hoggel

I have seen a lot of discussion about the infamous Secondary Air Injection system with some good information and some misinformation. I thought I would weigh in and see if I could stir the pot a bit.

A catalytic converter uses a catalyst to convert nitrogen oxides, hydrocarbons, and carbon monoxide into nitrogen, oxygen, carbon dioxide, and water. For the catalyst to be effective it must be at a relatively high temperature. If the catalyst is not hot enough it doesn’t operate efficiently and the catalytic converter can get clogged with hydrocarbons. One of the ways used to get the catalytic converter up to operating temperature is the SAI system.

On cold start the secondary air pump is turned on by the ECU. The ECU also sends a signal to open a vacuum control valve that operates a vacuum valve allowing the secondary air into the plumbing. The secondary air is injected into the exhaust before the pre cat O2 sensor.

The ECU monitors the pre cat O2 sensor. It is a closed loop control system that adjusts the air fuel mixture to make the engine operate as close to the stoichiometric point as possible. On cold start the fresh air from the secondary air pump creates a very lean condition at the pre cat O2 sensor. The ECU compensates by injecting more fuel resulting in a fuel rich condition in the exhaust from the head. That unburned fuel in the exhaust combines with the oxygen from the SAI system and burns thereby heating the catalyst up to operating temperature.

So what happens if there is no SAI system? The catalyst takes longer to come to operating temperature, harmful pollutants are released into the atmosphere, and the catalytic converter can clog over time.

In an OBDII car the ECU checks for the presence of the pump. On cold start the ECU checks that the O2 sensor signal goes through the stoichiometric, lean, rich, stoichiometric cycle with the SAI.

It is difficult to kluge the system so that it will pass the SAI test because the O2 sensor signal is a variable duty cycle square wave. In order to fool the system a square wave signal must be generated that is timed with what would have been the SAI cycle. But the pre cat O2 signal is also used for normal operation of the car so the kluge system must be smart enough to fool the SAI check without impacting the normal feedback loop.

Some tuners of other marques have modified ECU software to delete the SAI check but that is a Federal offense. If you install a kluge in your car to bypass the system and present it for inspection it could be a Federal offense.

It is my opinion that 993 implementation of the SAI system is definitely a poor design. But, we should make every REASONABLE effort to keep the system on our cars operating properly.

Discuss?

RallyJon

Your core contradiction is right there.

Enthusiasts will always try to improve on poor designs, or eliminate or replace them when they're inherently flawed. From engine destroying thermal reactors to wallet destroying SAI systems, Porsche has made some pretty boneheaded decisions over the years.

An environmentally conscious Porsche owner can keep the car 98% clean and improve performance by keeping a catalyzed exhaust. Fixing Porsche's poor solution for the extra 2% shouldn't cause even the greenest owner any lost sleep.

TheOtherEric

Jon hit the nail on the head.

Stealth 993

It's ONLY used to fire the cat off earlier on a cold start. Once the cat is hot, it will not run again.

With out the SAI, it will just take a little longer for the cat to heat up to operating temp. This is also to help combat the rich cold start condition.

In reality, the SAI is what a .002% emissions increase if not working? No facts, just pulled it out of my, you know.

Lorenfb

Check here for more info:

www.systemsc.com/pictures.htm

hoggel

I'm not sure about the 98%/2%. I think the SAI is part of an effective catalyzed exhaust.

But I'm in total agreement with the rest of your post. Especially this part, "Enthusiasts will always try to improve on poor designs, or eliminate or replace them when they're inherently flawed." I'm no wild environmentalist. I won't bomb anyone's house or even argue with anyone. I just don't think an electrical kluge of the SAI system is the right way to solve the problem, ethically, practically, and especially legally.

I'm not a guy who just talks (posts) either. If my car required a head removal only to clear the SAI passages I don't think I would pay the price. There should be a mechanical bypass solution. I think that new plumbing routed from the fitting where the secondary air enters the head closely around the head (get the air hot) to the exhaust should serve the same purpose. A thin plate could possibly be put between the head and the heat exchanger to insert the SA? I envision something like the old Holley carbureter double pumper fuel tubes distributing air into an adapter plate/s.

I'll be in the garage this weekend taking some measurements to see if it looks possible. I ordered the Durametric tool so I can drive the SAI system for testing. If anyone has already tried this I would be very interested in your success or failures. I'm not looking for a new product to sell. Ideally we could come up with a solution and publish a parts list and drawings.

?

AOW162435

I think David (dhicks) tried this very 'solution' and found it to work less than optimal. A search should unearth his thread on this.


Andreas

Lorenfb

"I ordered the Durametric tool so I can drive the SAI system for testing."

What? Unless I've overlooked something on his website, that capability is lacking:

Module Read Codes Reset Codes Actual Values ECU Information
DME Motronic 2.10 (Engine) Yes Yes Yes Yes
DME Motronic 5.2 (Engine) Yes Yes Yes Yes
Tiptronic Yes Yes No Yes
ABS 5 Yes Yes No Yes
Heating & A/C Yes Yes No Yes
Airbag Yes Yes No Yes

There's no mention in the above. Also, I personally asked Duram to incoporate
the PST2 "Short Tests" (SAI & Tank Venting) and he said he wasn't going to.
Also, one needs to monitor the cycle flags during the setting of the readiness
states prior to an emission test which is not in the Durametric system.

Bottom line: Without "forcing" a SAI mode, any testing and evaluations are not possible.

hoggel

Thanks for the tip toward dhicks Andreas. I'll look into that in a bit.

I check out your info often Loren. Thanks for the help. I'm intimately familiar with the vw emissions systems after swapping a 2004 1.8t, including all of the emission controls, into a 1990 mkII GTI. Created what may be the only ULEV mkII in the world? I use the vag-com tool for VAG cars. But I am just learning the intricacies of the Porsche implementation.

Maybe I should have called Durametric. I assumed from the page 10 of their instruction manual that the tool would initiate the SAI test? I need the tool to reset my Fuel Relay code right now. I'm looking forward to using it to see what it can do.

dhicks

I tried the plumbing route but was less than succesful due to the fact my bypass pipe was too big and could not deliver sufficient pressure.....somebody else tried it with small ID aluminum tubing and it worked a treat.....I will try and find a link

dhicks

Here you go
https://rennlist.com/forums/showthre...&highlight=SAI

Lorenfb

"I assumed from the page 10 of their instruction manual that the tool would initiate the SAI test?"

The manual may mean that you can turn the SAI pump on/off, but still not
be able to "force" the system to run a complete SAI test which utilizes the
O2 sensors to check the SAI effectiveness. The VW/Audi app from Vagcom
is much more powerful than the Durametric or other Porsche diagnostic
apps.

hoggel

That was an interesting read. I do remember clicking into that thread but I don't think I looked at the first couple pages. When I saw discussion about the electronic kluge I didn't follow the thread.

So the apparent success of jrb964's setup kind of shakes a couple of my assumptions.

1. I thought the SA injection site would have to be near the head to allow sufficient fresh air/exhaust mixing prior to the O2 sensor. It doesn't seem like the O2 sensor would get a good read of the mixed air with the injection site straight across the pipe. It sure is a convenient fitting though. And no CEL on jrb964's car?

2. I assumed that the SA would have to be hot to promote burning of the rich exhaust for cat heating. I made this assumption because the factory repair manual says the cylinder head temp should be >176 F before testing the SAI (volume VIII page 24-51). And for what other reason would Porsche cast the air flow through the head? Maybe I give the Porsche SAI designer too much credit?

In the previous thread it was discussed that this should be hidden from the emissions inspectors. I know here in Utah unauthorized mods to the emission systems are illegal. But common sense would say that if it can be demonstrated that the existing design is ineffective that the modification restores the proper operation of the system couldn't it receive approval or a CARB EO?

hoggel

Is there discussion of the emissions system in the factory repair manuals? I have the entire set but it seems that section may be missing. I looked at the pdf of the manuals and it seems the same sections are missing. Is the Porsche numbering system screwy or am I missing something?

Specifically, is the EVAP and SAI short test something different than the checks in the diagnosis manual, test points 20 and 29?

Lorenfb

"Is there discussion of the emissions system in the factory repair manuals?"

One of the best is the Porsche manual "On-Board Diagnostic Troubleshooting (OBDII)",
911 Turbo (993) - WKD 483 320

It discusses all aspects of OBDII as it relates to the 993, e.g. EVAP & SAI,
readiness codes, etc.