Forty key cycles to reset the fault codes I think (maybe it was 80???).
If one replaces an AMM with an adaptation of 1.32 (normal limits for late HFM and all ME cars) with one that needs no fuel correction, the fuel trim will be 32 percent too rich. The proper job would have one clear the adaptation back to 1.00 on the lower and upper partial and 0.00ms on idle.
The ML I did this morning I drove with the light on to see performance. As with most warm addapted engines it idled fine and moved off fine. Above 40% throttle the car started to go slower. This is a classis case and I know the answer already but I do it properly. I read the fault codes P0170 and P0173 (naturally) and I go to actual values to view the adaptation values = 1.32. So with the symptoms my evaluation is confirmed. I then clear the adaptation back to 1.00 and drive it about 3 miles. It has then partially readapted back to 1.11.
At this point I replaced the AMM and left the adaptation at the 1.11. I drove the car 10 miles and again read the adaptation of 0.99. All the driving symptoms are gone full power to the redline first and second gear.
This is final proof of the repair. Notice I also didn't clear the adaptation values from the 1.11 number of the short old AMM test. if I had left the numbers at 1.32 they also would have readapted back to .99 in 8-10 miles. The problem is that initially the mixture could be so rich it could set misfire codes or others before it did its own readaptation.
For HFM cars the readaptation could take 100 miles of driving so a controller stuck with at limit adaptation values can easily have other problems before it finds its way home.
Bosch Master, ASE Master, L1
33 years MB technician