This seemed like a fitting topic for my troubles too.
Some background: My car, 2000 model 1.8TS with 60k miles on clock, had a botched up cambelt change last spring (the timing was off by 8 degrees) and diagnosing this took for various reasons all summer.
Now that the timing problem was rectified, the performance came back to normal, but bad mpg and cold start troubles remained. Car was averaging 10,5l per 100km and would die spluttering about every other cold start. 2nd starting attempt usually was succesful and sometimes black vapour would come out of exhaust as it starts.
One and half tanks after the timing fix I decided to run ECU and throttle reset with AlfaDiag in hope of rectifying the fuel consumption and the start-up woes. No fault codes were active.
After 100 miles it started to seem that something is severely wrong as the fuel gauge was free falling. I drove to fuel station to check if it can be showing the truth. After fueling I calculated the consumption and came up with figure of 16,5l per 100km (!)
I hooked up AlfaDiag to car and went on test drive to highway. Car seemed ok, some hesitation in lower revs, but nothing major. 50 to 75 mph accelaration in 3rd gear took about 7 seconds so at least with full throttle the performance seemed ok even according to the clock. What was troubling is that I noticed lambda state stuck to rich for long periods while driving steady speed of 65mph.
Example from log:
0:0:14:34.858,Cold or Open Circuit
0:0:14:36.046,S.C to Gnd/Sensor lower limit
0:0:14:46.749,S.C to Gnd/Sensor lower limit
So, it really seems to be pouring fuel in for no good reason. No fault codes either. Temperature reading from ECU varied between 84 and 90C so nothing wrong there either. The oil had a hint of petrol odour after this episode and was quite black for the mileage it has done
I have a suspicion that the lambda heater circuit on the car side is somehow faulty as it takes up to 5 minutes when idling for lambda to start switching. The resistance of the heater is within spec when measured from the plug (9,2ohms). This behaviour occurs even with warm engine if the exhaust has cooled off. Switching starts much faster if you keep revs at 3k for some time or start driving immediately after start up. After that the lambda readings at idle or 2000rpm seem normal.
Could this have such a dramatic effect to the mpg and how can I diagnose the fault? Car had some universal lambda when I bought it and that heater of that had been grounded directly passing the ECU. This "fix" was removed when car was fitted with Bosch lambda. Maybe it was there for a reason afterall.
Some ideas would be more than welcome - I have voltage/resistance meter + AlfaDiag for diagnosing whatever you guys suggest.