Hard to start, EPC, limp mode, no codes-Possible solution, CPS? [Archive] - Audi Forums

: Hard to start, EPC, limp mode, no codes-Possible solution, CPS?

09-20-2011, 02:26 AM
This is long, but in the hopes that this may help some people here, I am posting it. Iíve got a 2005 A4 1.8TQ ( AMB ) 6MT, mostly dealer maintained, 52k miles, which Iíve owned for two and a half years. New throttle body in Jan 2011, havenít had the timing belt done yet. Lately, the car seemed to have a slower throttle response and seemed to lack power when pulling in midrange RPMs. It also seemed to stumble throughout the RPM range, but without being REALLY noticeable. Then something weird happened-I went to start it after a trip to dinner and it was really hard to start (it did after about 10 seconds of cranking). When it did finally start, the EPC light was on, it idled at about 1200 rpm, sputtered, and ran really rough with no power, but it got me home (~6 miles). When I got there, I turned it off and tried to start it again, it started and ran perfectly. So I thought it was just a fluke until it happened again, 3 days later, under similar conditions (the car had been off for about 1.5 hours). I noted that the temp gauge was halfway between low and all warmed up (2nd line clockwise from the left). Again, it got me home, and when I shut it off and restarted it, it was fine (EPC light off). I have never had any problems starting it when cold or when completely warmed up. The weirdest part, when scanned, no codes (generic code scanner, I donít have access to a VAG-COM). I have a P0441 because of a malfunctioning N80 valve (on order), but I donít think that would cause an EPC error (correct me if Iím wrong), and it existed for 2 weeks before this started happening. I started searching the message boards and the symptoms seem to match a few common problems, but I was particularly struck by the crankshaft position sensor problems people have had, especially that it didnít always throw codes. I tested the resistance between terminals 2 and 3 on this sensor (should be 730-1000 ohms), and it was 939-the other two measurements were infinite, as they should be. Though Iím not one who normally advocates just randomly changing parts, I decided that it would be prudent to change it, given the ease of replacement and relative cost and the seemingly high (anecdotal) failure rate. After swapping the sensor (thanks to the great write-up on A4mods), I started the car up and immediately it seemed to have more ďpepĒ, which was confirmed when I drove it-better throttle response and power throughout the RPM range. I know, not highly scientific-take it with a grain of salt.
Got home and started thinking about why the old sensorís resistance would be within normal limits and then I realized that the problem only happened when the engine had cooled a bit from normal operating temperature (n=2). So naturally, I decided to heat the sensor in water while measuring the resistance. Unfortunately, I didnít have a thermometer, but I can say that the resistance increased slowly and consistently to about 1300 ohms until it got to about a minute or two before boiling-then the sensor went a bit crazy and the resistance started fluctuating between 1600 and infinite (on the 2000 scale-I would later see that it was more than 5 kOhm). Then, as the water boiled, the sensor settled back to about 1300 ohm. Cooled it down, and on the way down, the resistance went to ďinfiniteĒ again, then as it cooled more, the resistance went to 1300 and decreased until at room temp it was 1033-didnít return to the 939 original baseline, but I think the informative part was what happened between RT and boiling.
To be fair, I know nothing of the sensorís normal values throughout the temp range, Iím certain that it wasnít designed to be boiled in water, and I wasnít about to do this with the new sensor. However, given that I seem to have recovered power, and have replicated the conditions (as near as I can tell) for the EPC fault and havenít seen it (itís been about a week now), I thought it was interesting that the sensor would behave this way. I also donít know how often the ECU tests the sensor-if it only tests it on startup, then it makes sense that when the sensor is in that intermediate temp range that the EPC fault would occur, though why even a generic P0322 code was not generated I donít know. Of course, I could be way off base here and the EPC problem will recur soon. If not, I hope that this helps someone out there dealing with similar intermittent problems and a seemingly ďnormalĒ system.