Stuck on open loop

[suixo]

Sooo, yeah, more troubles for a change. My idle is still hunting after changing the engine, so it must be something from the old.

That's not really the problem tough it may be related so I mentioned it. Problem is I'm stuck on open loop I think. I failed the emissions test 2 times already, changed the catalyzer after the first time and went from 1.7% CO to 1.0% CO, max is 0.5% CO. Lambda value is 1.01. CA18DET btw. The second time I passed the acceleration test (between 2000 and 3000 rpms) but the idle test I could see the CO dancing from 0.01% to 1.9% pretty much randomly.

I tested the ECU in mode I and II, no blinking, light stays off (both). They are on with the ignition switch, but turn off when starting the engine and stays off no matter what.

I have changed the lambda and ECT sensor but still no blinking. I don't think it's the MAF, but I may try that next. But assuming the MAF is ok, what could be the cause of constant open loop?

And btw, what exactly is the difference between the preface and postface AAC assembly? A friend has one from the preface and it's visually identical, I could perhaps try that on see if it changes the idle behavior.

[green_rs13]

the plugs on the AFMs are different. one has 3 pins the other has 4 pins. prefacelift has a 'mixture' screw.

I wonder if your ECU engine temperature sender is faulty.

have you done a 'smoke test' as CA's tend to have inlet manifold leaks.

[suixo]

I did change the coolant temp sensor. Is it possible to replace the MAF with the AFM? A friend has one. I could change the plug on the harness, but I don't know if the ECU is compatible.


No smoke test, my only option here is to blow in with cigarettes lol. I did spray some starter fluid but couldn't sense any change.

[Rochester]

What chip do you have ? I have had this problem before and may be able to help, pm me.

[skyshack]

You need to do continuity tests on the circuits.

Lambda, both signal and heater.
Temp sensor, it won't do closed loop until engine is up to temp.

Has it been lowered? Can rub though wiring in wheel arch. Whole world of pain splicing it back together.

[ANDY black s13]

if the MAF sensor elements (the delicate wires inside) are dirty it will increase CO,
or if something is blocking or restricting slightly the air flow to the tiny hole that feeds air to the sensor elements it will raise CO level
the dirtier the sensor or the more blocked/starved of air the higher the CO reading,
the single pole/terminal temp sender is for the gauge in the cluster,the two pin square (ish) sensor next to the single pole is for the ECU temp
If it is faulty the ECU will 'be told' the engine is still cold and it will stay on cold start,dumping in loads of fuel it can't burn = high CO
carefully clean the Maf element 1st, you will need another facelift/lambda MAF if you change it for testing

AAC valve pig nose/facelift are near the same,just the idle screw is different design iirc,the 'O' ring on the idle screw can perish and leak air

[suixo]

Temp sensor is new, I checked the resistance when cold it gave me 1.75ohms (it was a hot day). I'll do a resistance check when hot.
I have cleaned the MAF before, wouldn't hurt to clean again, but it should be pretty clean. I'd like to test the voltage but it's a pain to setup, I'll see what I can do.

It's a standard stock ECU, no chip.

As for the wheel arch harness I don't know, I'll take a look. I have it on stock suspension for now but the previous owner had it lower.

And my idle screw does seem to be able to flap around a very very small amount. I'll check that as well.

[suixo]

Loom is ok.
I changed the AAC and I can clearly feel an improvement. Less vibrations at low RPM and it's not idle hunting anymore. It's stable at 900rpm.
Temp sensor gave good resistance hot at around 0.25.

I guess i should do a harness continuity test....PITA, if that's good I don't know what it could be.

edit: continuity is fine. is don't get it

[suixo]

Went to a garage do an emissions test today, I'm getting 0.25% at 3000rpm (that's a pass) but around 1.5% at idle. The weird thing is the CA is supposed to be on open loop at idle, so my problem shouldn't even enter the equation.
I also received my AEM wideband today, I'll install tomorrow and use the P4 narrowband simulation for the ECU. Maybe it will tell me something.

Correct me if I'm wrong, but it's the white wire to the ECU (middle harness connector on the plug). Anybody knows if I have to do anything with the red and black heater wires? I think the AEM has it's own heater circuit, but I don't know if I just leave the harness ones open.


edit: I just realized i've got a 44f07 ECU, isn't that a non-lambda ECU?
edit edit: yup, I'm screwed. I need either 39f00, 39f01 or 39f05. Luckily there's one on ebay. Problem solved! (hopefully)

[ANDY black s13]

check post 16 http://www.sxoc.com/vbb/showthread.p...wer&highlight=
it seems you have a non lambda ECU,that could be the problem

[suixo]

Yup, I don't know what the previous owner was thinking. The fuel economy is beyond negative. And no wonder the light wasn't flashing in modes 1 and 2 if the 44f doesn't have them lol. I went through 3 lambdas already because of not paying attention

[ANDY black s13]

Yup, I don't know what the previous owner was thinking. The fuel economy is beyond negative. And no wonder the light wasn't flashing in modes 1 and 2 if the 44f doesn't have them lol. I went through 3 lambdas already because of not paying attention

easily missed in this case and will explain the fault,shame the previous keeper didn't mention it had been swapped (assuming he swapped it?)
do a wanted advert for one posted or ebay? if getting hold of one is difficult in your part of the world

[skyshack]

I have swapped non lambda and lambda ECU's in to Lambda equipped car. Short run up nearest dual carriageway and back showed both flashed LEDS in mode I and II. To compound it the non Lambda was auto and Lambda manual.

Engine has to be up to temp for closed loop.
Lambda and temp signals have to get to the loom plug. Back probe the ECU plug with a voltmeter.
Lambda wire is thin co-ax. Thin inner core is the signal and the outer braid is an earth shield, not connected at Lambda plug.

[suixo]

I already found and ordered a 39f00 from Latvia. It's a non-aircon but I removed mine anyways.

Are you sure they are not both 39f? FSM says mode 1 and 2 are not available on decats. I did test the wiring loom and it's fine.

[skyshack]

I already found and ordered a 39f00 from Latvia. It's a non-aircon but I removed mine anyways.

Are you sure they are not both 39f? FSM says mode 1 and 2 are not available on decats. I did test the wiring loom and it's fine.

Quite sure. If the Lambda sensor is fitted and the loom has the wiring all ECU will do mode I and II. What they won't do is heat the Lambda, no transistor in non-Lambda ECU.

[suixo]

I think I'm even more confused than before. I've been searching the VIN to learn about the factory details of the car. Turns out the production year is 1991 and the AP system was an option. 1992 was the year of the first owner. And on the technical paper of the car under "Observations and authorized reforms" it says "vehicle equipped with catalyzer ".

Is it possible the 44f is actually stock on the car? If it's a 91 production it should be de-cat, but it does have the facelift bumper(it may have been changed). They just added the cat and lambda as AP option but kept the ECU? This is really weird, maybe the emission test should be for decats (that would be a pass no problem).

[green_rs13]

91/92 is where the cars got all mixed up spec wise. what seats and carpet has it got in it?

seats with removeable headrests or the seats with holes in?

dark grey woven carpet or light grey furry carpet?

[suixo]

Actually the seats are the ones with the hole, but the dude told me he got them from another S13 he had. The ones before were different. I guess removable headrest. The carpet is light grey furry.

[ANDY black s13]

I'd go with you have a lambda plug in the harness so must be a facelift,unless the entire wiring loom has been swapped
does it have a viscous lsd diff (it may still have the VLSD sticker on top of the diff casing) or an open mono wheel of death one tire fire diff
was it first registered in the UK then imported to Tenerife as I recall some odd ball euro spec 200's with a mix of parts

[suixo]

I think it comes from Germany. There's German writing on the stickers.

As for diff well... when I tested rotating a wheel the other stayed static. I'm guessing worn VLSD?