Hi there,
I've been having a problem with the lambda light for quite some time now. Searched several forums but didn't get to the solution yet.
The case is:
- the Lambda/Check Engine light is ON irregurarely. Sometime within minutes after reset and in other cases after weeks of using the car
- from the moment the light is ON the car starts running unevenly, which is especially clear when running stationary: it varies between 500 and 1300 rpm and sometimes stops.
- When the light is still OFF the car runs great.
- the fault code that I find everytime is 1-2-3, meaning a coolant temperature sensor error.
- Although the resistance I measured at the ECU side of the sensor seemed to be OK, I renewed the sensor. But with no result.
- I also tried another ECU, a new air mass meter and put an extra earth lead on the engine block, but still no result
I had this problem now for over 6 months, and searched the internet for tips to solve it, unfortunately so far I didn't find it.
Perhaps there is someone who can help me on this issue?? I would appreciate that!

 

You can test to see if your O2 sensor is making the idle fluctuate by - disconnecting it and driving the car and checking for those symptoms. Yes a code should certainly get stored due to the lack of a correct signal - but the Ecu is tuned to have the car run under normal circumstances. The O2 sensor tries to make the mixture perfect so the converter works most efficiently.

For your temp code - Had a customers car years ago that had a recurring 123 temp code. Tried everything as you have. Finally replaced the spark plug wires - and no more 123 code.

Is the 123 code the only one stored - no code related to the O2 sensor?

 

I'd look at the MAFS (Mass AIrflow Sensor). It creates all kinds of weird symptoms. I had a similar situation with a '93, the geniuses told me it was the ECU, etc, nothing worked until I started changing all electrical components and the MAFS swap did it, like magic! (I have a large spares box)

 

Thanks. I already changed the MAFS, but unfortunately in my case it didnt solve the problem

 

I will try Hoonks suggestions asap.
There is only 1 fault code btw and that's 1-2-3. Nothing O2 related.....
Let you know the result.

 

I first tried new spark plug cables. And the light cam on again.
After resetting it was idling smooth, and I unplugged the O2 sensor. Result was very uneven idling. Done that for some minutes, but no light came on and also no fault codes were stored.
Later with cold engine I could notice no difference when I unplugged the O2 sensor.
Hmm, nothing to go on with, I guess.
Would there be any other suggestions???

 

The o2 sensor signal is ignored for the first 90 seconds or so after a cold start. The sensor has to heat up to XXX degrees to give an accurate signal - so the ECU ignores it.

Disconnect O2 sensor, erase any codes. Disconnect a battery terminal and hold them together for 10 seconds to hopefully drain any memory from the ECU. Reconnect battery and drive with O2 disconnected - check for poor idle when hot as noted before.

 

Drove until heated up with O2 disconnected and codes erased. After 15 min the Lambda light came on. No sign of uneven running/idling, didn't actually notice the sensor was disconnected.
It gave fault code 2-1-2. As it should, I think.
The 1-2-3 code didn't show yet.
What can you make of this?

 

If the idle changed to worse after disconnecting the o2 sensor as you mentioned earlier - you might have a bad sensor. The output from that old style o2 sensor is easily measured with a voltmeter. Disconnected with engine running at idle the sensor should output between 0 and 1.2 volts. (on the black wire from the sensor, as long as the engine is not running too rich or lean)

Plug the sensor in, warm the engine and measure the voltage at idle, at the connection between the black sensor wire and the green wire. The voltage should be fluctuating usually from .2 to .8 volts. Depending on the sample rate of your voltmeter - the reading will jump around, and using that method that's ok. As long as you see a .2 and a .8. - if you use an oscilloscope and you will see a sine wave, fluctuating from ~.2 to ~.8 volts. That sine wave switches from high to low faster with higher rpms.

If the sensor is stuck on a voltage - disconnect the sensor and connect your voltmeter to the black wire. Then try making the mixture lean - pull a vacuum line off to introduce an unmetered air leak, and make the mixture rich by clamping off the fuel return line to raise the fuel pressure (at the regulator on the intake manifold) or spraying a flammable carb cleaner into where that vacuum line came from. If the voltage does not fluctuate you have a bad O2 sensor. You might want to pull the sensor and visually inspect it. I have found many broken sensors. - Careful, sometimes/rarely they strip the exhaust manifold during removal. (easiest fix then is a used manifold) Use caution and your favorite penetrant backing in and out slowly turning it slightly further each time to hopefully help prevent that.

Or if the car starts and runs fine with the sensor disconnected - and the check engine/lambda light does not bother you (you could always remove the bulb!) Simply leave the o2 disconnected and drive away happy.

The US cars don't have a lambda light, it's a check engine light here.

 

A bit strange perhaps. As reported the first time I disconnected the O2 sensor the engine was idling worse.
This afternoon however I noticed no difference when disconnecting it. Nor in idling, nor in driving. Only the lambda/check engine light came on and I got code 2-1-2.
This evening also no difference. I measured the voltage between black and green. It was fluctuating between 0 and 0,35 and gave a steady 0,51 when I pulled one vacuum hose. But then I did idle unevenly.

My original problem was that the engine was running unevenly when the lambda light came on (that's the same light as the "check engine" in my car) and the fault code I then read was 1-2-3. You thick that might be solved by leaving the O2 sensor disconnected?

 

Another fun thing to do - warm engine up, and with black o2 wire unplugged connect voltmeter to it and ground. Measure voltage, should be stable. Touch green plug to ground (sending 0 volts to the ECU making it think it is lean) and watch voltage from sensor change. Then lick a finger on each hand and with one finger touch the positive battery terminal and the other hold the spade connector in the green wire (providing a small voltage through your body to the ECU making it think it's rich) - The ECU will enrichen or lean the mixture and there should be a voltage output change from the O2 sensor if working normally.

On a car that old I would remove and inspect the O2 sensor for damage -

As far as leaving it unplugged - The O2 sensor can only change the mixture very little. It's really there to keep the catalytic converter at the exact mixture needed for the converter to work the best and to keep from damaging the converter usually from overheating and melting the ceramic honeycomb. Is there any catalyst material left in your converter? Usually by now it has broken up and blown out the tailpipe.

The ecu does not know if the converter is intact or working. That was solved later when rear O2 sensors were installed beginning in about 1999?

I trust you have looked for any obvious air leaks - such as the intake manifold gasket - they like to break and let extra air in on one cylinder. And If I had a working catalytic converter I would not run without an O2 sensor -

 

Hi. After fiddling with it last week and reconnecting the O2 sensor the Lambda/Check Engine light stayed OFF for 250 km. But now it came ON again. WIth the same code 123.
Hoonk, the tests you suggest focus on the O2 sensor. However it seems that that sensor workts fine. You think it still can be the cause of the 123 (coolant temperature sensor) fault code? Or are there other solutions for this?

 

I am having a similar issue with code 1-2-3 on my 93. After I clear it, the code returns about 500 miles later.

Did you ever find a resolution?