[onapprovedcredit]

Was recommended to post this here since the earlier 9K’s have engines more present in the 900s.

Idle Issues - what's left to diagnose?

I recently purchased a 1988 9k with 176,000+ miles on it. Overall the car drives great and has a very small list of to dos given what I paid for it. The most troublesome issue I am facing right now is a idle/cold start issue that does not seem to cure with the normal remedy's. So far I have...

www.saabcentral.com

Link has what is going on as well as a video of startup - feel free to leave your suggestions. This is really the biggest issue the car has outside of general maintenance. Would love to get this dialed in!

 

[jvanabra]

Looks like a bad MAF to me, but I would check the condition of the ignition system and O2 sensor as well.

 

[onapprovedcredit]

For further diagnosis I installed a vacuum/boost gauge - at idle (slightly high idle at 1000-1100rpms) vacuum was clocked at 18Hg. Under load gauge read a max of 5psi boost. Does this sound correct? I have factory spec logs that I’ll be looking through tonight to see if this is normal or a sign of what is wrong.

 

[jvanabra]

No - that's base boost. Full boost is 11-12psi. Probably a separate issue - cold idle issues are usually MAF, low boost is usually an APC malfunction (solenoid, pressure transducer, clutch/brake switches.

 

[onapprovedcredit]

No - that's base boost. Full boost is 11-12psi. Probably a separate issue - cold idle issues are usually MAF, low boost is usually an APC malfunction (solenoid, pressure transducer, clutch/brake switches.

That’s great information, thanks.
Confirmed my 02 sensor is in working condition, unplugged at connector and noticed a drop in throttle response and rougher idle than I already had.

will go ahead and unplug the MAF to see if there is a difference in operation. Is there another way to test it that is more accurate?

 

[jvanabra]

MAFs are essentially untestable, and running it unplugged is not advised unless you're trying to diagnose a no-start condition. The engine relies on the MAF for mixture. On LH 2.2 cars (which a 1988 is), you only have two indexes - engine speed and MAF - for mixture. If you take one of those away the car is in "limp-home mode" and will barely run, if at all. Also, "limp-home mode" isn't a thing. If it does happen to run, all you've proven is that conditions (air density, engine temperature, etc.) are just right and only having engine speed is sufficient for the ECM to dole out fuel.

Be sure you have good performance from the O2 sensor (proper swinging voltage) and you're certain the ignition system is in good shape (old wires = high resistance or arcing) and you don't have any vacuum leaks. At that point, it's probably the MAF. Last time I bought an LH 2.2 car that didn't need the MAF replaced was in like 1995. They only last 80-100k or so before they drift too far out of spec. LH 2.4 fixes that by adding a calibration function to the ECM, something all modern cars have. But LH 2.2 has no such capability, so once the MAF is "worn" you get poor mixture control (usually lean running) followed by unstable cold idle (due to lean mixture). Once the O2 sensor warms up you get closed loop and fuel trim to stabilize idle, but until then it's a rollercoaster.

I miss my '80s 9000s dearly - they were supremely underrated. Phenomenal cars!

 

[onapprovedcredit]

MAFs are essentially untestable, and running it unplugged is not advised unless you're trying to diagnose a no-start condition. The engine relies on the MAF for mixture. On LH 2.2 cars (which a 1988 is), you only have two indexes - engine speed and MAF - for mixture. If you take one of those away the car is in "limp-home mode" and will barely run, if at all. Also, "limp-home mode" isn't a thing. If it does happen to run, all you've proven is that conditions (air density, engine temperature, etc.) are just right and only having engine speed is sufficient for the ECM to dole out fuel.

Be sure you have good performance from the O2 sensor (proper swinging voltage) and you're certain the ignition system is in good shape (old wires = high resistance or arcing) and you don't have any vacuum leaks. At that point, it's probably the MAF. Last time I bought an LH 2.2 car that didn't need the MAF replaced was in like 1995. They only last 80-100k or so before they drift too far out of spec. LH 2.4 fixes that by adding a calibration function to the ECM, something all modern cars have. But LH 2.2 has no such capability, so once the MAF is "worn" you get poor mixture control (usually lean running) followed by unstable cold idle (due to lean mixture). Once the O2 sensor warms up you get closed loop and fuel trim to stabilize idle, but until then it's a rollercoaster.

I miss my '80s 9000s dearly - they were supremely underrated. Phenomenal cars!

sure enough it was the MAF - put in a new to me Bosch MAF and started right up @ 530 this morning after sitting night.
Jvanabra- your explanation of MAF and o2 sensor helped me out so much! I couldn’t understand at first how the MAF could be bad if the idle cleared up once warmed - now the Saab is much more responsive when cold started and in all RPM ranges. No more bogs. Thanks again!

 

[jvanabra]

Glad you got it solved!

 

[rawill]

I am wondering if it is the same fault on my 93 9000 CSI.
I suspected a dodgy 02 sensor, but on reading this post, and the OP's posts on the 9000 forum I am rethinking.

It revs to about 1400rmp, then usually settles to about 900/1000, say at a set of lights.
But more of an issue is the "surging" on "float" throttle.
Fine of acceleration.

Also the Check engine light comes on, but it has no stored codes.
I know technically this is the wrong forum, but my car is not genuine T5, too early for that it the CSI configuration.

 

[jvanabra]

Is that the B234I? I'm afraid I know nothing about them... I assume they are LH 2.4.2?

 

[rawill]

Yes, and yes.
B234 not lots different from the earlier model, in its pre T5 configuration.
I assume LH2.4 also, but have never had to know!
I know next to nothing about them.

 

[jvanabra]

Saab switched to LH 2.4.2 in '91 - even turbos with DI/APC use 2.4.2 then I think. Makes sense that's what yours would have as I think Motronic was only on V6 cars.

There are various things that will cause surging idle, including a bad O2 sensor, a vacuum leak, or a problem with the evap system. The MAF if possible, just like any car, but in my experience with LH 2.4 and 2.4.2 it's a lot less likely due to the calibration function. I mean, they certainly do go bad, you're just less likely to know it.

 

[onapprovedcredit]

Saab switched to LH 2.4.2 in '91 - even turbos with DI/APC use 2.4.2 then I think. Makes sense that's what yours would have as I think Motronic was only on V6 cars.

There are various things that will cause surging idle, including a bad O2 sensor, a vacuum leak, or a problem with the evap system. The MAF if possible, just like any car, but in my experience with LH 2.4 and 2.4.2 it's a lot less likely due to the calibration function. I mean, they certainly do go bad, you're just less likely to know it.

Rawill, something that worked for me to confirm my vacuum system was in good working order was to pick up a boost/vacuum gauge (Bosch has one for $25 online as a cheap tool, would not recommend for long term in fact I am upgrading mine next month) - according to Marshall (another gauge company) 16-20 is a good and functional range for a vehicle vacuum system.

this may be way off for our cars or in general but seeing this put my mind at ease and allowed me to focus on other potential issues, in my case an MAF. Hope it’s a quick fix for ya!

 

[jvanabra]

Unfortunately, on a fuel injected engine, a vacuum gauge doesn't indicate the things you might like it to indicate. Proper engine vacuum is a function of everything between the mixture sensor (in this case the MAF) and the combustion chamber, but a vacuum gauge only measures what's between the throttle body and the combustion chamber. That leaves a major opportunity for unmeasured ("false") air, which is a real problem for mixture (and thus idle) control.

A vacuum gauge is a great tool and can tell you a lot about engine health, but unfortunately it is not a reliable measure of air leaks... For that you need a smoke machine or a pressure testing device... Or really keen eyes.

 

[onapprovedcredit]

Unfortunately, on a fuel injected engine, a vacuum gauge doesn't indicate the things you might like it to indicate. Proper engine vacuum is a function of everything between the mixture sensor (in this case the MAF) and the combustion chamber, but a vacuum gauge only measures what's between the throttle body and the combustion chamber. That leaves a major opportunity for unmeasured ("false") air, which is a real problem for mixture (and thus idle) control.

A vacuum gauge is a great tool and can tell you a lot about engine health, but unfortunately it is not a reliable measure of air leaks... For that you need a smoke machine or a pressure testing device... Or really keen eyes.

thats good to know! I’ll have to keep that in mind and not consider my vacuum system 100% just yet.

 

[Jim Mesthene]

My favorite for finding vacuum leaks is a can of chlorinated brake cleaner. It's closely related to Halon as a fire extinguishing chemical.
You can spray it at suspect vacuum areas without fear of a fire. It's so effective as an extinguisher you can hear the motor stumble if it sucks some in. You can spray a whole area like the charcoal canister.
Once it's narrowed down, you can further test with carb cleaner at the same point. You can see both by reading the O2 sensor if you want.
Don't forget throttle body seals and intake manifold gasket.
Old School, but it worked for me in 1988.

 

[jvanabra]

That's a great call - especially using the chlorinated variety. Never occurred to me!

I usually turn towards a pressure test, as it most closely simulates what happens under load, something even a smoke machine won't. It goes double on c900s and early 9000s with those $#@! rubber grommets in the intake manifold.