Ok, so here's a crappy dataset of the O2 sensors on a cold startup. I forgot my stopwatch, so I just kept counting to 5 and taking a picture (5s sampling time) of my scanner, and then plotted the numbers manually. I just bought a datalogger which should be here tomorrow, so Tuesday morning I should be able to post better data.

My reader shows "O2Sxy", and I'm assuming X represents the sensor number since it's after "S", and Y represents the bank. Yes?

O2S21 goes to zero pretty soon after startup. It's 'mate' doesn't. Hmmm. As soon as I heard the air pump turn off, O2S21 started to climb.

Thoughts Ahsai?


I also took out the alternator to get a better view of the solenoid and pneumatic valves.

Pneumatic valve tested out OK: I tested it by blowing on its inlet tube that the air pump connects to. Couldn't push any air thru. I then drew vacuum on the valve's vacuum port, and now I could blow air thru it.

Electric solenoid valve tested out OK: I tested it by sucking the port connected to vac, and it was blocked. I applied 12V to the valve and I could now suck air thru it.

So it appears that those valves are working correctly, which is a good thing.


But that data. Hmmm. Still no CEL and all the other monitors (Cat, O2, etc.) pass and say OK.

 

Excellent plot. For the O2Sxy coding, please re-read post #13. I've already explained there (x is the bank # and y is the sensor position: 1=pre and 2=post).

I think you've found your problem. The SAI doesn't set because your orange and blue curves are not staying low enough during the first minute of cold start.

Are you sure your legend of the curves are correct? It looks more like to me yellow and grey should be from one bank and orange and blue from the other. That would have made a lot more sense. The yellow is the pre-cat and grey is the post-cat, from the same bank and their readings are spot on.

Orange and blue both don't "see" enough fresh air on the same bank, possibly due to carbon build up at the exhaust. You would need to remove the header and check the SAI ports to correct this problem.

But first thing first, if you can confirm once again the legend of the curves, we can do a more accurate analysis.

 

Thanks Ahsai. You're helping me narrow in on the problem.

For the "XY", I understand there is a pair of pre-cat O2 sensors, one on bank 1, one on bank 2, and a post-cat pair, one on bank 1 and one on bank 2. What I'm not sure about is how my scanner is referencing them with their nomenclature. Am I assuming correctly that O2S12 means sensor 1 on bank 2?

I just double checked all the pictures of my scanner that I took the data from, and all of the O2 sensors are correctly labeled in the graph. Strange...

I've never taken the headers off, but I have removed just about everything off the back end of the car to get at the coil packs a couple months ago. Is it a simple process? And I just poke-out any carbon buildup that I see in the SAI ports? Where are the SAI ports? Are there holes in every exhaust port in the head, or just one on each side?

In the last two months I've changed the plugs, coil packs, AOS, water pump, thermostat, starter, and window regulator. Man, now more work. : / And I thought Alfa's were bad!

 

Ok, seems like I'm not getting my point about the "XY" across despite trying it 2 times

I'll just list it out for you this time. Note these are per the OBDII standard and that's also what your scanner SHOULD be using.

O2S11: bank 1 pre-cat
O2S12: bank 1 post-cat
O2S21: bank 2 pre-cat
O2S22: bank 2 post-cat

Suppose your scanner is good and the graph labels are correct. Here's what I don't think it makes sense. The red will be bank 1 pre-cat and the gray will be bank 1 post-cat. Why the post-cat sensor was able to detect the O2 after 40s yet the red pre-cat still can't detect the fresh O2? One possibility is you have a worn out bank 1 pre-cat sensor.

Also, the orange line will be bank 2 pre-cat which detected the O2 perfectly but why the blue bank 2 post-cat didn't? One possibility is you have a worn out bank 2 post-cat sensor.

I would like to see another set of data from a different scanner and if they show the same, I think changing out the O2 sensors first (~$100 each Bosch direct replacement) may not be a bad idea.

However, if it turns out what I said was right in the previous post, you will be looking for SAI ports. Once the header is removed (just 6 to 8 header bolts, best soaked with Kroil and remove with impact gun), the port or holes are right there for you to see and carbon buildup will be obvious (if that's the problem). There was a thread I helped another gentleman and it has very nice pictures of the ports once the header is removed. I'll see if I could find that thread.

 

Check out this thread for photos of the header and SAI ports
http://www.renntech.org/forums/topic...comment-277522

 

Thanks for listing the names out, I couldn't find that anywhere in the previous links you gave, only talk about the different banks/sensors. That's exactly backwards to what I was thinking, so I'm glad I see it correctly now.

I don't know what to think about the O2 sensor data. I don't want to go spend $200-$400 just to find out it was carbon build up, so I'll wait for my datalogger to show up tomorrow, take some data on Tuesday morning, see what it says, then likely take the exhaust off Tuesday night regardless since I know there's got to be some issues there.

Thanks for the link showing the ports. I also found a boxter post where the guy flushed the whole SAI system with Sea Foam and a little hand pump. I'll likely do the same. I just wish it wasn't a rush job cause of this smog thing; why couldn't Porsche throw up a CEL if the monitor isn't readying for nearly 400 miles?? Then at least I'd've known. Oh well, time to dig into the car yet again.

 

Yeah, I agree we should collect more data first. If you suspect the sensors, you can switch them between the banks and see if the readings follow the sensor or the bank. That will give us even more data points.

In any case, seems it's a good idea to pay for your registration first to avoid penalty as I think it may take some time to diagnose and fix your problem.

 

Re the CEL, it's possible your system is at the edge of failure and it intermittently passes some of the tests (within the SAI tests) . When some of the tests fail, the DME is giving it another chance before it flags the CEL. There are quite complicated algorithms inside the DME to make it not too trigger happy and yet be able to detect persistent failures.

 

Ahsai you are the man, I don't have to deal with the smog requirements in MI, but if I did I'm glad you are on this site. Thank you for sharing all your knowledge and expertise in all things 996.

 

Thanks! I'm always happy to spread the joy of diy and diagnostic

 

Soooooooo. Yesterday I got my datalogger and all the stuff to clean out the carbon once I drop the headers. I scanned the OBDII when I got home (like I've been doing every time I stop!) and guess what? Secondary Air System is now ready! Even with that not-great-looking O2 data, the system is finally green across the board (SAI was the only one not ready). And it only took somewhere north of 300 miles (it wasn't up to 400 yet like I'd thought). Sheesh. Never a CEL, and the car has been running better than it ever has after all the work I did. Anyway, just got my smog certificate and Mr. California DMV is happy.

I will still take some data on my O2 sensors in the next few days and plot it out, as the system is obviously not 100% healthy. But now I've got two years to find the time to pull the exhaust and clean out all that carbon and/or replace any O2 sensors.

 

Heh. If it's any consolation, Mr. Murphy's Law stipulates that if you had not ordered the datalogger, the problem would not have cleared on its own. You were screwed either way.

 

Sorry if this does not help... I did not read all the posts...
If your looking to reset the Readiness Monitors this is the work sheet that I came up with from searching the net last year:

Start engine, idle cold for approx. 2 min, 10 secs.

Accelerate to 20-30 MPH, Maintain steady speed for approx 3 min, 15 secs.

Accelerate to 40-60 MPH, Maintain steady speed for approx 15 mins.

Decelerate and come to a stop. Idle in gear* for approx 5 mins.

1..As soon as the engine starts, idle the engine in drive for two and a half minutes with the A/C and rear defrost on. OBDII checks oxygen sensor heater circuits, air pump andEVAP purge.

2. Turn the A/C and rear defrost off, and accelerate to 55 mph at half throttle. OBDII checks for ignition misfire, fuel trim and canister purge.

3. Hold at a steady state speed of 55 mph for three minutes.
OBDII monitors EGR, air pump, O2 sensors and canister purge.

4. Decelerate (coast down) to 20 mph without braking or depressing the clutch. OBDII checks EGR and purge functions.

5. Accelerate back to 55 to 60 mph at half throttle. OBDII checks misfire, fuel trim and purge again.

6. Hold at a steady speed of 55 to 60 mph for five minutes.
OBDII monitors catalytic converter efficiency, misfire, EGR, fuel trim, oxygen sensors and purge functions.

7. Decelerate (coast down) to a stop without braking. OBDII makes a final check of EGR and canister purge.


just hold the following RPM ranges from 2-5 minutes:

Idle
1500
2000
2500
3000
Idle

the checks usually happen before the 5 minutes is up, but without a scan tool hooked up you won't know when it happens.

 

Heh. Now you just have to find a place in SoCal where you can go 55 for three minutes without rear-ending someone or getting shot.

 

That's nice. It's quite possible since the system is at the edge of failing so sometimes the SAI tests pass and sometimes they don't. Good you caught it when the tests passed.

Let us know if and when you have more data to look at.