cfc928gt

Idle speed for automatics is with the car at normal operating tempurature and the transmission in drive.

TheoJ

Yes, 675 +/- 25rpm is normal, idle, not in drive, and warm engine. Initially when cold it is at 1000rpm and gradually steps down to 675. Now, 675 is not too exact to be honest, but 900 or 1100 is just too high IMHO.
Spec on "D" selctor? Not sure but I would say about 650 or so, a little below idle but not much. If the idle is high, it will drop considerably, because of the increasing load from the Torque converter at higher rpm's. I can check if you like.

Carefully check the crank ventilation, oil filler, dipstick seal, valve covers etc if you did not do that already.
regards

cfc928gt

I'm really surprised to hear idle in drive for an automatic is somewhere around or less than 600 rpm. If you're saying 650 rpm in neutral is in spec I would think in drive it would drop to 550 or 575 and I can't imagine it would idle very well. A 5 speed S4 idles at 675 +/- 25, I don't think an auto is supposed to be less than that in drive.

stevedavies

Theo, at cold rpm is approx 800rpm. I tried pressurising intake again yesterday looking for leaks. Took it to 10psi, heard small leaks at dipstick, oil filler cap etc but couldnt find anything significant.
I agree 900 & 1100rpm (displayed) is too high, but 900 sounds better than 1100rpm!! I am making progress, hoping driving more mileage will allow LH to adapt to lower rpm. Youu mentioned crank ventilation, is that part of vacuum ?

TheoJ

Cold could be 800, sure. But hot is pretty close to 675. Let me check RPM's on my car, just to avoid confusion. I get back to you tomorrow.

The crank is supposed to be "vacumend, air sucked out" and routed back into the intake. Oil vapors are digested. That creates a low pressure which should hold. I once saw a leak at the oil intake (connected to the crank) which caused a high idle, and the dipstick o-ring is also famous.

Hilton

Steve - out of interest, did you disconnect the battery before using the Spanner to do the idle adaption?

The LH doesn't retain any parameters in flash memory, only volatile memory, so it loses all adaption parameters when the battery is disconnected.

If you disconnect the battery and leave it for 20 mins, then re-run the idle adaption, you should be able to get it to adapt correctly for your vehicle. Any change from spec at that point indicates a problem (e.g. unmetered air leak, O2 loop, MAF, LH, sensor inputs etc.)

edit: don't trust the tacho needle - my 89 S4 reads 100 rpm out, but I had a datalogger wired into the car which shows the rpm signal off a coil (from a pin on the diagnostic plug), and could verify that the idle speed was 675 (it actually fluctuates slightly - from around 665 to 690).

Also note that the table posted above, the far-right column is the specs for a GT (idle speed 775), due to a lumpier cam. All S4's should be 675.

stevedavies

thanks Theo

TheoJ

Hilton, not to start an argument, but take the schematic and find RPM signal. Check where it originates from, find the EZK, and see that the TDC Sensor is the source, Ezk processes and provides the pulse train, and the LH, kickdown, diagsocket, and dash are wired on this signal. Nothing coil related on the diag bus. I'm afraid, and trust me, you're wrong where.

Steve, RE idle RPM: I will check mine and let you know. No problem. I can even put my freq counter or scope on it if you like.

regards

Hilton

Interesting Theo thanks.

I'm not sure on the exact origin of the signal - I just recall I had to tell my Datalogger software that the car is a 4-cyl to correctly decode the rpm signal, so I assumed its a coil signal.

I can't find my diagram of the pinout for the 19 pin connector to see what it actually is

edit - just found the pic, its in the 89 Service Tech document.. pin 14 labelled tD. Any idea what I'm actually logging? Its the same as LH pin 1 I think?

the flyin' scotsman

Steve............the table that I posted is directly from the Porsche Service Information for your model year of car. The table lists, starting column 2:

ROW S4
USA S4
ROW GT
USA GT

Sorry if there was any confusion but the bottom line is your car should idle 675 +/- 25 rpm.

The eprom in the LH for an auto trans is programmed to compensate engine idle for gear engagement and a/c operation. I learned this from John some years ago now when I was experimenting with GT eproms (both LH and EZK) in my car. I dont know how much compensation perhaps John can share.

John metioned he did not open the ECU when testing............is it possible the previous owner installed a different eprom? When you have it out to test with the loaner LH I'd open your LH and have a good look.

the flyin' scotsman

Theo...........I posted on this thread that the rev. signal could be found at one of the coils. I stated this from a brief search on Rennlist but perhaps it was in reference for an older model. If this was inaccurate at least it gives Steve something else to ponder in his search for a true accurate idle.

Appreciate your 'positive input' to this thread...........

stevedavies

So after 20 miles yesterday the idle at hot went down to 900rpm and was steady. Today its back to 1100rpm again!! Its exactly the same as before. Its really wierd. I am going to take up John Speake offer of a loan on a LH and EKZ ECU. its about the only thing left to try!! Its really strange after all this time it reverts to 900rpm for a day and is now back to 1100rpm. Will be trying some of the latest suggestions above on the weekend and report back

stevedavies

malcolm, re the LH ECU, I think John said my LH was unopened when he checked it, so assume its original eprom fitted

John Speake

Yes, the LH had never been opened.

Hilton, it is only necessary to disconnect battery ground for a couple of seconds to loose all the adaptation stored SRAM data. Steve did do a battery grond disconenct before reading fault codes and doing idle adaptation. Lots of emails have been exchanged off list ! The rpm signal on the diagnostic socket is the same as pin 1 LH.

the flyin' scotsman

So prior to the spanner work we had a 1100 rpm idle.

Batt. disconect, full spanner diagnostics including idle adaption resulting in 900 rpm idle after short trips.

A day or 2 later the idle reverts back to 1100 rpm idle.

Given the above I'd still question what signals the LH is receiving to affect the idle after all was reset such as the temp II.