Hello everyone, I've been having an ongoing problem with my car and in an effort to eliminate things dissembled and cleaned the two secondary air valves.

WARNING: I do not recommend this as a permanent fix as there is a possibility of bolts or screws coming lose and falling into the engine and destroying it.This post / article is educational more so than a repair.

I did this for two reasons firstly it was a test to see if a valve was malfunctioning and it happened on the day I was driving to get the yearly vehicle inspection done - how convenient, after driving home again and clearing the code thankfully the check engine light did not illuminate when the vehicle was checked.

Here is the error that keeps reoccurring intermittently:
What it actually means is low air flow on the left side into the exhaust as the computer does not see the O2 sensor react (or react enough) to the extra oxygen being added to the exhaust flow.
I had already removed the air pump cleand it, checked brushes and bearings and it checks out fine.

I could not find good photos of this valve anywhere on the net - maby I was using the wrong search terms?? So here we go.
Sorry but I started taking photos halfway through the process as I did not think it would be so hard.

Initial disassemble was easy only three screws. They are very tight - you may need to use and impact driver.







To me it looks like lots of carbon buildup on the other side of the one way flap valve as is not closing properly. Maby I'm wrong and this is normal??




This does not look normal to me as there is carbon under the valve stopping it from closing fully







Here is the hard part. The screw that hold the one way valve together are stuck, and i mean really stuck. You can see on the first one on the left I broke both off both not realizing that the screws are spread out on the other side (look at next photo). The other thing I did not know was the plate with the one way valve separates from the rest of the body. Even with using a Dremel and taking off the back part of the screw (next photo also) the screw has been turned into a wedge inside the aluminum and still refuses to come out. I could only get half a turn an after that the head broke off anyway. I ended up drilling out what remained of the screws - please excuse the poor effort with drilling them out I was getting frustrated and impatient at that point (out of character for me).

Side note: Some people will say why not use a solvent to clean the carbon off - I did try but exhaust deposits do not dissolve readily so you need to abrade it off. The other problem is one part has a rubber coating and may be damaged by soaking. Disassemble and gentle mechnical cleaning is the only way (I think) may someone has a better idea ie ultrasonic cleaner??







All clean and prepare for reassembly:

I used a combination of brake clean, carby clean and Cif (or Jif) cream cleaser with a non scratch scourer pad. I found that rubbing with soap or cream cleanser worked the best being careful to hold the vacuum actuator upright to stop water entering the diaphragm via actuator shaft. I also held the actuator valve open with vacuum to clean the sealing surfaces. The rubberized side of the plate i cleaned mainly with my fingernails and rubbing with Cif. All was blown dry with compressed air.









Here is the completed assembly. I had to enlarge the holes and used machine screws with high strength high temp locktite. (yes it's on the thread inside and not just on top). The other problem I had was as the screws were bigger than the originals the nut interfeared with the main valve housing so I had to modify both the nut and housing to make it sit right.









After all this the error still happens so I'm still in the process of diagnosing whats going on. It's hard to replicate on demand as it only happens on cold start now and then, after that you usually need to wait until the next morning.
The thing that's really giving me the sh**s is when I watch O2 on Star Diag during startup it does not read anything in actual values (just shows a ratio of 1) until after the air injection function has completed and by that time the O2's seem to read normal.
I cant use my Tektronix scope as wide band O2 sensor signals cannot be look at directly like old style sensor and require the control circuitry inside the engine ECU to get a sensible reading.

So at the moment I'm leaning toward a possible bad (ie slow and not dead) O2 sensor or blocked air injection ports in the head.
Next I will try to do a flow and back pressure test and see if thats the problem.

More info to follow.

If your going to reply with answers like that please dont.
Yes, I've been through all the tests except the current draw as I don't have the breakout box. Yes and I do know how to use scan tool. I've been doing automotive electronic board level repair for 20 years. Yes I've done all manner of mechanical repairs also Australian American European Japanese.