SaabCentral Forums banner

1 - 20 of 38 Posts

·
Registered
Joined
·
249 Posts
Discussion Starter #1
1990 vert turbo 5 speed. LH 2.4
Could use some help trouble shooting. Recently,I started having issues with my idle fluctuating. Once the engine is up to op temps, it revs up and down between about 800 to 1500 rpms. Driving down the road seems ok.
CEL is on and code pulled initially was 12223 - "Air fuel mixture to lean" and then later got this code and another: 12112 - "Oxygen sensor self-compensation circuit problem (system unable to reduce idle speed to correct level)"

What I've done so far:
1. checked for vacuum leaks - appears all good. If i pop a vacuum line off, engine runs noticeable worse. But doesn't seem to run terrible bad other than the idle hunting.
1. Checked 02 sensor output voltage - fluctuates btween .4 and .7 until idle hunting starts, then its fluctuates radically between milli volts and .5 volts or so.
2. Replaced 02 sensor with spare with same results.
3. checked AIC valve resistance - 8.9 ohms between pins (2 pin AIC valve) - Good according to Bently.
4. Removed AIC and flushed out ports with brake cleaner. There was some black gooey stuff in the ports. Re-installed but idle still hunting.

Let me know if anyone has any thoughts on what i should check. At this point i'm considering throwing a new AIC at it.
Thanks,
 

·
Registered
Joined
·
2,512 Posts
With my humble experience with these cars, I read the issue as Air Mass Meter related but have not had turbos to play with - so this could be a waste of your time to read the following.

Items worth scrutinising very carefully.....

Idle Speed.
Remove and Clean throttle body.
Confirm Throttle stop screw - 1/2 turn clockwise after first contact with throttle plate. Throttle stop is to prevent butterfly from jamming in barrel - not an Idle Speed Adjustment.
TPS - With throttle at rest, confirm switch contacts 1 & 2 are closed when hot and cold. Contacts should open at initial movement off throttle stop.
Check / confirm one way check valve in 8mm vacuum hose from manifold to valve cover PVC tee.

As to codes 12223 and 12112....
This is difficult to focus to one item - especially as this below will not be an orthodox approach .....
In any case, the elements of interest here are centered around the AMM....

AMM
They do age. They can loose their accuracy of measurement at low air flows as the wire sensitivity drifts out of spec.

12+ at AMM plug.
Should be minimal 'difference' between running B+ and 12+ pin of AMM. ( eg: alternator output as reference). Any more than 0.5-1.0 volts drift from the alternator o/p may suggest too much voltage loss across the loom.

GND at AMM plug
This is a difficult one......On the last 900 I had had, I replaced the ground wire from the AMM plug to the head...including thicker gauge wire, new JPT terminal ( to all AMM wires) and ring terminal.
The reason behind this step, the car had constant codes of 12112 and 12111 - idle/lambda mixture errors. With the use of a $800 hi-resolution Fluke 289 meter, I could measure DC mV across the gnd wire suggesting poor gnd connectivity and or noise in the circuit. This action of the wire/lugs replacement cleared the error 12112 & 12111 codes.

GND at ECU.
After the result of the new gnd wire to the AMM, I added new/extra gnd wires at the ECU connector as insurance.

Yep - overkill but car gained an extra 50km to a tank of fuel after all the new gnd wires.

GNDS at Head/Chassis/Battery..!
.... need to be 100%


Give the AMM plug a shake when the engine is running.
Does it still have the protective boot on the connector?
 

·
Registered
Joined
·
249 Posts
Discussion Starter #3
After posting last night i went back out and was hunting around vacuum hose leaks again, and found a cracked piece of hose between charcoal canister and hard black vacuum line in front left fender well. It didn't look or feel like it was leaking but i replaced. Let car run until back up to operating temps but idle stayed normal. Let car run for another 5 or 10 mins but idle fluctuation never returned. I'm not convinced the issue is gone though because sometimes the idle hunting doesn't return until i drive the car for awhile.

Answering Rodentmasters post:
i'm going to remove and clean the throttle body next.
Both the intake manifold and ground and ground near radiator were recently, thoroughly cleaned and tightened so i think i'm good there.

I checked the connector at the amm by wiggling and found no difference. I did notice that part of the rubber boot has been chewed away by something but the wires all look ok.

I will see if i can check the voltages you suggest and read up on this in the Bentley.
Really appreciate your well thought out and articulate reply, Rodentmaster. Thank you!
 

·
Registered
Joined
·
249 Posts
Discussion Starter #4
Update: I removed and thoroughly cleaned the throttle body. While out i checked the TPS for continuity and all checked well. I adjusted the TPS and the throttle stop to spec and reinstalled everything.
Afterwards, she started but barely and idled really low. If i revved it up and let it drop quick, she would die.
After about 3 restartes she surprisingly began sound ok or normal idling at about 850. I let her warm up until the fans kicked on. So far so good. I turned the A/C on and let it run like that for minutes. Still running ok. Turned the A/C off and within few minutes she started running rough and then the idle swung up to about 1000 - 1100 and the CEL came on. Let her run for few more minutes with no difference. Same issue. Pulled the codes but this time only got the 12223 - Air Fuel Mix to lean. Not sure what to try next. Might start taking closer look at the AMM.
I have a spare AMM from an 87. Different part # but the connector is the same. Can i use this to see how it runs with different AMM?
 

·
Registered
Joined
·
3,381 Posts
Update: I removed and thoroughly cleaned the throttle body. While out i checked the TPS for continuity and all checked well. I adjusted the TPS and the throttle stop to spec and reinstalled everything.
Afterwards, she started but barely and idled really low. If i revved it up and let it drop quick, she would die.
After about 3 restartes she surprisingly began sound ok or normal idling at about 850. I let her warm up until the fans kicked on. So far so good. I turned the A/C on and let it run like that for minutes. Still running ok. Turned the A/C off and within few minutes she started running rough and then the idle swung up to about 1000 - 1100 and the CEL came on. Let her run for few more minutes with no difference. Same issue. Pulled the codes but this time only got the 12223 - Air Fuel Mix to lean. Not sure what to try next. Might start taking closer look at the AMM.
I have a spare AMM from an 87. Different part # but the connector is the same. Can i use this to see how it runs with different AMM?
I doubt you can use the 87 AMM for a test ( different versions of the Bosch Jetronic between my87 and my90)
Some possible considerations for lean mix related cel code... Failing O2 sensor ? Fuel pump pressure delivery or faulty return valve ?
 

·
Registered
Joined
·
249 Posts
Discussion Starter #6
I swapped the O2 sensor out with spare and got same results - see my original post. I suppose both O2 sensors could have same issue but unlikely. I will go back and monitor the O2 sensor volt readings again though.
I have a brand new Glow shift fuel pressure gage which is reading from the fuel rail - pressure readings are 30psi at idle and jumps up to 35psi or so when i rev the motor. Bently says 27.6psi at idle is good.
By "Return Valve" do you mean the Fuel Pressure Regulator? I dont know of any other valve inline with the fuel return to the tank.
Also, i was wondering if the FPR was leaking fuel through the diaphragm but when i pull the vacuum line i would expect to see some fuel dripping but i do not see any.
 

·
Registered
Joined
·
12,820 Posts
It sounds like it's running very rich. What does the exhaust smell like?
Very rich (other than leaking regulator) is usually air mass meter, NTC circuit (resistor, connector), vacuum leak, or ECU.
 

·
Registered
Joined
·
2,512 Posts
This.....
Checked 02 sensor output voltage - fluctuates between .4 and .7 until idle hunting starts, then its fluctuates radically between milli volts and .5 volts or so.
With this O2 probe voltage oscillation ..... 0.4 < > 0.8 ...... the core operation of fuel mixture control appears sound which, should suggest all items are working .........then ...
"until idle hunting starts"......
Is this when the engine warms up or at any indiscriminate time?
 

·
Registered
Joined
·
5,543 Posts
A hunting idle and low O2 sensor voltage definitely suggests actual lean running. If vacuum leaks are for sure eliminated, I would want to correlate NTC sensor readings with O2 sensor readings and a IR temp gun, then I'd probably replace the MAF.
 

·
Registered
Joined
·
249 Posts
Discussion Starter #10
Started it up cold little while ago. running really rough and bogged on acceleration. Vacuum was at about 15hg. Checked resistance on NTC just before start up - 1.5koms.
As she got closer to operating temps started running better and idle came up to normal. When the fans kicked on the idle went up to about 12 and even up to 1500 rpm.
During all of this i had the meter watching the O2 signal. I never seen it get out of the millivolts. I'll have to get more precise readings from the O2 sensor next time. When the idle was up around 11-1200, vacuum gage reading was at 19-20hg.
Directly after i shut it down, checked the NTC - 224ohms Temp gage read almost 200f
I also checked the plugs before i started it up. They all look clean and normal accept for maybe cyl 1. Plug insulator a little more brown then the others but certainly not sooty.
 

·
Registered
Joined
·
249 Posts
Discussion Starter #11
This.....


With this O2 probe voltage oscillation ..... 0.4 < > 0.8 ...... the core operation of fuel mixture control appears sound which, should suggest all items are working .........then ...
"until idle hunting starts"......
Is this when the engine warms up or at any indiscriminate time?
Only after warm up to operating temp. Typically after fans turn on and off. At this point sometimes the idle will go up and down. Other times like this last test, idle went up to 11-1200 and pretty much stayed there.
 

·
Registered
Joined
·
249 Posts
Discussion Starter #12
It sounds like it's running very rich. What does the exhaust smell like?
Very rich (other than leaking regulator) is usually air mass meter, NTC circuit (resistor, connector), vacuum leak, or ECU.
Jim, exhaust seems normal to me. Not smokey or odd smelling. I think the NTC is functioning normally (See my previous posts).
For vacuum leaks, after i let it get to operating temps. I started pulling vacuum lines off the manifold one by one and plugging. vacuum gage maintained same reading - 19-20 hg depending on the what the idle was doing (which was high).
The only sort of test i've done so far with the AMM was pull the plug while she was running. She immediately started running really bad and wanted to die. Plugged it back in and it went back to normal.
 

·
Registered
Joined
·
2,512 Posts
.....do you have any chance of obtaining a can of spray freeze?
If so, spray some freeze on the metal heatsink of the AMM when the car is in the fault condition.......
or.....
Find a way to deflect the hot air from the fans - away from the AMM to see if the behaviour of the fault changes..
 

·
Registered
Joined
·
2,512 Posts
Whether the early versions with the metal barrel or the plastic body with a metal base at the connector, the DC amplifier inside is mounted on the metal for temperature stability.....a heatsink.
Just like the metal block/ heatsink for switching/DC amp - driving the ignition coil.

The clue which needs to be reviewed for the idle fault here is the nature that the issue is promoted after a heat cycle of the fans.
90deg air blowing across the AMM that it possibly becoming unfaithful, will be soaking up the heat.
In normal situations, the AMM should be able to work in such conditions but if the internal DC amplifier is showing signs of breaking down, the additional heat will assist in exposing the issue.
Cooling the AMM down may mask the issue.
 

·
Registered
Joined
·
3,381 Posts
Whether the early versions with the metal barrel or the plastic body with a metal base at the connector, the DC amplifier inside is mounted on the metal for temperature stability.....a heatsink.
Just like the metal block/ heatsink for switching/DC amp - driving the ignition coil.

The clue which needs to be reviewed for the idle fault here is the nature that the issue is promoted after a heat cycle of the fans.
90deg air blowing across the AMM that it possibly becoming unfaithful, will be soaking up the heat.
In normal situations, the AMM should be able to work in such conditions but if the internal DC amplifier is showing signs of breaking down, the additional heat will assist in exposing the issue.
Cooling the AMM down may mask the issue.
I like the logic Rodent , Maybe a backyard approach but in the absence of freeze spray , disconnect the Fans and sit an ice pack on the AMM ?
 

·
Registered
Joined
·
249 Posts
Discussion Starter #17
I see where Rodent is going with this. I dont have freeze spray of course but i could set something up to deflect the hot air from the fans away from the AMM and even get one of my garage fans setup to keep the amm cool. I will say that when the fans come on, it gets uncomfortable hot standing next to the front of the car.
As a side note, i bought a used AMM from ebay last night. if it turns out to not be the issue i will have a spare.
 

·
Registered
Joined
·
249 Posts
Discussion Starter #18
I got some better readings from the O2 last night as well.
I let engine cool down to about 120F and started it up. Here are some of the O2 sensor signal readings as the engine warmed up.
.56vdc @120F
.64vdc within few mins
fluctuate between .2vdc and .7vdc @130F
Idle increased @190F to about 1100 to 1200rpm
At this point O2 signal dropped to 154mvdc
O2 signal voltage continued to gradually drop to 113mvdc at which point the fans kicked on @ 200F
With fans on, O2 dropped down to 88mvdc and Idle was @ 1300-1400

I see now that Rodent thinking the fans heating up the AMM and initiating the problem can not be correct since as you can see from above test, the idle came up before the fans kicked on.
Sorry for misinformation earlier.
 

·
Registered
Joined
·
2,512 Posts
For anyone interested.....#013 / 019 Wire element/assy
274126
 

·
Registered
Joined
·
2,512 Posts
#005 AMM

274127
 
1 - 20 of 38 Posts
Top