John Crisalli

Thanks to all who posted about the oil consumption issue. The consensus is that 1900 mi/qt is not considered excessive. So if excess oil isn't there, has anyone done any research into intermittent SAI CEL's. Again, the CEL is set during the first 2 minutes of driving when the SAI pump is operating. If I let the engine idle for two minutes before driving, a SAI code is never set. As I stated earlier, I know the SAI ports are clear. Any insight as to what's causing the SAI CEL lights under those circumstances?

As usual, the wisdom of the group is always appreciated

Thx

JC
98 C2S widebody

pp000830

• The most common is a partially blocked check valve, a normal maintenance item to replace. I had this and even with what looked like air flow through it, it was impeding air flow enough to trigger a CEL.
• With Durametric software, the air pump can be turned on and off to check that it is running.
• There is also a third air valve I seem to recall that is vacuum actuated in-line between the pump and the check valve where the vacuum hose to it needs to be checked as in place and the valve and its solenoid control valve can be tested.
• If you leave the hose off the check valve and let the car cool overnight and cold start the car in the morning while you have someone observe/put their hand over the hose they should see a lot of air flow and it is probably quite loud for a very short period before the air pump shuts down as a normal course of its function. This would function as a shade tree mechanic's integrated system test for the blower and the solenoid control valve.

John Crisalli

Thanks for the reply. Let me reset to the scenario. The SAI CEL is set on start-up. It's not on every start-up. Sometimes the CEL is for the left bank of cylinders and sometimes it's only for the right, and sometimes it's set for both. If I let the engine idle for the 2 minutes that the SAI pump operates, a CEL is never set.

You don't need a Durametric to check the function of the SAI pump. You can remove the relay for the pump in the fuse box in the left side of the engine compartment. You jump terminals 30 and 87 with a jumper wire that has an inline 30 amp fuse. This will trigger the pump at any time, hot or cold. I replaced the check valve approximately 10K miles ago. I don't think it is clogged or inop. If it was, I would get a CEL upon every start-up. I have tested the vacuum valve and solenoid. Both appear to be functioning. I did replace all four O2 sensors and the issue disappeared for awhile, but it has returned.

When I trigger the pump I can watch the function of the O2 sensors with OB-II software. The O2 sensors voltage drops showing an increase of oxygen in the exhaust. The ECU will correct this in a short period of time by increasing the fuel supply. When I stop the pump, the O2 voltage levels increase and the ECU cuts the fuel to bring things back to specs. This tells me that the ports are open although there could be one or more clogged or partially clogged. There is an option that one or more of the O2 sensors is intermittently lazy on start-up which would be a cause of the intermittent problem. Unfortunately, my OBD-II software has a hard time connecting to the car and will only connect with the engine running which makes it difficult to test for a lazy O2 sensor.

So, the original question is still there. Assuming that the system if functioning correctly, what other issues could be causing an SAI CEL to be set? Has anyone ever had this issue?

Any info shared is always greatly apprecitated.

JC
98 C2S Widebody

pp000830

Since your CEL sometimes points to bank 1 and sometimes to bank 2 suggests it is not your oxygen sensors.
If your Oxygen sensor heaters were not working the fault code would be specific to this.
It seems to me it may be a flow issue where the amount of air injection is sometimes not quite enough and other times seems to be adequate. Sometimes tripping bank 1 sometimes tripping bank 2.
So the issue is how much air is the pump pushing out and is a faulty vacuum controlled valve or check vale incrementally impeding flow just enough to cause an intermittent issue. I still think I would consider replacing the check valve again as it may have an internal manufacturing issue or some debris got into it. Next, I would pull the solenoid controlled valve and inspect it to make sure it fully opens internally when actuated. Finally, I would seek guidance on how to test the air pump output as to if it is adequate.
The entire system is just not that complex and if the pump and supply devices check out it may simply be partially clogged ports that need the old bicycle brake cable rotor-rooter job.
Good luck with this,
Andy

techman1

Maybe my memory is playing tricks, but does anyone else recall a few reports of the wires around the pump and valve breaking down?

pp000830

You may have a point, however,
the insulation fails along a specific segment of the wire due to heat cycling. I would think if it caused a grounding condition or a short the symptom would be a blown fuse.

John Crisalli

Thanks for the reply. I like the idea of the bicycle brake cable as a roto-rooter tool. I have thought about ways of cleaning out those ports but never came up with an idea on how to do it. I will try this trick. I think if there is a clog, it'll be closer to the exhaust port rather than closer to the check valve. I might spray some ATF in the ports the night before to help break up the deposits.

I have the factory service manual and the OBD-II troubleshooting manual. Neither give a test for or an output volume or pressure spec for the SAI pump. If these were important they would be included in the manuals. So if the pump is working, I have to assume the the airflow is sufficient.

I would be very surprised if the check valve were inop. If it were, the problem wouldn't be intermittent. The CEL's would be set everytime on start-up and not just under cold engine acceleration.

I plan to plan to check the inspect the check valve, the vacuum valve, called the change-over valve by Porsche, and roto-rooter the ports sometime soon when we get some warmer winter days. I don't like working in the cold anymore.

JC
98 C2S Widebody

pp000830

See attached:

SwayBar

Have you hooked up an OBD2 tester to see if it's throwing codes?

I discovered my driver-side after-cat O2 sensor was bad, which when replaced, cleared up all my problems, and no CEL for 5 years now.

NoRush993/951

pp000830

Yes, and illegal, easily identified during an inspection and a disclosure issue if one ever sells the car.

PRSWILL

I went through this last year- and discovered there's a VERY SMALL little vacume line right by the SAI valve and I knocked it loose when I did the blow out. As soon as I took the air box off I could feel it wasn't connected. Once I connected it, no more check engine light. My CEL was coming on very intermittently and this fixed it. Worth checking.