Idling - poem of unplugging things, O2 sensor and more... [Archive] - SaabCentral Forums

: Idling - poem of unplugging things, O2 sensor and more...


Bartek
19-03-06, 08:01 AM
Hope this forum is for helping, not for killing annoying people :roll:

Still need help. Briefly what's going on.
Car is '92 LPT with Lucas CU 14.
Problem: swinging idle.
Symptoms: when starting from cold idle is rock solid about 1000rpm. After some time of idling it starts to swing rather down than up - form 1000 to 700 then back then goes to 700 then 900 etc.
I've noticed rmps starts to swing faster when I drive - first junction, foot of the gas, clutch and rmps swings - than if I let the car idle.

What was swapped: AMM for another 2, AIC for another good one. O2 sensor for another but I need to write more below about lambda.

I started from the obvious things- leaks, squeezed every hose, then unlugged all of them one by one. Blocked AIC hoses to see if there's any leak after throttle. No change. Checked faults - only one was stored but it was lambda and it was re-swapped. HG was recently changed with all other appropriate gaskets.

AIC - cleaned, checked for movement and resistance. I even tried Sturla method of blocking 2 big hoses going to and from AIC to find a good idling speed - I found it but even with AIC unplugged tickover was erratic - even at 650rpms level.

Blow off - checked.

Hall sensor - swapped for another whole distributor. No change.

Vacuum - at idling -0.8 bar (11 psi)

O2 sensor. I've tried to swap it for another NTK which was at another LPT donor car. It wasn't OEM so the car's started to run pig rich allowing me about 350-400km from one tank in city driving. Cheked with multimeter and sensor used to give me almost constant of 0.25 V which meant the richest mixture ever. It didn't react on throttle moving. There were some variations but in 0.23-0.26 V range. Swapped back for my previous one. My old one swings from 0.1 to 1.3 which could be bad thing BUT car doesn't react on UNPLUGGING sensor. Of course for a while rmps changes but swinging lasts. I guess it means that those strange lambda fluctuations are the effect of something not the reason.

AMM - checked another 2. With various option - 336 and original around 500. No change. Strange thing - unplugging AMM at all has always the same effect - car stars normally, CE light goes on, car runs fine for about 30 seconds (already warmed engine) then stalls and even foot on the acc pedal can't ressurect it. Those 30 seconds (at warmed engine) is exactly the time for engine to start swinging. With AMM unplugged tickover is OK.

NTC - resistance OK, but again - can't start it with another one just plugged in ( not plugged in ehgine block), also can't start fully warmed engine with NTC shortened to simulate car at operational temp.

Questions:
1. If engine goes well for a few first minutes of running (from stone cold) does it mean there's no vacuum leak?
2. ECU starts to see something after exactly 30 seconds of running with fully warmed engine and that thing is responsible for swinging. What could it be?
3. Oxygen sensor - should unplugging it has the effect of stabilising idling when it is shot?
4. Could it be ECU itself? Against this speaks the fact that the time of correct idling changes - from 3 minutes at cold engine to 30 seconds with fully warmed. AFAIK ECU takes control in less than 1 minute after starting the car - no matter if engine is cold or hot.

Heeelp :cry:

Willy Wonka
19-03-06, 08:27 AM
Since you get a constant reading on the O2 and the issue starts when warm, it could be that when the ECU is in mapped mode it's ok, when it goes to adaptive mode, some thing is preventing the swinging of the O2 values.
In adaptive mode the O2 should be varying as the ecu limits gas as well as adds air to control the idle speed.
I would check the supporting items to the AMM first. The manifold pressure transducer is a place to start. You can test it from the engine compartment by checking that the lines to it hold a vacuum. The temperature transducer would be next. Check it's resistance when cold and warmed. (search SC for correct values)

I was told by a Saab tech to try to set the idle manually close to 800, so as to minimize the swings by the ECU.

At idle with a warm engine the O2 readings should swing. If they don't you are still in mapped mode. The cable from the O2 to the ecu becomes a shielded one, so check that the connection is to the core wire, not the outside shielding cable. Check the O2 heater fuse.
Partially block the exhaust like a cat would and let it warm, check for O2 swings.
I'm sure you checked for fuel pressure. You should be able to hear fuel going past the fpr with the engine off and key on.

Luck

Bartek
19-03-06, 08:49 AM
Thanks Willy. I think it's not just a Lucas thing - the rules are the same for Bosch'ed cars, so every theory could help.

1. Since it's LPT it has no pressure transducer - only overboost cut switch. I unplugged it at all some time before.
2. Can't set idles manually - Lucas is quite original and it won't let you do this :).
3. Will check shield today to see if there's no unneeded voltage, but: if I unplug O2 sensor ECU gets into closed-circuit mode, which should make idling stable - if lambda is shot, right?
4. From what you say it could be ECU, 'cause once it gets into adaptive mode idles goes nuts. I'll try to swap ECU today for a known good one. One more question - what decides about turning from mapped into adaptive mode? Lambda preheater, which means lambda itself?
5. NTC sensor - one thing can't let me sleep well. If I have another one used sensor which gives proper resistance at least when it's cold why engine can't start with this sensor plugged alone? Coolant doesn't close the curcuit.

Willy Wonka
19-03-06, 09:53 AM
The temperature determines what mode the ecu runs at. I don't know about Lucas first hand, but I find it difficult to believe that they would not use a manifold pressure/vacuum sensor. It could be done, but what an approximation. Good luck. Lucas uses a throttle potentiometer. Check it.

The Lambda/O2 preheater are the two similar colored wires, of the 3 wires that go the O2 (bosch). It heats the O2 sensor.

cdaly
19-03-06, 11:43 AM
The Lambda/O2 preheater are the two similar colored wires, of the 3 wires that go the O2 (bosch). It heats the O2 sensor.
This depends. my '92 LPT has a three wire O2 sensor with black for signal and red + white for the heater. This makes it a titanium sensor. If the heater wires are both white, it's a zirconium one. They are not interchangeable.

O2 sensor. I've tried to swap it for another NTK which was at another LPT donor car. It wasn't OEM so the car's started to run pig rich allowing me about 350-400km from one tank in city driving. Cheked with multimeter and sensor used to give me almost constant of 0.25 V which meant the richest mixture ever. It didn't react on throttle moving. There were some variations but in 0.23-0.26 V range. Swapped back for my previous one. My old one swings from 0.1 to 1.3 which could be bad thing BUT car doesn't react on UNPLUGGING sensor. Of course for a while rmps changes but swinging lasts. I guess it means that those strange lambda fluctuations are the effect of something not the reason.
Did both sensors have the same wire colours? when you unplugged the sensor, did you unplug the black wire rather than the heaters?

Bartek
19-03-06, 12:38 PM
3 wires, black single one for signal for sure. I unplug this one. This is original part - Lucas a3g or something :). Voltage swings from almost 0.1 to 1.3 V.

I'm after replacing NTC for another used one and I've noticed some other symptoms. I was at my friend's place, he owns lucas conv. I checked how car behaved when unplugging things. Well, AMM off - both cars died. NTC off while running - his perfect idling car starts rumbling in a sluggish way at very low range rpms - immediately. In mine rpms rose a little and nothing else. I don't know what could it mean.

I also tried with another TPS still holded to another throttle body. No change at all.

My assumption is:
While being in adaptive mode ECU gets faulty signal from some device and while trying to adapt rpms it constantly swings 'cause it can't get proper mixture. But unplugging things doesn't change anything - doesn't exclude faulty signal, so what's going on?

900t
19-03-06, 02:47 PM
Maybe the ECU itself is faulty?

Bartek
19-03-06, 02:59 PM
AFAIK there's no other way to check it but swapping another one?

I've recently went through similar symptoms at one of the saabnet users. It appeared car had misaligned camshafts, but I guess it can't be connected with my problem because my idling is steady with cold engine. One more question - after all descriptions could I be sure for 100% it's something reliable from ECU or ECU itself? Or do I have to think about some mechanical problem (engine wear, leaking injectors, etc) ???

RoccoH
20-03-06, 05:56 AM
..
O2 sensor. I've tried to swap it for another NTK which was at another LPT donor car. It wasn't OEM so the car's started to run pig rich allowing me about 350-400km from one tank in city driving. Cheked with multimeter and sensor used to give me almost constant of 0.25 V which meant the richest mixture ever. It didn't react on throttle moving. There were some variations but in 0.23-0.26 V range. Swapped back for my previous one. My old one swings from 0.1 to 1.3 which could be bad thing BUT car doesn't react on UNPLUGGING sensor. Of course for a while rmps changes but swinging lasts. I guess it means that those strange lambda fluctuations are the effect of something not the reason...

the swinging voltage means your lambda sensor is working perfectly! :cheesy: By loading the cat with a constantly changing rich and lean mixture the cat will work at its optimum. So at least the emission control part of the ECU is working correctly. On a Bosch LH injection system the swinging should be between 0.1 V and 0.9 V and it should swing every few seconds from max to minimum voltage (I think it was something of ten swings per minute).