Pretty sure its the AOS

I got a high idle cel P0507 about three weeks ago. I cleared it and it stayed away for two weeks. Now it comes back after a short trip. I cleaned the MAF and throttle body. No change. Did a search and came up with AOS. This is what it is doing, cold or with the a/c on the car idles fine. Hot with the a/c off it is idling at 1000-1100. I checked the oil cap. Seems to fit tight. I pull off the cap and the idle pops to 1500 after about 5-10 seconds and stays there. I took the cap off with the engine not quite up to temp, the idle jumped, but returned to normal after about another 5-10 seconds. These symtoms sounded like others AOS problems. Forgot to mention that the car is a 2002. So AOS or not. Is there any lines/hoses the lead from the AOS I should check for leaks first? If it is the AOS I'm debating about getting it from the bottom or top. I do have a very slight drip from the engine/trans - IMS/RMS. Could fix both. Then again the clutch is original (79k). LN upgrade? This could turn into a case of the "might as wells". Terry

check the rubber bellows that joins the aos and the crank vent.... sometimes it will crack. Any oil weeping in that area?

Redridge I was just reading your write up. I didn't get under the car to visually check the AOS area. Another thing I forgot to add is that I have no smoke.

the white smoke and the high pitch squeal is not always present.... follow all the vacuum lines and make sure there are no leaks. That includes the check valve that feeds to the varioram solenoid valve.... on top of the alternator.

I had no smoke on my AOS failure... still failed

Quotes source post: 7707992

Not symptoms normally arising from AOS failure. I don't recall anyone posting a P0507 error code appearing. In fact, one AOS failure out of 2 the CEL never came on at all. And first failure no smoke until very late in the AOS failure. The idle behavior when you remove the oil filler tube cap is normal. Let's see... P0507 is idle air control at stop, above limit. That's an air intake leak -- most likely -- but it doesn't have to be the AOS per se. There are several hose connections to the intake system that can if anyone develops a cracked hose or fitting (due to material aging, brittleness) could account for this error. Another possibly source of an air leak is the oil filler tube. About all you can do is visually inspect each hose and its fitting at the intake manifold and to where the hose goes for any bad hose or cracked fitting. Used to help isolate intake leaks by spraying some cleaner -- carb cleaner but you can use electronics cleaner -- that comes in an aerosol can at the various fittings and connections and hoses that connect the hoses and lines that connect the intake system to the other hardware. If you spray the cleaner at a leak the engine should react and from this you know you've hit a leak with the spray. You might let the E-gas system perform an adaptation. Do not touch gas pedal at all during this. With engine off swtich on ignition to engine on but not to engine crank/start. Wait one minute. Switch ignition off for at least 10 seconds. Read out fault memory. Sincerely, Macster.

I also have a P0332 Knock sensor, but I could not find anthing on that. I'm assuming that this vacuum leak can be quite small and I found one. It looks like the oil fill tube contacted the idler pulley and wore a small hole in one of the ribs. I'm trying some 5 min epoxy. I was trying to apply with the engine running and it sucked it in, so I know it was drawing air. I got my fingers crossed. I'll take it for a drive later.

Anyone change the AOS on a 2002 996? I tried to remove with the transmision out but could not reach the second bolt.

Quotes source post: 7709942

Your transmission is out?? Should be an easy access from underneath I would imagine. However, with the engine and trans in..........All you would need to do is lower the engine in the rear, and you should have enough room to access it.

Quotes source post: 7708794

My info is knock sensor 2 below limit. This diagnosis made with engine speed more than 3600 rpms and engine load greater than 45%. Possible fault causes: break in wiring or short to ground; corrosion at sensor connector; sensor loose; bad sensor. You can check wiring and eliminate the corrosion possibility by disconnecting the sensor and reconnecting it. This renews the connection electrical contacts. Then check the sensor is secured properly. I believe you can also swap one sensor for the other and see if the error code follows the sensor or stays put. This tells you the sensor is either at fault or it is something separate from the sensor. It is not clear from my info how protected the engine is with this error. I'd be a little leery driving the car with this error cause the engine's without its knock protection -- at least on one bank -- and these engines operate at incipient detonation due to the their high compression so there's not much margin if a knock sensor is not working. Sincerely, Macster.

Ok. Sealing small hole in oil tube did not solve the problem. CEL cleared then came back. Same codes as before along with P0154. O2 bank 2 sensor 1. I discontected and reconected the O2's. Macster, it seems the knock sensors are under in the intake runners. Must they be removed to get to the sensors? Are all these codes leading up to a single probable cause? The are no codes till the idle P0507 appears.

Quotes source post: 7722614

Let me see if I can be of some help.... P0157 is oxygen sensor behind converter (cyls. 4-6) short[ed] to ground. Oh, I'm using my OBD II Porsche models 1996 to 2004. Fault conditions: Short circuit of signal wire to ECM ground or housing ground. Note: If this fault detected the fault codes P1115 and P1119 may be stored too; however these fault codes are not real or actual faults. Cause: wiring harness. Another fault condition is short circuit of sensor grond to ECM ground to housing ground. Note: When this fault occurs all 4 oxygen sensors are store [sic]. [ I think this is poor wording that means if one fault detected all other sensors get their corresponding fault codes set as well.] Yes, another note: When P0157 is stored and faults of the heating of the other 3 sensors are displayed, only one oxygen sensor is probably faulty. Diagnosis conditions: P0157 is stored in the ECM when the sensor voltage is below 150mV. Sensor heating active. Test 1: disconnect all 4 sensors. Turn ignition switch to on. Connect digital voltmeter to sleeve pin 3 and 4 of sensor connector 2 behind converter. Read approx. 450mV. Connect a digital voltmeter to sleeve pin 3 and ground. Read approx. 720mV. If you don't get the above measurements the next step recommended is to remove the ECM connector and check wires for short circuit to ground. If no short to ground can be detected the ECM may need replacing. Test 2: Remove oxygen sensor connector 2 behind converter. Connect digital ohmmeter on pin side to pin 3 and the oxygen sensor housing. Read infinity. Connect a digital ohmmeter on pin side to pin 4 and the oxygen sensor housing. Read infinity. Connect a digital ohmmeter on pin side 2 and 4. Read infinity. Connect a digital ohmmeter on pin side 2 and 3. Read infinity. There's the sensor stuff. What I usually recommend doing is taking care of one thing at a time and then seeing what else remains. As for your questions about the knock sensors, I don't know if the intake manifold runners have to be removed to get to the knock sensors. I would hope not. I was taught to leave a perfectly functioning intake manifold the h*ll alone unless absolutely necessary. My OBD II reference doesn't have anything on P0332 but is has P1384 and P1385: Knock sensor (1 or 2) signal implausible. Let's see:... Knock sensor loose. Break in wiring harness. Short in wiring harness. Contact corrosion on the plug connector. Knock sensor faulty. Regardless which one of the above it is, you're going to have to have access to the knock sensor. But I'm still not saying to remove any intake plumbing. Not yet anyhow. Address the P0157 error first though. While you're "there" check engine wiring over real good for any signs of damage. Then see what if any errors come back. You might have to fix this in stages, or steps. Sincerely, Macster.

Sorry Macster meant to say that the codes don'y show up until the high idle P0507 shows. I assume that the o2 test procedure is the same for o2 sensors 1 and 2. I'm going to test them this weekend.

Quotes source post: 7723366

AFAIK and this based on my research into the sensors used on my 02 Boxster, the #1 and #2 sensors are the same. In my Boxster's case the #2 sensors have longer wiring leads but otherwise are interchangeable. Treat sensors with care. Do not get anything on the tip. Do not attempt to measure resistance through the tip. The sensor's one delicate gizmo believe it or not and while they'll last and last mis-handled they can be ruined in no time. Sincerely, Macster.

Update - Problem solved. The hose clamp that holds the coupler on the rear intake crossover tube was half on the rubber half on the pipe. This caused a vacuum leak.

cool! glad it was an easy fix

Quotes source post: 7780135

Easy to remedy...but hard to find. Did you chase down the fix? Nice to know that you didn't have to replace your AOS.

I was getting a high idle and knock sensor code. Idle was ~1100rpm with the a/c off. A few weeks ago it was really hot and I drove with the a/c all the time. Then I got a lean 02 and knock sensor codes, but no idle code. This lead me to think that there must be a vaccum leak letting unmetered air into the motor. Then it was just a matter of checking every possible place for air to leak.