your engine (that results in a CO output of .7% and HC of less than 250 ppm) follow the last half of the last reply in this post:
http://www.ferrarichat.com/forum/sho...d.php?t=202441

By doing this you ensure that if and when your O2 sensors fail then you don't blow your engine through an over lean situation OR you don't burn your cat (and possibly car) through an over rich situation. With Motronic 2.5 cars not having Check Engine Lights wired in (folks there is no connection for these) if you dropped an O2 and your CO screw was not well set you could potentially not notice for long enough that other bigger problems arose. The CAT Thermocouple device flashing the Slow-down lights was designed to warn people when a fire risk was present from overheating cats. 2.7 cars had the CELS.

383 Ohms as being default:
I spoke to Gabriele Melilli at Ferrari in Maranello, Italy on a recent factory visit. He was 61 years old and worked on many F-car engines, including the F-119, and he told me the MAFs were set to individual values as would produce the numbers referenced in the user's manuals. He said many mechanics believe there is an empirical value and that has caused many problems.

Suspicion that MAFs were wired in series with the injectors - not the case.
The MAF produces a voltage for a given airmass (voltage needed to ensure the wire is kept at a given temperature counteracting the cooling effect of the air flowing through it). The converts the analog signal to a digital one and uses this as a parameter (amongst others) to identify a fueling coefficient which then triggers the injectors to open for a given amount of time. I messed around significantly with the MAF units (see the post linked above)

(This means that you can replace the MAFs with MAPs quite easily and you can monitor the MAF output voltage for a given temperature and RPM by using the voltage ouput on pin 3)

Finding the ideal resistance for your CO screw:
When is the CO setting used?
ONLY in open loop NOT EVEN at WOT
1) The gas analysis numbers don't change when the CO screw is varied during the test cycle in closed loop, or even if the engine is restarted. Though after an ECU reset there is a small variance which returns to the pre-reset value after 10-15 minutes (proving the ECU's do learn something).
2) The GAS analysis numbers at 383 ohms (with ECU pin 31 disconnected) were way off the charts with a VERY rich mixture (HC's were in the 900 ppm range), and WOULD vary with the CO screw during the test cycle.
3) Putting the car on a rolling road with a resistance preventing it from red-lining, changing the CO screw setting at WOT resulted in no changes to the gas output (boy was that noisy though)

CAT testing information:
I put the MAFs to 383 ohms and ran open loop. The cats ran at 890 (1-4) 950 (5-8) deg. C - I don't have the thermocouples which would have triggered surely. Incidentally at WOT they stabilized at 730 with less than 5 deg. C differential.

I tuned the CO screw in open loop with a gas anaylzer tuning the CO/HC until in spec according to the manual using the richest end of the range. The resulting ohm reading (using pins 1&6) is the "right" number for that bank of cylinders - These need NOT be the same, though widely differing values indicate you may have some other reason causing the imbalance, such as fuel pressure differences, injector performance differences, valves etc.

Using the gas analyser, the CO screw setting that produced the CO and HC numbers in the range as printed in the manual were 520 ohms for bank 1-4 and 430 ohms for 5-8. The Cat temperatures dropped to below 600 Deg. I fine tuned the screws further using the procedure in the post I enclosed above.

With this range the COMBINED HC fell to 250 ppm and CO to .7. When the O2 sensors were looped in (I put a small switch on the wire grounding pin 31 on the ECU), this all averaged out to .6 ppm CO and 160 ppm HC. This would leave the cats to reduce the HC by 110 ppm in closed loop and by 200 ppm in open loop. This is very little unburned fuel (thus probability of overheat) compared to the 800 ppm which would be the case with the 383 ohms. The post cat numbers were .07% CO and .03 HC - I think cleaner than the air going into the engine....

Lower Cat temperatures (still within operating range) = lower risk of fires.