Lambda light - on my Sport that had a misfire for years

[Laird Scooby]

Originally Posted by rogerthechorister

Well now hot start is fine after changing Crank Position Sensor. Safi the garagiste made a bit of a meal of it, though, at my expense!

Now the puzzles. Safi did an emission test, and it is good but it says it is for a non-cat car. How can that be? She is 1995/6, and must have and indeed does have a cat. Second how different are the cat and non-cat standards for that year? Third, I am SURE but have not yet checked that I have seen lambda values on tests of some of my volvos, but this test does not state a lambda valu.

Am I imagining things?

What was the CO% at idle? Wouldn't worry about the Lambda value, that relates to the machines percieved operation of the mixture control and for an MoT pass should be between 0.97 and 1.03 and is a measure of stoichiometric operation and how close it is to that - 1.00 is stoic, less is richer and more is leaner.

It's entirely possible he set the machine up on manual mode and some machines won't operate in cat-test mode when in manual as the cat-test follows a set procedure to ensure the cat is fully warmed up etc.

[rogerthechorister]

Garage has fitted the new crank position sensor (and totally overcharged me but it is close enough to walk home which saves some money) and we have hot start and the lambda light is staying off.

But they did the wrong emissions test thinking for some unimaginable reason that she was 1994 and catless whereas she is 1996 and must and does have a cat - so she is back there for them to do the RIGHT emissions test (at their own expense). Not only are the emission levels different (I think) but also am I not right that a test of a 1996 car would shoe the actual lambda?

[rogerthechorister]

Yes I gather the emissions had to be done on manual as otherwise the test device would have generated an MoT report so if there was a bad emissions fail it COULD have been a "stop now" report.

CO at idle was 1.37. In April for her MoT it was 0.99. I am however a bit worried about the HC. This test it was 228 but in April it was 17 ppm (out of the 1200 allowed).

[Laird Scooby]

Originally Posted by rogerthechorister

Yes I gather the emissions had to be done on manual as otherwise the test device would have generated an MoT report so if there was a bad emissions fail it COULD have been a "stop now" report.

CO at idle was 1.37. In April for her MoT it was 0.99. I am however a bit worried about the HC. This test it was 228 but in April it was 17 ppm (out of the 1200 allowed).

That CO level seems a tad high, should be 0.3% or less unless they've relaxed the MoT regs for your year car - i did hear something about 1994-96 cars now had a wider pass limit than before.

Just looked it up and it should be under 0.3% as i thought, Pg4 of this pdf confirms it :

https://assets.publishing.service.go...th-edition.pdf

There's a lot more info in there and you can even look up your specific model in the pages after that, might take some finding though!

On the bottom of Pg4 there's a footnote explaining Lambda Value in a slightly different way but meaning the same as i said.

[rogerthechorister]

Fresh test done without charge. The correct test.

Lambda perfect at 0.14 first test and 0.997 second test.

HC perfect. 200ppm is allowed and I have 26 first test and 19 second.

CO disaster. 0.3 is allowed at fast idle and 0.5 at natural idle, I have 1.12, 1.00 and 0.93.

So, too rich.

Garage says most likely air filter, then spark plug then air leak then exhaust leak then AMM.

California smog advice adds o2 sensor (sounds a fair bit like the way a lambda sensor works) a MAP sensor which I have never heard of, TPS which I think is a simple switch on a 1996 940, and ECT - unlikely in that temp gauge is sitting bang in the middle although I think she has two, one for the gauge and one for the ECU.

So I will order air filter, dizzy cap, rotor arm, HT leads and plugs (which I had been planning to do anyway). I have an unused AMM in stock. Some of those bits I can do myself mostly for fun - and I could do an Italian tuneup too.

[Laird Scooby]

Originally Posted by rogerthechorister

Fresh test done without charge. The correct test.

Lambda perfect at 0.14 first test and 0.997 second test.

HC perfect. 200ppm is allowed and I have 26 first test and 19 second.

CO disaster. 0.3 is allowed at fast idle and 0.5 at natural idle, I have 1.12, 1.00 and 0.93.

So, too rich.

Garage says most likely air filter, then spark plug then air leak then exhaust leak then AMM.

California smog advice adds o2 sensor (sounds a fair bit like the way a lambda sensor works) a MAP sensor which I have never heard of, TPS which I think is a simple switch on a 1996 940, and ECT - unlikely in that temp gauge is sitting bang in the middle although I think she has two, one for the gauge and one for the ECU.

So I will order air filter, dizzy cap, rotor arm, HT leads and plugs (which I had been planning to do anyway). I have an unused AMM in stock. Some of those bits I can do myself mostly for fun - and I could do an Italian tuneup too.

Before shelling out oddles of £££££ on things that may or may not be necessary, do you have a multimeter? If so, find the Lambda sensor and the connector for it on the bulkhead.
If it's a 3-wire sensor, there will be one coloured wire and the other two will be black, grey or white, coloured wire is normally green.
Get the engine warmed up and with the ignition off, insert the +ve/red probe into the back of the plug in the coloured (green?) terminal and connect the -ve/black probe to a good earth point. Initially set the meter to 20VDC range.
Start the car and observe the voltage produced. If it's battery voltage or exactly zero, i've remembered it wrong and you'll need to consult a wiring diagram for the Lambda sensor output wire - the other two are the heater for the sensor.
Assuming it is reading something, it's likely to be in the region of 0.5V, in which case change the range to the 2VDC range. You should see it fluctuating around the 0.45V point going up and down rhythmically from there. If it's lower than 0.45V, particularly 0.3V or less as the mid-point, your Lambda sensor needs renewing.

The lower the output voltage on a (working) Lambda sensor, the leaner the mixture. The ECU will correct a lean mix by increasing t inj or injection time so the voltage will rise above the 0.45V mid point (which corresponds to 14.7:1 air:fuel ratio) and the ECU will then read this higher voltage (usually 0.7-0.9V) and correct this by decreasing t inj producing a constant rich-lean-rich-lean-rich-lean cycle to average things out where they should be.

If the output of your Lambda sensor is below 0.45V for most of the time or it's slow to rise, the ECU will try and correct this "lean" mix by making it richer. When the sensing element is getting old/fails it no longer produces voltage under rich conditions so will run rich under a failed/failing Lambda sensor condition.

If you find the Lambda sensor is faulty, when you renew it, remove fuse #1 before you start and replace after you've renewed the sensor. This will erase the changed fuel trims in the ECU and reset to factory settings, where the ECU can relearn things much quicker.

From the figures you've provided, i'd suggest the Lambda sensor would be the first place to start looking. There is no MAP sensor (Manifold Absolute Pressure) on the Bosch LH-Jetronic system, instead they use a MAF (aka AMM in USA speak) to determine airflow into the engine. Systems that use a MAP sensor work out the airflow into the engine by engine speed and manifold pressure, generally increasing the fueling with higher manifold pressure.

[rogerthechorister]

That is useful info tx, but the lambda as tested is spot on (as I said) and the lambda light is off. I am therefore not suspecting the lambda sensor as such. Unless you can suggest why I should.

[Laird Scooby]

Originally Posted by rogerthechorister

That is useful info tx, but the lambda as tested is spot on (as I said) and the lambda light is off. I am therefore not suspecting the lambda sensor as such. Unless you can suggest why I should.

The CO is far too high, indicating either a failed/failing Lambda sensor, the Lambda light is nothing more than an engine management light and won't come on until the Lambda sensor has completely failed.

The two first things to look at would be Lambda sensor output and cat condition. It may be possible to revive the cat with some Cataclean and an Italian tune-up but first make sure the Lambda sensor is actually giving the output it should or the ECU will be running the engine rich no matter how many air filters and sets of plugs etc you throw at it.

If the Lambda sensor is fine when tested as i described (given that it's running rich, it should be showing in excess of 0.5V - however i suspect it will be shoing less than that!), check the resistance of the CTS (it is separate from the gauge sensor, the CTS is between #2 and #3 inlet tracts on the head, just below the inlet manifold, the temp gauge sensor is nearer the 'stat housing) using the resistance range on your multimeter.

Each pin to a good earth point should give ~2500 Ohms +/- ~250 Ohms at 20C, on a hot engine it is significantly lower, can't remember the figure exactly but they're on the forum somewhere and i think are about 300 Ohms when up to temperature.
Also check pin to pin which should be the sum of the two pins to earth, about 5000 Ohms at 20C and 600 ish when hot.

[360beast]

327ohms +/- 30 ohms at 80°

[Laird Scooby]

Originally Posted by 360beast

327ohms +/- 30 ohms at 80°

Thanks Luke, it's not something i carry in my head but the site that hosted the 7/9xx Pocket Data Book has disappeared. Was fairly sure it was ~300 Ohms at 80C but couldn't quite remember.