Fuel Rail/System Pressure Low 2005 Audi A8 D3 6.0 W12

[Amiran Tsanava]

Okay.
I'm back with results.

I would do cold start and log ECU2 until fully warm with both evap valves disconnected and/or plugged. Before we saw lambda corrections going crazy after it warmed up - now we hope to see none of that as it's the only source of unmetered fuel that can affect both banks on each ECU.

Cold start, EVAP connector for the ECU2 unplugged and the hose blocked.

LOG-11-01D-02A-032-705-73F

Then do a warm idle log with ECU1 keeping both EVAP valves disconnected and/or plugged.

Hot start. EVAP for the ECU1 unplugged and th ehose blocked.

LOG-01-01D-02A-032-705-73F

 

[nugentp]

OK - this is good and helpful. To make 'apples comparison' easier it would be best if you can please take two more logs:

- ECU1 COLD start with its EVAP disconnected & plugged
- ECU2 WARM start with its EVAP disconnected & plugged

The results so far so good behaviour of ECU1 during the warm start and no crazy lambda corrections.
For ECU2 during cold start there is possibly an observed issue in Bank 4 - a sticking injector or lambda probe problem - need to analyse a bit more but having the two other logs will help a lot.

You can see COLD start on ECU2 showing Bank4 being corrected more heavily than Bank3...



Meanwhile - WARM start on ECU1 shows lambda correction much more sensible ranges, but all banks trending to be rich.

None of the silly +/- 20% corrections seen while EVAPs connected so they may have been contributing to the badness.

 

[nugentp]

After you get the next two logs, you could try a short drive to see if things feel smoother but you don't want those EVAPs blocked long term especially in high temps.

Another thing you can do is swap over lambda sensors from Bank 4 & Bank 2 and see if the excessive richness follows that sensor to the other bank... or if there really is a problem with one or more injectors in Bank 4.

 

[Amiran Tsanava]

Great.
So, we do have a one little step forward, I guess.

When doing ECU1 cold start log, does it mean I have to wait until it fully warms up? Or few minutes can be sufficient?
Doing ECU2 warm start do I have to leave the ECU1 EVAP disconnected and plugged?

 

[nugentp]

I would like to see the cold start logs at least 10-15 min duration so that lambda sensor is awake and temperature is building. Absolutely YES keep the EVAPs disconnected and plugged for these tests.

 

[Amiran Tsanava]

Okay.

First time I did the test, I started with ECU2 EVAP (keeping the ECU1 EVAP untouched) and only after recording the log I proceeded with doing the same disconnect/plug with the CEU1 (keeping the ECU2 disconnected and hoses plugged). So, both EVAPs were disconnected only when testing ECU1 hot start.

This time, I made these tests both EVAPs disconnected and hoses plugged from the cold start.

ECU1 cold start and 20 minutes log:
1-LOG-01-01D-02A-032-705-73F

ECU2 hot log for few minutes proceeding after recording the ECU1 without shutting the engine down:
2-LOG-11-01D-02A-032-705-73F

Didn't touch the accelerator pedal not once.
After finishing with ECU1 the engine had been warmed up to 90°C when I switched to the ECU2.

Can repeat this process again if you want me to record a cold start with ECU2 and hot ECU1 one more time with disconnected and plugged both EVAPs this time.

Didn't swap the O2 sensors (bank 4 & bank 2) as the cable length is different. I have replaced all 8 sensors not long ago while rebuilding this engine. So, they are brand new and I hope they won't be an issue.

 

[nugentp]

Getting interesting now - look how good ECU2 behaves on the warm start...



And the ECU1 cold start looks like...




So with EVAPs out of the equation, when the engine is warm everything looks basically fine with ECU1 possibly leaning towards slightly rich but well within lambda correction range (for cruise / part-load anyway).

Based on the vast improvement in these lambda correction scores it does now seem wise to replace both EVAP valves with new OE parts (or quality equivalent aftermarket), but I don't think thats the whole story here.

On a cold start we saw that the lambda correction needed for ECU2 is much higher than ECU1, so it makes me wonder about temperature compensations - need to look at WD to establish if they share the same coolant temp sensor input. Maybe one ECU is connected to the sensor physical and communicates the temp over CAN Bus. I assume each ECU gets its own air temp sensor from each MAF so that is something else to check... do you see same coolant, oil and air temps on each ECU in measuring blocks ?

 

[nugentp]

air temp on each ECU needs to be pretty much identical +/- 1 or 2 degrees C (my gut feel) so that would be something to check.

 

[Amiran Tsanava]

I'll wait for the engine to cool down and try to do the same test with cold start ECU2 once again to be sure and afterwards I'll try to find those evaps online.
Do I have to stick to the part number 077133459 or maybe I can pick something else as well?
Audi TT, S4, S6, S8 share that part looking quite similar.
Doesn't it do the same functionality for any of those vehicles?

air temp on each ECU needs to be pretty much identical +/- 1 or 2 degrees C (my gut feel) so that would be something to check.

You mean intake air temperature, right?
I'm totally sure it is calculated using MAFs.
I'll double check them next time I'm doing the test today.

 

[nugentp]

That EVAP solenoid part was superseded by 1C0 906 517 A according to PL24. It's silly money from Audi for a genuine OE item.

I see a Bosch equivalent on aftermarket though - 0 280 142 353 for sensible money.

 

[nugentp]

Make sure and look at air temp and coolant temp on each ECU. Both MAF values are wonderfully consistent in terms of g/s numbers but it's a separate signal for air temp into each ECU so an errant sensor or bad connection might skew it off.

 

[Amiran Tsanava]

Here's the log done today one more time for cold start with ECU2 for 20 minutes:
02-LOG-11-01D-02A-032-705-73F

And hot condition with ECU1 for 5 minutes:
01-LOG-01-01D-02A-032-705-73F

Also, I have checked intake temp readings from both MAFs and MAF2 is 38°C while the MAF1 is 37°C so 1 degree difference definitely can be observed.

Conditions for today's test was totally identical as before.
Started engine and recorded data while idling without a single touch of the accelerator pedal for the whole period of testing.

Along with this, I want to show you couple of logs recorded after reconnecting and unplugging EVAPs back and driving about half an hour in heavy traffic from traffic light to traffic light with laptop on my knees:

ECU1
001-LOG-01-01D-02A-032-705-73F

ECU2
002-LOG-11-01D-02A-032-705-73F

I didn't feel any excessive vibrations or something abnormal during this drive.
All seemed to be good.

 

[nugentp]

Well that +/- 1˚C difference in air temp is in the noise - no smoking gun there. I had OpenAI create graphs from the latest logs and you can make your own mind up but these look a more like normal/expected behaviour to me - there are occasional spikes into 10% correction territory... note how on ECU2 one bank was corrected from lean and the other was corrected from rich for first ~3 minutes before they converged.

If I was going to summarise things, I would say to get new EVAP solenoid valves fitted - they look like they were causing the crazy 20-25% corrections and must have been sticking open when operated. There *might* be an underlying issue with one or more injectors in Bank 3 or 4 when cold so keep an eye on those.

You could try some short drives with the EVAPs blocked to see how it drives on the road and take more logs. I would strongly suggest to keep the journeys short (max 30min) and on a relatively low amount of fuel in the tank. This is to minimise vapour build up in the hot climate you have... and you can manually vent any fumes by momentarily opening the fuel cap at the end of the drive. I would advise against long extended journeys or with a full tank of gas while the EVAP solenoids are plugged off.

No doubt the engine light will be on with the EVAPs disconnected but it shouldn't affect driveability for some short tests that hopefully confirm the occasional misfires have gone away.

NOTE - if you do experience misfires on next test runs, log the misfire counters - ideally per cylinder to see if there is a trend... Its a pain in the ass that you can only log one ECU at a time.

 

[Amiran Tsanava]

Okay.
Startig to look up those EVAP solenoid valves online today.

By the way, may be you didn't notice I have edited my previous reply for adding some new logs.

Along with this, I want to show you couple of logs recorded after reconnecting and unplugging EVAPs back and driving about half an hour in heavy traffic from traffic light to traffic light with laptop on my knees:

ECU1
001-LOG-01-01D-02A-032-705-73F

ECU2
002-LOG-11-01D-02A-032-705-73F

I didn't feel any excessive vibrations or something abnormal during this drive.
All seemed to be good.

May be this would tell more about this issue.

 

[nugentp]

Those latest logs are identical to the previous ones - I think you must have got the files or links mixed for sharing !

 

[Amiran Tsanava]

I have just checked those links and they are definitely different.
Try to look them up one more time please

 

[nugentp]

Maybe sharepoint is conspiring against us here - or maybe I inadvertently already processed your latest logs with EVAP connected... I checked and yes thats what happened - I must have grabbed the links after your edited post when I thought I was pulling files with EVAP disconnected. These actually do look OK in broad terms so one has to suggest the EVAP badness is intermittent. Its all rather odd as a single injector fault or O2 sensor fault could NOT explain the odd behaviour caught before. Given the low price of Bosch EVAP solenoids I would replace them both and continue to monitor the situation if further misfire events are detected.

 

[Amiran Tsanava]

Today I'm going to remove EVAPs and try to take a closer look.
Is there any way of testing those valves whether they are functioning correctly or not?
What do they actually do?
Operated by a pair of 12 volt cables they turn on/off allowing or blocking air and vapor movement through it. Isn't that correct?
Can I test those valves on conductivity of air while feeding them 12v? Therefore defining their working state.

Found couple of replacement options in local store:

Option 1 for approximately $27





Option 2 for approximately $35

 

[nugentp]

You could flow test them on a bench but it's hard to replicate if they are working correctly with a duty cycle based control signal and the high temperature in the engine bay.

For the sake of $25-35 each for brand new Bosch parts I would just replace them.

 

[Amiran Tsanava]

You could flow test them on a bench but it's hard to replicate if they are working correctly with a duty cycle based control signal and the high temperature in the engine bay.

For the sake of $25-35 each for brand new Bosch parts I would just replace them.

Totally agree.
I'm going to replace them, but there are a lot of Chinese fake spare parts marked as BOSCH in my country and I'm trying to avoid that.