I am not familiar with the smoke test to know if it exerts any pressure or vacuum on the intake tract to reveal a leak. Checking the intake with vacuum would be realistic but pressurizing it is easier. It would be necessary to locate the air pressure supply away from then engine in order to better hear any leak if a leak can’t be felt with your fingers. It is possible that a leak will not reveal itself with pressure but needs vacuum but a pressure test is an easier thing to try first.
It seems that your fuel pressure tests are being done under contrived conditions. I maintain that checking fuel pressure at the injector side of the distributor under dynamic conditions, meaning when the engine starts, runs for 2 seconds and then stops is more realistic in order to attempt to determine why the started engine stops.
ok4450, have you tried inserting a long 3mm wrench in the small hole next to the air sensor plate? Remove the plug and insert long 3mm hex wrench and turn one full turn clockwise.[The fuel adjustment is very sensitive.Put a reference mark on the body.]
The fuel distribution block has very fine screens at each port. Make sure these are not clogged
Whatever you do, don’t separate the the fuel distribution block.It will forever leak at the seam.
Also make sure the plunger that is controlled by the metering plate going into the fuel metering{distribution] block is not stuck in the bore.
Please excuse me for not being around the last few days to update this. I’ve gotten involved with several projects that (as per the usual) have snowballed into something much bigger than they ever should have been and they’re chewing my time up.
(Murphy’s Law at work again.)
Will be back later tonight because after dinner I’ve got more fish to fry, unfortunately.
To address a few things, it’s not really manifold vacuum that affects the air sensor plate. (except indirectly)
It’s ported vacuum from above the throttle plate.
At times when I’ve managed to keep the engine wheezing for a few seconds the vacuum gauge needle is bouncing around but that’s to be expected on a very rough running engine. (it bounces around 14 to 17 and in the past 17 has been normal for this car and locale.
The vacuum solenoid in the intake tract and the breather pod on the fuel distributor boot have been more than quadruple checked and even blocked off to weed them out of this.
Oldschool, I have tried rotating the air sensor screw more than a turn in each direction by using the 3MM tool and nothing changes.
The screens are clean, the pressure relief valve in the fuel distributor is not sticking, etc. and I’ve even tried adding and removing a few shims on that while playing around with it. No change.
The fuel plunger moves smooth as silk with not a hint of binding and I’ve even scrounged 2 more fuel distributors out of my stash and tried them; all to no avail.
Even changed the fuel pressure regulator and warm up regulator. Nothing.
The CIS system is basically a simple one and it’s pretty easy to understand how it works. Most problems involve air leaks (rough idle or hesitation) or poor running when hot or cold. (Control Pressure Regulator)
That’s why this thing has just totally whipped me. It makes zero sense, and even more so considering it was running like a top one minute and 2 hours later - won’t run and has not run since.
Any chance something wandered into the exhaust pipe while the convertor was off, got blown downstream while it was running and is now plugging up the exhaust? Ever tried running it with the cat completely removed?
im not in the same league as these other esteemed mechanics, but one thought of mine after reading all these is you mention lack of fuel. You also mention the fuel pump, checkvalve and assembly. by chance when you replaced the Cat did you work on the fuel pump? (or fuel lines) i am thinking of the possibility of a collapsed or mangled fuel line on a hose barb fitting. scenario of gets enough fuel to push through to start, but cant keep up with the volume needed to keep it running? maybe the inner fuel line ‘liner’ was accidentally loosened while installing, and is showing these fuel starving symptoms? with not enough fuel entering the float it CANT lift, thus mimicking the same issue as a faulty fuel float.
Nothing wandered into the exhaust because it was done in one shot. Changing the converter is about a 10-15 minute deal and the car will not run with the cat in place or disconnected.
It will start right up for about 2 seconds but that’s due to the gasoline shot from the cold start valve. After a number of attempts the cold start feature goes away (normal) and it will not start, or cough, at all.
Cranking the engine over (with injectors removed and placed on the intake) does not produce any spray pattern from the injectors. Raising the air sensor plate with a magnet or with the 3MM adjustment tool does cause the injectors to spray.
It’s not a lack of fuel or fuel controls. The system maintains the proper fuel pressure with the pump relay in place or with a jumper. The correct fuel pressure is present in the fuel distributor where the injector lines are connected.
It just seems to come back to the air sensor plate not rising from the incoming air; or at least that’s the only theory I can come up with. With the intake tract disconnected I can place my hand on the throttle body, crank the engine over, and it will suck my hand down on there pretty good. The suction gets progressively weaker as the path through the intake tract lengthens.
I’ve removed the entire tract, again, and currently have it off the car. Not a sign of boot cracks, tube cracks, loose clamps, and no visible problem with the turbocharger at all. (Originally, I was even suspecting that maybe the impeller was seizing up and screwing up the airflow but the impeller turns freely with no slop.)
Both the vacuum solenoid and breather pod in the intake tract were blocked off to weed out a potential problem there.
See why this thing is frying my brain? I could only imagine what a poor mechanic would be going through if saddled with a problem like this. I’ve been dinking with this thing for months now and just flat don’t get it.
From the readings suggested previously, hope it helps!
Engine Temperature / NTC Sensor
One of the least known sensors is the engine temperature sensor (aka NTC II). It is a semiconductor resistor, also know as a thermistor, and the NTC stands for Negative Temperature Coefficient. This means that the sensor’s resistance goes down as the temperature goes up. If the ECU applies a fixed voltage to the NTC resistor, it will receive a smaller signal back as input from a cold resistor with higher resistance than from a warm one. This sensor is screwed into the water passage on the intake side of the head between intake runner #2 and #3 on the 900s and just under runner # 2 on the 9000s. Test the sensor by probing the two contact points with a multimeter to measure the resistance which should be:
i have a freind that used to make that car and i will ask him tomarrow. he worked for SAAB for 20 years and now owns 3 SAABs and all in perfect condishen
It’s not the temperature sensor because the temperature time switch as it’s called has only one purpose; to activate the cold start valve for a certain number of seconds depending on temperature. The cold start valve works fine; and is apparently the only thing working fine.
So here I am; an ex-long time SAAB tech with quite a bit of CIS experience and a stack of service school diplomas; and not a clue to go along with any of it.
Once upon a time I had a VW rabbit that had some sort of electrical setup that powered the ignition through the starter circuit with the starter engaged, but then through a different circuit when the key was released. It would start, but stop immediately when I let go of the key, because the fuse on the firewall was blown. It was a flat piece of metal, not a traditional glass or plug in fuse. The spark stopped when the key was released.
Of course, the fact that your Saab starts only after sitting leads away from this idea, but what the heck. You seem to have tried everything about the gas supply.
Have you tried applying vacuum to the fitting on the air sensor plate from another source? Like a vacuum pump, or some other suction device or even your dumb teenage nephew? I don’t know how these cars work, but if you are not getting vacuum from the engine, and the car runs using vacuum from elsewhere, you have pinned down the problem.
The boost pressure only comes in when the pedal is pushed pretty hard. Air is still being pulled through the fuel distributor housing (which also houses the air sensor plate) and this airflow, even under boost, is what keeps the plate raised.
Whenever the plate is raised, even a small amount, the injectors will start spraying.
(On a normal system anyway.)
My bad, the car does have the temperature switch as it’s a Lamda car. Non-Lamdas do not have it.
The switch tests fine but even if faulty it would not cause a problem like this. (It provides for enrichment but would not cause a total lack of fuel to be discharged from the injectors.)
I’m going to put it back together this weekend (again) and see what happens but I have zero optimism.
You said the front of the intake tract had little vaccum. Later you added that the back of the tract had good suction so the thought is there is a leak somewhere in the intake tract. If vacuum is not developed in the tract=no ported vacuum=no plate lifting.
This is the solenoid I was referring to as maybe leaking back into the intake tract. There’s a brief discussion on the normal operation. What happens if it gets stuck open? I don’t know exactly how it’s plumbed, they don’t show that kind of detail but thought it was worth mentioning.
There is certainly a lot of mystery meat as to how it works but have to wonder if what you are looking at is a symptom of a different problem.
(1) Knock sensor
ok4450, this is my gut feeling. Fuel/computer relay is cycled alot.Probably worn slam out.
It is also located at the RH kickpanel. Alot of moisture, especially if antennae is located at that area. I know most antennae were located on rear fender, but this area is a moisture accumulator.
I know you want to rule out the computer. Don’t be stubborn. If you have a spare, replace it.
Exactly why I originally suggested looking in this area back a few days ago. OK’s a sharp cookie so it’s likely to be something obscure. However, I have no first hand knowledge of the design so could be barking up wrong pipe…
When dealing with thermistors, be aware that depending upon their beta (curve of temperature change vs. resistance) and bulk, they can actually change resistance when you measure them with lower impedance multimeters. High beta thermistors change resistance very fast with small changes of temperature, even just the temperature change caused by current from a low impedance multimeter. I have even seen small ones oscillate with a multimeter, at a very low frequency. It is indeed a semiconductor, specially doped for a negative temperature coefficient.
