AUA 1.4 misfire cylinder 2 - trim readings

[PlasticMac]

Hi

Sorry about the delay, back at work last week and very busy and I’m a bit of a woos when it comes to working on the car in the rain

So I ran the car up for 5 mins from cold and looked at measuring block 32 and 33. It was not up to temperature by the end of this but the idle revs had settled back to around 700 rpm at the end

So this is measuring block 32

View attachment 111540

It didn’t move through the whole test. I have no idea if that is normal. I’ll give this a try again on an actual run with the engine fully up to temp

This is measuring block 33

View attachment 111541
This was more animated but this is a typical reading after about 5 mins. At the beginning I saw it hit 8% but as the engine temperature rose this is a more typical value. If I raised the revs to 1500 rpm and then let it drop to idle I could see negative values and a fair bit of fluctuation as the mixture adjusted, so would say what I was seeing was what I would expect. I did take a short video clip of this if it’s any use.

The other thing I discovered is that although the VAG401 can take recordings the file it saves makes no sense, lots of Chinese and what looks like Arabic characters in the file but nothing that makes any sense. I need to do some googling on that.

As it’s the first day here when it’s not raining and the sun is actually out I’ll see if I can get out this afternoon and make those resistance measurements on Lambda 1 for you

best wishes

J

Ross-Tech say change the log file type to .csv, then open it. Works for me, but I didn't open it before changing the file type
Mac.

 

[birtyA2]

Ross-Tech say change the log file type to .csv, then open it. Works for me, but I didn't open it before changing the file type
Mac.

Ooh, thanks Mac, that’s a thought, I’ll give that a try or maybe it’s an early excel format

J

 

[tuizner]

thanks J - same here - bit wet last week.

just took a run out and you can see here whats happening. As soon as warm (blue), once the engine speed goes to idle (black) the misfire occurs (red).

I cannot see any other conincidences than the rpm + misfire.

So looking at it in reverse: what is it that revs is doing that stops the misfire?

• more oil pressure
• lower MAP (should maybe log MAP next time)
• better volts from alternator to coil? (how to measure that?) - note i did replace the belt yesterday and i felt it might have helped a little - could be psycho-sematic though...

have checked timing and crankshaft pulley timing amrk aligns with camshaft 2x holes.

next plan is to whip off the cam housing and see if anything tappety has happened or there is any oil line blockages but need to buy a pulley holder doodah first.

 

[Robin_Cox]

I was reading through the AUA fuel injection chapter of the workshop manual earlier (there is a separate pdf for the BBY version) - the image of the lambda plug connection pin numbers appears to be exactly the same as the one above.

One thing that I was wondering is whether the knock sensor (G61) for cylinder 2 or linked connectors could be an issue as this is how the misfire is detected. There is another pair of measurement block pages to look at - 22 (cylinder 1 and 2 are data items 3 and 4) and 23 (cylinder 3 and 4 are data items 3 and 4). These display the degree of ignition advance for each cylinder, and if one cylinder is dramatically different from the others that could be an issue?

There is also one other thing that came to mind as regards the idle versus higher revs question that is mentioned in the context of "additive" rather than "multiplicative" issues relating to lambda function that I didn't fully understand ; the gist of it appears to be that additive issues are encountered at idle and go away when revs increase, and "multiplicative" ones get worse with increasing revs. Something to do with "learned values" in measurement block 32 so the mixture richness at idle or higher in the rev range are wrong (air leaks etc.)

 

[birtyA2]

thanks J - same here - bit wet last week.

just took a run out and you can see here whats happening. As soon as warm (blue), once the engine speed goes to idle (black) the misfire occurs (red).

I cannot see any other conincidences than the rpm + misfire.

So looking at it in reverse: what is it that revs is doing that stops the misfire?

• more oil pressure
• lower MAP (should maybe log MAP next time)
• better volts from alternator to coil? (how to measure that?) - note i did replace the belt yesterday and i felt it might have helped a little - could be psycho-sematic though...

have checked timing and crankshaft pulley timing amrk aligns with camshaft 2x holes.

next plan is to whip off the cam housing and see if anything tappety has happened or there is any oil line blockages but need to buy a pulley holder doodah first.

View attachment 111547

So I got the connector off ok. The Lambda plug is supported by a bracket. I can’t feel anything that holds this on, so does it just pull forward to get it fee of the bracket or am I missing something here? I might be able to get the meter reading from under the front, just waiting for a passing shower, or I hope it’s a passing shower

 

[tuizner]

yes you just pull the plastic connectors out of the metal bracket in the direction of the bracket. No locking spring or anything - its just a slide rail on the plastic of the connector - you can just see the two "rails" below the 704

 

[tuizner]

I was reading through the AUA fuel injection chapter of the workshop manual earlier (there is a separate pdf for the BBY version) - the image of the lambda plug connection pin numbers appears to be exactly the same as the one above.

One thing that I was wondering is whether the knock sensor (G61) for cylinder 2 or linked connectors could be an issue as this is how the misfire is detected. There is another pair of measurement block pages to look at - 22 (cylinder 1 and 2 are data items 3 and 4) and 23 (cylinder 3 and 4 are data items 3 and 4). These display the degree of ignition advance for each cylinder, and if one cylinder is dramatically different from the others that could be an issue?
View attachment 111549

There is also one other thing that came to mind as regards the idle versus higher revs question that is mentioned in the context of "additive" rather than "multiplicative" issues relating to lambda function that I didn't fully understand ; the gist of it appears to be that additive issues are encountered at idle and go away when revs increase, and "multiplicative" ones get worse with increasing revs. Something to do with "learned values" in measurement block 32 so the mixture richness at idle or higher in the rev range are wrong (air leaks etc.)

Thanks for keeping looking. I did unplug the knock sensor (only one for all but it does seem to be broadly in line with the #2) which had zero effect - still misfire.

Where on the manual is this please?

Will take a look at blocks 22 and 23

 

[Robin_Cox]

Thanks for keeping looking. I did unplug the knock sensor (only one for all but it does seem to be broadly in line with the #2) which had zero effect - still misfire.

Where on the manual is this please?

Will take a look at blocks 22 and 23

It's the section on roughly pages 48-50 in my pdf of the AUA workshop manual - after discussing the misfire detection measurement block pages this is then overleaf before getting into a section on lambda function and measurement data items.

 

[birtyA2]

yes you just pull the plastic connectors out of the metal bracket in the direction of the bracket. No locking spring or anything - its just a slide rail on the plastic of the connector - you can just see the two "rails" below the 704


View attachment 111551

Cor a stubbon little chap. So managed to get it off with a bit of squirt. So working with your diagram posted earlier I get about 5.6 ohms between the two pins indicated and nothing elsewhere like you.

 

[tuizner]

I was reading through the AUA fuel injection chapter of the workshop manual earlier (there is a separate pdf for the BBY version) - the image of the lambda plug connection pin numbers appears to be exactly the same as the one above.

One thing that I was wondering is whether the knock sensor (G61) for cylinder 2 or linked connectors could be an issue as this is how the misfire is detected. There is another pair of measurement block pages to look at - 22 (cylinder 1 and 2 are data items 3 and 4) and 23 (cylinder 3 and 4 are data items 3 and 4). These display the degree of ignition advance for each cylinder, and if one cylinder is dramatically different from the others that could be an issue?
View attachment 111549

There is also one other thing that came to mind as regards the idle versus higher revs question that is mentioned in the context of "additive" rather than "multiplicative" issues relating to lambda function that I didn't fully understand ; the gist of it appears to be that additive issues are encountered at idle and go away when revs increase, and "multiplicative" ones get worse with increasing revs. Something to do with "learned values" in measurement block 32 so the mixture richness at idle or higher in the rev range are wrong (air leaks etc.)

found an older log about a week ago where i looked at block 22 - cyl 1 and 2 always at 0 degrees even when misfiring. i think i need to read up a bit more...

 

[birtyA2]

I was reading through the AUA fuel injection chapter of the workshop manual earlier (there is a separate pdf for the BBY version) - the image of the lambda plug connection pin numbers appears to be exactly the same as the one above.

One thing that I was wondering is whether the knock sensor (G61) for cylinder 2 or linked connectors could be an issue as this is how the misfire is detected. There is another pair of measurement block pages to look at - 22 (cylinder 1 and 2 are data items 3 and 4) and 23 (cylinder 3 and 4 are data items 3 and 4). These display the degree of ignition advance for each cylinder, and if one cylinder is dramatically different from the others that could be an issue?
View attachment 111549

There is also one other thing that came to mind as regards the idle versus higher revs question that is mentioned in the context of "additive" rather than "multiplicative" issues relating to lambda function that I didn't fully understand ; the gist of it appears to be that additive issues are encountered at idle and go away when revs increase, and "multiplicative" ones get worse with increasing revs. Something to do with "learned values" in measurement block 32 so the mixture richness at idle or higher in the rev range are wrong (air leaks etc.)

Ah yes this is an interesting point, so should be zero at idle for all cylinders. I wonder if they are?

 

[tuizner]

Cor a stubbon little chap. So managed to get it off with a bit of squirt. So working with your diagram posted earlier I get about 5.6 ohms between the two pins indicated and nothing elsewhere like you.

View attachment 111555

Star man - thanks! - I believe the Lambda now is OK then. Any fluctuation in readings then is effect and not cause.

Now then it can only be back to basics?
spark (but why 2 only?)
compression / valves (more likely in my mind)

I will run another hot compression test tomorrow and check that

Will also try and do a voltage leak test on the alternator and block - difficult with the battery in the back though - cables arent long enough

Valves are a bit more tricky but aim to whip off the cam housing and look for anything obvious

 

[birtyA2]

found an older log about a week ago where i looked at block 22 - cyl 1 and 2 always at 0 degrees even when misfiring. i think i need to read up a bit more...

View attachment 111556

Thinking about Robins additive, multiplicative point. This scan is incrementing and misfire detection is enabled, then the count stays at 72 and the misfire detection disables when you apply some revs.

 

[PlasticMac]

Perhaps a silly question, but do you actually feel the miss fire, (on two), or is just that the error is being logged?
Mac.

 

[tuizner]

Thinking about Robins additive, multiplicative point. This scan is incrementing and misfire detection is enabled, then the count stays at 72 and the misfire detection disables when you apply some revs.

Some useful data at Ross also

Fuel Trim Info - Ross-Tech Wiki

wiki.ross-tech.com

"Add" means additive trim, which is addressing an imbalance at idle. When the ECU is using additive trim, it is telling the injectors to stay open a fixed amount longer or shorter. The malfunction (e.g. vacuum leak) becomes less significant as RPM increase. For additive adaptation values, the injection timing is changed by a fixed amount. This value is not dependent on the basic injection timing.

"Mult" mean multiplicative trim, which is addressing an imbalance at all engine speeds. The malfunction (e.g. clogged injector) becomes more severe at increased RPM. For multiplicative adaptation values, there is a percentage change in injection timing. This change is dependent on the basic injection timing.

 

[PlasticMac]

Is miss fire detection based on the ECU knowing when the "bang" of combustion should happen, and reported via the knock sensor, (I'm guessing), when it doesn't happen when it should?
If so, what's the signature difference between a miss fire, and a knock? Knocking is when the fuel pre-ignites, and the ignition timing is adjusted to prevent it. So a miss fire is when the "bang" is late?
Mac.

 

[tuizner]

Perhaps a silly question, but do you actually feel the miss fire, (on two), or is just that the error is being logged?
Mac.

You feel it.

You feel the misfires happening at idle after 3-5 minutes from cold. When cold (and I guess running rich) all is sweet. when warm, the engine starts to run rougher. it hunts a bit and rocks on the mounts.

Then when EML comes on (after so any misfires) you can certainly feel it as the cylinder has been shut down.

Does anyone know hat knock sensor voltages should be?

 

[PlasticMac]

Is miss fire detection based on the ECU knowing when the "bang" of combustion should happen, and reported via the knock sensor, (I'm guessing), when it doesn't happen when it should?
If so, what's the signature difference between a miss fire, and a knock? Knocking is when the fuel pre-ignites, and the ignition timing is adjusted to prevent it. So a miss fire is when the "bang" is late?
Mac.

The above is incorrect, misfire detection is down to the the camshaft and crankshaft sensor, not the knock sensor:

By using the cam and crank position sensors, the ECM can see which cylinder is approaching top dead center on the compression stroke.. If the engine speed drops when the #2 cylinder (for example) is supposed to be on the power stroke, then the ECM knows the #2 cylinder is misfiring. (Credit Google).

The crankshaft sensor triggers the spark also. I think the crankshaft sensor is a hall effect device, which counts the teeth on the crankshaft rotor. Is there a diagnostic to check the crankshaft sensor "pulses" to see if they are all the same?
Mac.
Mac.

 

[tuizner]

idle from warmish

grey is a bit of revs - checking MAP - seems to work OK

nothing obvious to me.

o

 

[tuizner]

The above is incorrect, misfire detection is down to the the camshaft and crankshaft sensor, not the knock sensor:

By using the cam and crank position sensors, the ECM can see which cylinder is approaching top dead center on the compression stroke.. If the engine speed drops when the #2 cylinder (for example) is supposed to be on the power stroke, then the ECM knows the #2 cylinder is misfiring. (Credit Google).

The crankshaft sensor triggers the spark also. I think the crankshaft sensor is a hall effect device, which counts the teeth on the crankshaft rotor. Is there a diagnostic to check the crankshaft sensor "pulses" to see if they are all the same?
Mac.
Mac.

where is the crankshaft sensor? i had a look yesterday but all i could find was the speedo sensor