Seat Alhambra emissions MoT failure

As a follow up, the emission readings were as follows Normal Idle: CO 5.5% HC 760

Fast Idle CO 3.5% HC 330

This also means my earlier statement about high HC was incorrect - it's high on both readings.

If it's that high then the lamda sensor will never switch so it's unlikely you'll know if it's working. Have you checked the injectors aren't leaking?

Is it running on all 4 cylinders and is the idle quality good? If its not running on all 4 nice and smoothly, and these VW(?) engines do tickover nice, then attack the HT system / compressions FIRST.

If the above is ok then its running very rich, obviiously, way over the

I'd like to know how the garage tested the o2 sensor without the car entering closed loop running without a gas torch or artifical air leak!

Do not run the car in this condition any more than you have to as you'll be replacing the o2 sensor and CAT knackered from excess fuel / overheating.

You have a basic sensor problem, probably the CTS or ATS. We assume the engine is capable of properly warming up to operating temp? If not, i'll stop now from laughter.

You dont mention to say which engine this is- i'll assume its the 2.0 VW

Later cars used a MAP sensor which replaced the MAF. If its the former, then check the output from the MAP and the sense line for splits.

I am certain you will find the ECU is fine after testing.

TBH, it doesnt sound like your garage has a clue what its doing with what is a simple fuelling fault.

Tim..

No injectors haven't been checked, although petrol usage is minimal when running on LPG (cold idle only). Any pointers on how to check them? Re your comment on the lamda sensor; I assume you mean that with HC so high there should never be any O2 present in the exhaust. Does this mean the sensor o/p would always be high and I should challenge the claims of the tester who is telling me it has been checked on an oscilloscope and that it is working correctly?

Tim Thanks for response Firstly yes, Engine is VW 2.0L 8v. Not too sure on the ECU make. The engine reaches temperature fine; oil at 100C or so according to trip computer. I know this is working because I had the thermostat replaced a while ago because it had been running cool.

I agree with you about the validity of the O2 sensor check as in suggested in my previous post, although it was a new sensor put in 2 weeks ago.

Idle quality has been variable - sometimes very good, sometimes uneven and occasionally dies at junctions after deceleration (usually when cold). HT system has been checked but not compression.

Excuse my lack of knowledge on the sensor acronyms, I don't know what CTS and ATS stand for (guessing temperature), if so, wouldn't the fault be recorded on the code record? Similarly for the air flow/pressure outputs. (I have checked the fault codes myself with a PC based VAGCom system).

You believe the problem is not with the O2 sensor, what do you think might be the cause of the intermittent low reading recorded on the ECU?

Garage doing the work & MoT is local and has good reputation with colleagues and friends, however I agree they may be out of their depth with this. However my expeience with other local garages (including the main dealers), is that they are even more clueless. Should I be looking for someone more specialist, and if so do you have any general recommendations? At the moment I'm a bit stuck, the car has failed the MoT and therefore cannot be driven insured. BTW, if the CAT is already knackered, would that account for the emission readings?

thanks, Tom.

If you warm the engine up you can check what temperature the ECU thinks the engines at & what it thinks the lamda reading is. The low lamda fault would occur due to the excess fuel, which if the ECU thinks the engines cold is inevitable. You don't get a direct fault for the engine temperature/external air temperature sensor reading cold.

Duncanwood

Thanks for your comments. I've just spoken to the mot garage who have the car at the moment and they say the temperature readouts on their diagnostics were as they should be when they checked the ECU readouts on Weds i.e. the temperature sensors appear to be OK. Having said that I think I would like to confirm that for myself when I get the car back as your (and Tims) assessment of the problem seems very likely to me. Unfortunately that won't be until next week now..

thanks Tom

As you can glean live data from the VAGcom you should be able to see exactly where the fault lies! (As long as you know what you're looking at)

Obviously the ECU must be seeing actual engine temperatire- 90 odd degrees when hot, rising smoothly from cold start. Air temp will vary depending on heatsoak or not, as long as it shows a sensible value based on ambient temp then that will be fine.

You should also see throttle position (as indicated by the TPS) which at idle should be around 0.8 or lower volts. (or under 10% depending on how your VAGCOM displays)

Mass airflow (if MAF fitted) should be similar to the tps reading.

You'll also see o2 switching (if its in closed loop) and the mixture being trimmed back and forth in response to the o2 sensor too.

A variable artificial air leak downstream of the throttle plates can be used to good effect to see if the ECU is capable of switching to closed loop, and to see if the o2 is capable of responding to whats actually in the exhaust.

Hope this helps abit more!

Tim..

How does it run on gas? If you can get it through its test on LPG, then at least you will have a usable car while you get the problem sorted.

Yes

The temp sensor for the temp gauge may not be the same one as the temp sensor for the ECU.

: The car has had a relatively recent service and distributor cap and HT : leads so I think an ignition/misfire type problem is a low likelihood.

Hmm. I think I would have a Good Hard Stare at the ignition system, nonetheless, for two reasons:

2) The ignition system has been worked on about the same time as the problem appeared.

I would strongly suspect a track in the cap, a dud HT lead, a bad connection or a whiskered plug ...

Ian

But not high CO

Correct- on this application there is a CTS for the ECU and a temp sensor for the gauge.

Tim..

If the lambda sensor signal is definitely switching then it must be seeing around lambda = 1. It's possible that you have a leak in the exhaust system - the engine runs rich (hence high CO), then air leaks in to give an overall lambda = 1 mixture at the sensor.

Err yes, but the OP is seeing very high CO and HC's *at the tailpipe*

Tim..

Yes, but even so. The engine is at (say) lambda = 0.9, hence high CO. Oxygen then leaks in to return to lambda = 1 at the sensor, but the CO only gets diluted by (say) 10% so is still relatively very high. It's too late for that oxygen to burn the CO to CO2 (it's too cold) and it is probably too much for the catalyst to mop up as well.

Agreed, it sounds unlikely, but I have heard of similar effects before, and I'm wracking my brains to think how else you could have very high CO and HC ( = rich at engine) with apparently lambda = 1 at the sensor. It's rare for sensors to get that far out of spec and still be switching properly. It would be nice to see the actual scope trace for the lambda sensor.

An obvious question to the OP is: what were the other figures on the emissions test? We know HC and CO, but what about O2 and lambda?

The low lamda fault implies the Lamdas not 1

True, but I very much doubt the garage who tested the o2 sensor and found it switching were telling the truth.

Tim..