Are you testing the Voltages on the engine loom side i take it?

the best way to test your Lambda or O2 sensor is to get a 16pin KKL cable off eBay and download FiatECUScan for free, you can see the signal plotted onto an voltage Vs time graph and you can see it switching from 0.9v to near 0v.

Testing the open circuit voltage on the loom side wont give you much information apart from the power feed to the heating element and sensor. not unless you have a oscilloscope and back probe the O2 sensor whist its plugged in, but you need to know what the signal wire is.

having the O2 unplugged with the engine running wont damage the engine, it will just throw up a check engine light code and may run rich, in that case it would default back to its pre-determined perimeters.

 

Picture of FES whilst Idling

 

Hi!

That 0V check I did just doesn't make any sense
I just wanted to know if the sensor was getting any current... Well, it seems so.
But I will do a full test with MultiECUScan next weekend... (Why didn't I remenbered that earlier?)

Wanted to check the heater resistance too. Is 8.6 ohms ok?

Just finished a 170 km trip with no lambda and the car seems to be more responsive in every way and not consuming more fuel... Will confirm that when refilling.

Many thanks Chris!

 

and I'm getting this error code too: P0130

 

Well,
- graph is a straight line.
- lambda sensor signal - constant 983,4 mV
- lambda sensor integrator - constant 1.000
- lambda sensor status - always "top limit"

I throttled a bit, but values are the same... does this mean lambda sensor is really damaged? Should I check for other problems?

 

Hi!

I'm wondering, does MultiECUScan read voltage directly from sensor or reads voltage given by the ECU?
It could be that the ECU is assuming 983,4 mV thinking that sensor has failed completely, when actualy it could be outputing other values... What do you think?

 

As I said in previous post, unless you have oscilloscope and you know what colour wires is the signal, ground, voltage feed etc... then it seems that a new O2 is needed. Was the error code thrown up before you unplugged it?

The graphed values for the O2 on MulitiECUScan is calculated by ECU.

You could try forcing the engine into a lean condition to see if the O2 sensor changes state by taking a intake vacuum hose off...

But the O2 sensor will read rich in an open loop condition whilst the car is warming up. Make sure the engine is at running temp.

 

thanks for the feedback Chris!

the error fault code has always been there, before lambda unpluged and even after being reset for several times...
I've also been pluging MES with engine properly warmed up.
Strange, last time tried to read the sensor signal on MES with the the lambda disconnected and it always reads 983,4 mV, ocasionally droping to 0V...

I'll be trying different readings this weekend, pluging a voltmeter to the signal wires (grey and black) to confront with the MES readings.

I also wonder what a virtual oscilloscope like this one
Soundcard Scope
in my laptop could read feeding the signal through the mic input...

 

P0130 rings a bell. I'm sure that was the code I kept getting when I had an NTK lambda and not a Bosch unit.

I bought a universal Bosch LS615 kit and fitted that and no more codes!

I read somewhere the NTK has a different ohm rating to the Bosch what is why the codes kept coming back.

 

I checked this one's resistance and was about 8 ohms, so I guess it's a Bosch one and probably the first one it ever had.

 

If you want to fix the problem, am i right thinking that you do? then you can force the Lambda sensor to read a lean or rich condition by injecting propane into the intake after the MAF (to force rich condition) or disconnect an air intake vacuum hose to draw more air into the mixture (Lean condition) and the Lambda should respond to this.

but skipping this, i think its just worth while in just replacing the Lambda sensor and do you MES scans after as a comparison.

HTH

 

I think it's easier to disconnect the vacuum hose, the one fitted in the air intake rubber tube, right?

 

Running rich won't be doing your CAT any good. They are expensive(ish) to replace.

 

Yes, and I suspect this one is ruined already

 

Anyway, I tested a microphone on my laptop with a virtual oscilloscope. It only started cliping with inputs above 1 V, so I think is possible to input the o2 sensor signal through the mic in, and then comprare with values given by the obd2 interface if you don't trust them...

Chris, will try that hose thing by the weekend. Will post news here. Many thanks

 

And could a sensor cleaning do any good?

 

I had issues with my lambdas hitting 0v, managed to contaminate them somehow, would work fine for a while, then flag the engine management light.

Tries cleaning them with a 10min 3k revs session whilst parked, but marginally seemed to bring one back.

I had to replace them in the end with a used set from Autolusso (£40 for 2) and all good.

Tried cleaning the old ones with contact cleaner and a blow torch, just to give it a go, but no luck

 

could anyone confirm that this is the vacuum hose, please?
http://www.alfaowner.com/Forum/atta...tachments/alfa-147-156-and-gt/160955d1359411313-lambda-sensor-checking-hose.png

 

No, that's the crankcase breather ... there's another one behind the throttle body.

The only vacuum I know of is the one to the brake servo

 

Pulling the crankcase breather will still create a vacuum leak and lean condition.

 

Hi everynoe!

Just made an interesting test.
Managed to measure the lambda signal while engine running at proper temperature with an external voltmeter.

At idle I allways get very low values, say less than 0,040 V. Accelerating, lambda reacts instantly and outputs more than 0,800V. Returning to idle output drops to less than 0,040V again...

So, lambda is not dead, but not working properly. Could it be other fault causing this?

 

You really need to get a 16-pin KKL VagCom USB cable, download MES and view the waveform on the plotted graph in MES, cable costs peanuts off eBay, and MES is free, the waveform should look same as picture i posted (at idle), and frequency should increase with revvs. but don't assume its working correctly because you're getting some voltage, get VagCom cable & MES 1st then see how the waveform is.

HTH

 

Hi Chris! I have already done that! The obtained graph was a straight line at a fixed value of 0,9834 V as posted here
http://www.alfaowner.com/Forum/alfa-147-156-and-gt/365717-lambda-sensor-checking.html#post5164079

I guess the ECU is ignoring the lambda sensor and assuming rich mix all the time because the voltage output is not correct...

I made the voltmeter test just to confront with values given by MES and to make sure that the lambda sensor was really dead or not...

 

Just checked if my 146 lambda probe was the same model as my 156, but not lucky :\

 

Can you upload a screenshot of the graphed waveform of the Lambda sensor at idle and at high revvs?

 

Remember that a Digital Volt meter cannot sample fast enough with switching voltages, that is why you cannot rely on voltmeter readings for this type of application. just bear in mind.