Car: 1997 Saab 9000, 2.3 turbo, automatic transmission, 362,000 miles.
I figured it was about time to consult the brain trust here, this seems to be more of a generic problem than Saab-specific...
Driving around today suddenly the check engine light turned on as I got close to home. There are no symptoms, car appears to drive normally. So when I got it home I hooked up my Tech2 to the diagnostic port.
Although crude by today's standards the Tech2 does a pretty decent job on these old barges. Not only does it display and clear both generic and specific codes, it can display live and freeze-frame data as well as capture data for later playback. Data can be displayed in numerical or graph format.
Code are P0130 with P0340 and P0133 pending. (I cleared them but they'll of course be back unless the underlying issue is fixed.)
Live data shows upstream O2 sensor varying between 0.1 and 0.9 volts as engine RPM changes. Downstream sensor stays steady at 0.3 volt. Long term fuel trim is 0.0%, short term varies between 7% and 18% as engine RPM varies.
So, as far as diagnosis since long term fuel trim is spot on and there is significant variance in short term trim this would appear to say that whatever happened just happened recently, which agrees with the check engine light having just now turned on. The O2 sensors appear to be within operating parameters. So it looks to me like I have a vacuum leak that the Trionic ECU is compensating for by adding additional fuel (short term fuel trim).
There's quite a bit of plumbing under the hood for the turbocharger, so seems like time to start spraying some carb cleaner around looking for leaks. (Any other ideas from the peanut gallery are of course welcome!)
Also adding @jack62 to the discussion
Most likely it is a vacuum leak.
That's my thought as well, there's certainly old vacuum lines and fittings on this thing, will check for that.
Hi Chuck. With multiple codes you have to pick one and then try to resolve it.
I found the P0133 code more interesting so I chose that one
On the P0133 code we can ignore the P0130 because it's just saying there's a problem and the P0133 is saying exactly what the problem is.
It's saying as the computer is switching the mixture "rich/lean, rich/lean, rich/lean...." the HO2S1 voltage isn't responding in "real time".
It's lagging beyond the programmed parameters.
But when we look at your HO2S1 voltages, they look "textbook". Switching from 100mV to 900mV.
Your downstream O2 sensor voltage at 300mV looks a little on the low side but if it's holding steady it could be OK. Or it could be a clue.
The P0133 code can be anything from a vacuum leak to an exhaust leak Upstream of the HO2S1 to an electrical issue. Add into the mix that this is a turbo and it compounds the possibilities.
But these codes are "pending" and O2 sensors are continuously monitored components.
This seems like an intermittent issue and exhaust/turbo leaks aren't intermittent.
Here's a case where looking at the HO2S1 voltage in Freeze Frame from when the P0133 code was thrown would be useful.
Is it possible that the HO2S1 voltage signal is intermittently dropping out?
I've dealt with a P0133 code before but I've never been under the hood of a Saab so I checked it out to see if anything about the circuit was "unique".
There didn't seem to be anything unique. Heater+, Heater-, O2 sensor signal, O2 sensor Ground.
So I can be analytical but I have a hard time picturing wiring diagrams in my head. For me the best way is to map them and so that's what I did.
I'm just posting it, Chuck, in case you can find it useful.
You can enlarge it in your browser or else scroll sideways.
I had to splice 2 pages so that's why it looks like there's two ECU's. It's just two diagram pages sliced together. The pinouts are what matters.
So here's the circuit description and I tried to color code the words (approximately) to correspond to the wiring diagram.
I'm not good with HTML or Font tags so the Heater circuit is Red and Black and the Signal circuit is Blue and Green.
In the case of the O2 sensors, each cable has 4 leads. Two of the leads are for heating the sensor (until the exhaust system is up to temperature) and the other two are used to convey the signal generated by the measuring cell to the ECM.
Of the two used for heating; one lead supplies voltage directly from the battery for the heating element. The other side of the heating element is connected to the ECM (pin 50 for the front & 51 for the rear). The ECM grounds these leads as required.
The other two leads connect the reference ground and sensor voltage to the ECM. The sensor voltage comes in on pins 23 (front) and 70 (rear). The negative legs are spliced together and connect to the ECM on pin 47. Depending on the O2 content of the exhaust gas the cell will generate a voltage of between .1 to .9 volts.
So nothing "unique" about that HO2S1 circuit.
If we aren't talking about an air leak (intake/exhaust) then what could be the reason?
Here, the computer is providing Grounds to control the circuits. So where is it getting the Ground from?
Could there be a problem there?
Looking at the wiring diagram we don't see where the computer is getting Ground for the O2 sensors but when I looked up your Ground Distribution Circuit I found this:
Where is THAT in the wiring diagram?? O2 sensor heater Ground, ECU Ground?? That's useful information.
From what I can tell these Grounds are on an intake manifold bracket. What if there's corrosion, rust, or a bad connection causing an intermittent issue?
Maybe putting some sandpaper to those ground points on the bracket and their stud connectors on the wiring could be worth considering.
You have that Tech2 and there's O2 sensor tests I've come across using an ISAT scan tool which yours can easily do.
Also the Ground Distribution.
I'll post those links in case you find them helpful.
https://www.manualslib.com/manual/1756299/Saab-9000.html?page=123#manual
http://z90.pl/saab/pdf/9000wiringdiagrams.pdf
But frankly, I'd like to graph that HO2S1 voltage when the P0133 code is being thrown to see if the signal is dropping out.
I don't have to tell you, Good luck with that when dealing with an intermittent issue.
...... just my 2 cents from "the peanut gallery"
Thanks, some excellent tips on further diagnosis. You are correct that there's nothing really unique to Saab about the O2 sensor configuration on this car, sensors are bog-standard Bosch 4-wire jobs and there's nothing really unusual about the fuel injection and emission control system, it's all period-standard Bosch technology. I have the complete wiring diagrams for the car and know exactly where that "G112" ground is, as well as the ECU ground, so will check those as well as checking for vacuum leaks.
As far as data everything seems to be near textbook aside from short term fuel trim and possibly the downstream O2 voltage being a little low. (Replace it? I don't like to just throw parts at a problem but really it should be closer to .5 volt.). Of course an intermittent problem isn't necessarily going to show up in a short diagnostic session. One thing I have not done yet is data capture with the Tech2 while driving the car, that might reveal more clues. As you say getting more data off that upstream O2 sensor when the P0133 comes back to check for dropouts would be interesting.
Yes, intermittent problems are real "fun" to track down! Fortunately this car does not have to go through inspection for nearly 2 years and the computer seems to be compensating OK for the fault, it runs just fine. It's been a few years since I've had to do any wrenching on this thing aside from oil changes so I can't really complain too much given the age and mileage.
Hey Chuck - Do your 02 sensors have a pre-heater or they the old ones without the integrated heating coils?
They have preheaters - the Tech2 shows the pre-heaters operational. My first thought looking at the codes was an O2 sensor problem but it looks like the data doesn't support that. (With the Tech2 I have all the diagnostics that were available to dealers and the ability to set car-specific settings, marry key fobs, etc. Not too shabby for early 1990s tech.)
I could ask around on the Saab enthusiast sites of course but this seems like a generic type problem that could just as well happen on a Toyota or Honda.
The voltage differential in the upstream 02 sensor seems suspicious.
From what I've always read, 0.1 to 0.9 volts in the upstream sensor is within normal range. There are no symptoms while driving and the only thing in that data I'm seeing that looks abnormal is the short-term fuel trim adding fuel.
https://www.fixdapp.com/blog/oxygen-sensor/
Some more data, today I let the engine warm up and did the carb cleaner test, spraying vacuum lines and all connections to the intake manifold. No reaction, no leaks found. Took the car on about a 10 mile drive, codes have not re-appeared, car runs flawlessly. Short term fuel trim is now near zero with occasional variation plus or minus 2% or so, way down from where it was when the codes came up yesterday and seems perfectly reasonable given this beast's age and mileage.
While driving I used the Tech2 to take a data snapshot while varying speeds and goosing the accelerator. Being 30-year-old technology though the Tech2 has very limited data storage and can only record a handful of parameters for about a minute. (I've been thinking of getting a more modern scan tool just for better recording capability.) Playing back the data, short term fuel trim stayed within a few percent of zero. However I did see a short dropout on the upstream O2 sensor where voltage dropped to zero for several frames. I also saw variation on the downstream O2 sensor. Although most of the time it stayed around 0.3 volts, in some frames it would drop to 0.1 volts and then spike up to 0.75 volts before settling down again. So it does appear there is some intermittent issue going on with the O2 sensors.