ilko

No hunting or uneven idling and that's after about 60 days of not starting the car at all. It fired right up with no issues whatsoever.

I also purposely did not crank with the DME relay out to raise oil pressure like I usually do after a long time of hibernation. Wanted to make sure that it doesn't upset the start up procedure in any way.

My understanding is that the SAI pump only runs for the first 30 seconds after a cold start. It is interesting that the code appears when the pump is apparently not running. Or does the pump switch back on later for self diagnosis or any other reason? Does anybody have any insight on that?

Lorenfb

"It is interesting that the code appears when the pump is apparently not running. Or does the pump switch back on later for self diagnosis or any other reason? Does anybody have any insight on that?"

On a side note: A CEL does not necessarily occur when a fault code is present
in the DME ECM, i.e. the CEL comes on usually at a later time. This has been
observed when driving and setting up the 993 monitors (readiness codes) and
none of subsequent cycle flags were being set. A check of the fault codes
indicated a 410 code but there was no CEL. Once a fault occurs, the ODBII
diagnostics terminate testing on the 993 system. So, a fault code may be present
without a CEL.

ilko

Thanks Loren. Can a generic OBDII reader see the soft fault code? Or VAG-COM is OBDII mode? I haven't really played with VAG-COM on the Porsche so I don't know what it's capable of.

Lorenfb

"Can a generic OBDII reader see the soft fault code?"

Yes, any OBDII scanner should read the fault code even if the CEL
hasn't turned 'on' yet.

Jeff U

I have had a similar situation. Each time I have gotten a CEL it has been after the car has been running 15 - 20 minutes. (1996 C4S)

I also puzzle on why it should occur here and never upon start up?

I have noticed when the light has come on. It's when decelerating, just as I lift off the throttle, or push in the clutch. The RPMs drop toward idle and the CEL comes on.

Error code PO410

william_b_noble

depending on the generic reader, you can read all the OBDII mandatory codes (they all do that, even the $14 harbor freight unit), or those codes plus extra stuff. My unit reads all sorts of stuff off of my suburban, but on the porshce it is limited to the mandatory codes and the dynamic data such as O2 voltages and stuff. being able to see the dynamic data is really helpful to getting the cycle flags to set after changing the battery or clearing a fault.

So, nobody yet has posted a successful result with this simple circuit on a 97 or 98 model year car, only the 96. I would go modify my circuit to try another approach if someone can tell me what the 4 sensors read during the SAI motor run time so I can adjust the circuit accordingly.

4X4SCHE

I don't have the problem with my 993, but I have reverse engineered an OBD II ECU used on a different vehicle.

I believe that reason the circuit doesn't work for later cars, (and why the CEL comes on after the engine has been driven for some time) is due to a test of the SAI valve that is done with the engine warm. To perform this test the air pump is activated but the SAI valve controlling the air flow is not opened. This test was likely added in the later 993 ECU.

Under these conditions the oxygen sensor voltage should not be affected. A fault will be detected if the valve has a leak, or if you have added this circuit.

It would be relatively simple to add this extra signal to inhibit the operation of the circuit unless both the pump is on and the valve is open.

John

Lorenfb

"I believe that reason the circuit doesn't work for later cars, (and why the CEL comes on after the engine has been driven for some time) is due to a test of the SAI valve that is done with the engine warm."

A simple test could be done to verify whether this really occurs or not,
and results in a non-typical SAI fault:

Once the cold start SAI test has been completed, just unplug the air pump.
If the previous faults don't re-occur, then the system does test the valve.
I doubt this test was incorporated prior to the 996, as I've never heard
the pump run once the engine is warm on the many 993s ('97/'98) I've setup
for emissions.

Again, the CEL will not come on at the same time the fault occurs.
A perfect example was a 993 I did this week, which the shop indicated
that they had fixed all mis-fire problems, but that the readiness codes would
not set. The vehicle had no CEL. Once the trip began, no cycle flags were
being set. A check of the faults indicated mis-fires without the CEL being
on, which prevented any readiness tests being run.

william_b_noble

Loren - maybe I'll have to bite the bullet and fix this properly by getting the gunk out of the ports - is there any place that shows a cross section so I have a clue what the little holes above the valve stem do once they turn inside the head? In the past the procedure of using solvents like "sea foam" and compressed air works for a while - I made block off plates so I can put pressure on one port at a time but it still takes days and a lot of patience, hence my desire that a bypass circiut work - but maybe a better way is to add some ports onto the manifold for air injection rather than using the ports through the head?

Lorenfb

This may be helpful. See the bottom of the page: http://www.systemsc.com/pictures.htm

ltc

Excellent information.
Yes, it would be trivial to add this hardware logic to the current circuit design and produce a new PCB.

ltc

Do you have any details on this new (SAI valve) signal in a 993?
I don't have any wiring diagrams or technical information for 993's.

I have 6 boards left over from the 100 boards built, so I should be able to hack together a modification and then see if someone with a 97 or 98 can test it.

Jeff96-993

The spirit of Rennlist, always working. I do hope (expect) that you'll figure each variation out, just like you did for the earlier years. Will be fun to see the end result, as well as what's learned along the way.

william_b_noble

The existing circuit is activated by the presence of +12. I see a couple of possibilities - the simplest is to add an PNP transistor, such as a 2n2907 in series so that both the SAI and the pump have to be active. I suppose that another of the same FETs could (would?) be a better choice but I am not as conversant with using FETs - so any advice could be helpful. a third way is to use a relay. I made my circuits with a quick disconnect, so it should be easy enough to patch that in, but I can't this week.

ltc

If you wish to go old school, you could recall the days of DTL (before TTL came into the mainstream) and build a DTL NAND gate with just a couple of diodes and resistor added to the existing circuit ... (no, no need for vacuum tubes)

http://en.wikipedia.org/wiki/Diode–transistor_logic

It is just a matter of confirming that the new signal is active high true (i.e. 12VDC = ON = VALVE OPEN)