So continuing my SAI slog, I opted to remove the alternator yesterday to gain access,
and created some high tech test equipment after work today.

http://986forum.com/forums/uploads02...1492047137.jpg

http://986forum.com/forums/uploads02...1492047206.jpg

The video result shows...

https://www.youtube.com/watch?v=G4_9qSA-m7A

The changeover valve is operating the cut off valve and air is being pumped into the exhaust manifolds.

Looks like buzzing out the wire harness is next.

James, I admire your resourcefulness and tenacity.
Here is a link to a tool that may help with the harness diagnosis:
https://www.youtube.com/watch?v=Y0HjaR0tj7Q
HF sell one similar but probably not as effective?
How does the ROW Flash sound now ?

I'm doing this.
I just ordered a solenoid valve for $5...if it doesn't work out, I'm not out much. But it will work great. ;)
I'll be happy to eliminate vacuum from the system and lose the electronic changeover valve. Vacuum sucks (pun intended :D).

I agree the whole vacuum circuit seems over complicated and prone to failure. Metering the air into the exhaust manifold to maintain the A/F ratio at the O2 sensor will be the tricky part. It will be interesting to see your concept.

Buzzed out the wiring harness, and without a 2003 wiring diagram that's useless. Found that the ground lead for the change over valve goes back to the appropriate ground leads. The power lead was quite funky. First off, the connectors on the "Cartronics" computer mounted to the firewall seem to be pinned differently for different DME's. I'm working with a DME 7.2 schematic, but the car is DME 7.8, and the blue and red wire for the change over valve shows up in a completely different place on the cars computer connectors. Also, conductivity from the change over valve positive wire shows up in 3 pin locations! Am I looking at incorrect schematics, or do I have shorts between *three*, count them 3 wires in the harness? 2 shorts between 3 wires seems unlikely. I manually cycled the change over and shut off valves over 100 times to reduce the chance the recently replaced shut off valve is sticking due to carbon buildup. Doesn't seem to be sticking. The car goes back together tomorrow and gets driven to see what happens. Whose idea was this whole SAI thing anyway? I need to send them a Thank You card. :D

Maybe some people will understand why in desperation people less qualified and less tenacious than James do the ROW mod and dump the whole system. The carbon footprint of all the diligent work done to trouble-shoot and repair these systems far exceeds the benefit of SAI in these old cars that run low miles these days.imho

I saw a report that said the SAI system was shown to have very little to no benefit in emissions.
I'd do the ROW flash if there were somewhere near me that could...would rather do it myself. :D

The zealots of the "Save the Planet" crowd will complain at both of us. But I agree with you, That is why the Rest Of the World do not use the SAI system. And the Euros have excellent green cred?
The same problem exists in Architecture. Trying to get an Architect/designer to understand lifetime cost of 'green' designs or total end-to-end carbon footprint is very difficult, Because they are pandering to emotions not trying to find the best for the planet.
In theory , you can do this yourself. Just send your DME to ECU Doctors- $500
http://www.ecudoctors.com/porsche-boxster-ecu-performance-tuning-flash-2.html
Porsche 993 Performance ECU Tuning Flash - Specialized ECU Repair
If you Search(I did !) there are hundreds of Posts from smart Porsche diy guys bewildered by unfindable faults in this finicky system. The individual components are simple and the system looks simple on paper but fixing it is nonsense.I say that because if you read these threads to the end , the fix is seldom pinpointed. It could have been Drive Cycle or some other Readiness trick. Meanwhile people keep failing SMOG tests ,waste a lot on unnecessary parts, waste money with Mechanics. It is very easy to cause collateral damage when working on this system.Often you need to remove Bank 1 inlet manifold so it is a lengthy task. Many cases have multiple futile, expensive attempts at a fix. All this for a system that runs for 90 seconds max at start up. There ae better ways to save the Planet imho than throwing time and money at the SAI system.
Oh and those who thought they fixed it - some are back a few months later with the same Codes- again !
Even Jaguar don't have this problem ! Why? They used Delco !

I don't think the SAI is there to do anything for emissions other than during the rich start mode. I think it's main purpose is to test the function of the 02 sensors so they can reduce emissions during the drive. I could be wrong though.

Correct. It is an ineffective 90 second max cumbersome band aid to reduce unburnt HC. Just drive 5 miles less in a year and your account with the Green God will be absolved. She has an FAQ on this very subject !

I read the design documentation. They're saving the planet, ~90 seconds at a time. The system does run at points while driving as a self check.

Actually James ,I think you are saving the Planet and us specifically because you tirelessly perused this problem with inventive,inexpensive test equipment (on video too !) and listed all the tests you did.
You explained all the frustrating results.You explained how the system worked and demonstrated it using a vac pump simulation. I do not recall anyone else doing so much to completely understand the SAI system and use it as a teaching aid.
You bought parts, hunted/fixed vacuum leaks. You even discussed DME and complex wiring/connector faults. All this in one Thread. Very useful to future SAI warriors.
Despite all this effort, no resolution.
Partially the unresolved issue is a lack of a 100 % reliable Drive Cycle protocol for SAI faults. Imagine if the only problem is : You already fixed all SAI faults days ago.All that is missing is an effective Drive Cycle to complete the Readiness Tests.
Ahsai has Posted a simplified Drive Cycle that worked for him. Some of the previous Drive Cycles were impossible to do on public roads. Note the link below- some of the gotchas - like fuel level !
https://rennlist.com/forums/996-forum/967184-smog-geeks-help-needed-obdii-monitors-not-ready.html#post13818633
Doesn't Bentley have the 2003 Wiring Diagrams ?

I moved, and alas my Bentley, as well as my 101 Boxster projects (which is excellent by the way, thanks Mark!) is still lost in the flotsam and jetsam that accompanies every move.

However, I do tend to try and give back to the forum as it is based on give and receive. A long time ago I gave away some parts as I have on occasion (forgot about it actually) and that returned a PDF containing all the information needed to continue the pursuit of this gremlin. So with a new lance in hand, I'm charging this windmill again.

With that in mind, I recently bought a couple of clusters to try and fix mine. A base from a visit to Woody's at an incredibly reasonable price that will stay between him and me, and an 02S that I picked up for $129 on Ebay. Both have been disassembled and re-assembled several times with the battle scars to prove it. The S was dropped and the back plastic cover is now broken.

Both are available for the cost of shipping if anyone wants them. First dibs on either one goes to a certain person who took the time to send me a passel of PDFs :D

James,
With all the thrashing around you've done in the bowels of the engine - is it possible that some SAI component is reinstalled backwards by accident?
I ask because I mark all my components with Paint Pen Marker just to help my feeble memory.
These are the only ones that seem to work. I use them for plasma cutting and welding also.
https://www.amazon.com/Super-Met-Al-1296-1324-Squeeze-Action/dp/B000KL7OYO
The Sharpie version I found to be useless.

This is a great question because the one item that can be easily installed backwards is the vacuum check valve. Every time I remove and install it I question it (the black side goes against the black of the intake manifold). Simply knowing black on black isn't good enough for me, I have to blow air through it and think about where the vacuum is coming from each time to be sure. The rest of the components won't fit in backwards and the system works when wired around the computer control so it appears the mechanical components are installed correctly.

Very gracious of you James not to take offense at suggesting you screwed up :-).
I meant it as a warning to others to mark the orientation but then realized that the reason others may have a fault is that some(which?) component was already backwards.

Buzzing out the harness using the correct wiring schematic for DME 7.8 showed that the harness to be in good shape. To buzz out the harness disconnect all connectors on the ECU.

http://986forum.com/forums/uploads02...1492286054.jpg

Use an extension wire to the multimeter to check the resistance between the wire of interest in the harness and all of the pins going into the ECU and to the car body. This will find a broken wire, or a short between two wires or a short to ground. I use a sewing pin to reach into all of the connector pin holes to check for continuity.

http://986forum.com/forums/uploads02...1492286190.jpg

Everything checked out, so other than the bad relay there was nothing wrong with the SAI found.

The car is back together, and I even painted the rear bumperettes during the down time. I used sandable primer, sanded to get rid of the texture on the bumperettes, 3 or 4 coats of duplicolor general purpose black, 2 or 3 coats of clear. let dry several days then sand to remove remaining imperfections with 2000 grit paper and finish with polishing compound.

http://986forum.com/forums/uploads02...1492286639.jpg

With any luck , this is the end of this string.

Well, my old friends the bank 1 and bank 2 SAI codes popped back up shortly after my last post on this string and I kept grinding away at it and found a second problem in the SAI, this one of my own making.

http://986forum.com/forums/uploads02...1492911578.jpg

That little black and white cylinder in the middle of the picture is the SAI check valve. It provides vacuum to the changeover valve to open and close the shut off valve. Running the SAI pump and manually switching the changeover valve (the video I posted) to open and close the shut off valve taught me what the pump sounds like when the system cycles. On a cold start up I heard the pump running and it was clear the system was not cycling, the shut off valve was closed, period.

So either the changeover valve was not getting an electrical signal, but the wires buzzed out good, or it was not getting vacuum. The check valve is brand new and functions out of the car, but its vacuum source was blocked. Looking closely at the picture a clean area to the left of the rubber manifold tube the check valve fits into can be seen. This is where the rubber manifold coupling was and the hole inside the rubber boot for check valve vacuum was blocked by the plastic manifold tube. Moving the coupling to the right unblocked the vacuum hole.

I fixed that and drove 30 miles and no CEL. Its supposed to be cold tomorrow morning so the SAI will definitely run when I start it up.

On the upside I learned about O2 sensors and how they work as well as long and short term fuel trims (mine are spot on perfect!) as well as some additional information on the MAF sensor. It's been frustrating but I walk away with a much better understanding of the SAI system and the engine control system as a whole.

One other side note, Bosch is used as a subcontractor to design the engine control systems (including SAI) for many different badges. They use the same parts in many of their designs and they also subcontract for parts so you can find the exact same parts in different brands of cars which is good to know if you're not into paying extra for a "Porsche" parts box.

Hopefully that fixed it. If not, you may want one of my SAI delete circuits. :D

Who knows? I may be ordering one in the morning. :)