Rough idle and black smoke on full throttle - overfueling?

[efeballi]

Hello all,
My 2004 TDI has had a rough idle and black smoke on WOT since a couple of months. The smoke is not freight train-severe, but is noticeable nonetheless. I used to attribute the smoke to the long times I spend in stop-and-go traffic, as the low temps do not allow for carbon burnoff at the cat. But recently I did a long drive and the black smoke is still there after ~200 miles of driving.
The rough idle appears to be the biggest problem though, especially when cold the engine sounds weird and sometimes shakes the car (which might well be the DMF). All idle problems go away when I touch the throttle and push the RPM to ~850 (its ~780 by default). During the same long drive, I noticed that there's a noticeable discrepancy between indicated and actual fuel consumption, gauge shows 6.3 l/100 km while I calculated 7.4 l/100 km from the receipt.
I checked VCDS group 013 and all injectors are within .5 of each other.
I'm inclined to think that all these point at the same problem (overfueling) but it might well be a wrong fuel consumption index setting, DMF on its way out and normal black smoke under full throttle.
What are your thoughts?

edit: car's a 2.0 tdi dsg, the high-ish fuel consumption is caused by the bikes mounted on the roof rack and even 7.4 isn't abnormal.

 

[KERMA]

camshaft.

 

[Franko6]

Camshaft... it's possible, but that's not much of an answer...

You don't say the total kilometers on the vehicle. The PD injectors are more likely to overfuel with age. It's something of a problem that the in all PD's; the ECU optimizes much of the programming for air/fuel/ injection timing. The one thing you can control is the torsion value, which there were several who chimed in on a thread that is several years old, "The Question of Torsion Value". You might look at that.

There are things you can control to change things, like the turbo, injectors or ECU, but to optimize the best setting, you have to use the VAG-COM and three blocks in the VCDS: Engine module/ Measurements/ block 04, 13 and 15. The American counterpart sounds like it's like the American version 2.0 TDI BHW engine. The most common position of the block 04, torsion value is -.5. We use that as a starting point. The block 15, third block shows Liter per Hour(lph) fuel usage (FE).

Adjust the cam sprocket in relation to the cam; moving the cam bolt clockwise makes the torsion value more positive. Counter-clockwise, more negative. Adjust to find the lowest FE. That is usually a range of +/- 1.0 in the torsion value, that the FE will show the lowest readings. Usually, moving the torsion value toward the most negative reading will give you the best overall FE; usually between .6lph Then, run the engine up to 2,000 rpm and check the block 15 lph reading. It should be between 1.8 and 2.4 lph.

The lph reading is a virtual sensor, in that it is a group of various reads from the engine sensors to estimate FE. Usually, the best Torsion value reading will show the lowest FE reading. When the best reading is established, the block 13 injector balance will be very close. Apparently, that is what you have.

The other issue with this engine is the cams are known for a potential wear issue, usually about 200,000 km. The hotter the environment, the more quickly the wear becomes apparent. This can be seen by removing the valve cover and inspecting the edges of the exhaust cam lobes for sharp edges. If you find that, you have a worn cam. The arrangement of the valves is E-I-E-I-I-E-I-E. The two ends are exhaust, the two middle are intake.

The only way the engine can improve it's performance, cylinder to cylinder, is increase fueling. That is usually the first sign of a worn cam, puffs of smoke at takeoff.

Other than that, the EGR valve is prone to failure when the plastic gears strip out. The EGR is also very likely to become sooted and stick in a slightly open position.

Good luck with your project. I think your first need is someone who has the VAG-COM software and dongle.

 

[efeballi]

Thank you for the replies.
The car's at 215000 km, all stock. Appears like camshafts may really be worn. I live in a pretty warm environment, we spend at most a week below 0C.
I do have VAGCOM, I checked the torsion value and it's set to -0,5. If i understood correctly, I should play with it to get the lowest idle fuel consumption indication, it's 0,5 lph currently with A/C off. Will check fuel consumption at higher RPM.
VAGCOM shows two faulty glow plugs (not related to my issues, but cold starts were rough in winter) so I have to remove the valve cover to replace them + the glow plug harness. Will check camshaft wear when I'm there.
Will post readings from VAGCOM here.

 

[Franko6]

Which engine do you have? In the US, we do not have any of the 2.0 twin cam engines.

 

[WrEkkED]

Have you checked your in tank lift pump? Will also give those symptoms.

 

[efeballi]

Hello,
It's a BKD engine, one of the first 16V tdi's as far as i know.
Is there any way to diagnose a problem with the tank lift pump, other than removing it?

 

[efeballi]

Readings from VAGCOM:
Idle fuel consumption: fluctuating between 0.4 and 0.6 l/h, 0.5 average
Group 013 inj. readings: -0.18 0.13 0.19 -0.15
Fuel consumption at 2000 rpm: 1.3 l/h
Going by what Franko6 says, that's really low!
However, 1.8-2.4 l/h does seem really high to my eyes as the car uses 6 l/h at 100 kph with all ancillaries (A/C etc.) running.

 

[WrEkkED]

Most people just listen for the pump. You turn the key as if you are cylcing the glow plugs/turning on accessories but not turning over the motor and you would hear the pump prime from the back seat.

Symptoms:
1) hard start - needs more cranks
2) slightly less performance in normal opperation, horrible performance in the higher RPMs
3) black smoke when starting from a stop and a WOT - I assume this is due to the turbo not having enough pressure to spool properly
4) more vibration in the steering wheel, especially at higher loads - as if your engine mounts are starting to wear out
5) slightly worse fuel ecconomy - i gained maybe 50km per tank

Theres a great video on youtube how to DIY for changing the pump. It took me maybe 15 minutes.

 

[efeballi]

Sweet mother of Merckx, I have all of these!
I'm trying to find the part no. of the lift pump. I'll replace it right away.
edit: part no appears to be 1K0 919 050

 

[WrEkkED]

For your car I'm not sure the part number. Let us know if it fixes your problem. Don't be alarmed if you don't hear the new one prime. My new one doesn't make any noise at all which is very interesting.

I don't know about your engine, or how you took care of it, but I have 200k on my PD -BEW and when I replaced the head my cam and valves looked brand new aside from being a little bent LoL

 

[efeballi]

It's a badge engineered Golf Mk5. We got it new and took pretty good care of it in its 17 year lifespan. It was not exactly babied while driving but not overly stressed either. However, I should note that my country is known to be a meat grinder for all mechanical equipment and some car companies are bringing concepts/mules here to test their durability.
I can DIY the pump change with the car sitting in the front yard but I'm reluctant to deal with the valve cover and glow plugs without the help of a shop, so it'll be a while until I can report on the condition of the camshaft.

 

[WrEkkED]

Do you hear a popping sound from the valves when its running? When I pulled my head at 180,000k the cam looked brand new. A good mechanic can probably tell you with a degree of certainty if the cam is bad without opening it up.

 

[efeballi]

I finally got around to swap the fuel pump today, the M14 triple square bit took forever to ship.
Low fuel light was just about to come on, less than 1/4 fuel was left. There wasn't much of a mess and I had covered everywhere with towels. Car's idle sounds MUCH better, especially when cold. So WrekkeD, you were right in that the fuel pump was shot, thank you.
I'm not sure what you mean with the valve popping sound. I looked at some YouTube videos and couldn't find anything conclusive. If you could point me in the right direction, that would be great. I could also record the sound of the car idling, if that helps.

 

[WrEkkED]

If you don't notice it, you don't have it (what my mechanic told me). Asside from youtube videos I've never personally heard valve pop. But on those videos it sounds very pronounced. Almost like the sound of the injectors but much louder. Not sure we are allowed to post links, but if you go to youtube and search "tdi bad cam" you will see examples.

I'm glad that it was as simple as the lift pump! It's crazy how underfueling gives the same symptoms as overfueling as well as boost leaks.

 

[efeballi]

1 week update:
Car starts and idles much better, but some weird&intermittent clicking noise still remains. Fuel economy is also better but I can't see much difference in performance. I believe the clicking noise and the intermittent judder is the DMF, but I'm not sure what exactly to look for. After listening to TDIs for hours on YouTube, I'm pretty sure it's not the camshaft that's doing the clicking noise.
Acceleration below 1800 rpm is still not great and I still get some black smoke at WOT, I believe there's a boost leak and/or a VNT actuation issue. The rich fueling might have caused some fouling in the vanes, will do a VNT actuator test. I'm also designing an adaptor for the intake hose to 3D print, to do a leak test.

 

[WrEkkED]

May be time to bite the bullet and take it to a shop. I had a weird sound once I couldn't locate. Shop found a bolt on my alternator was loose within 30 seconds.

Are you able to take a video or sound clip of the noise?

My only other suggestion is to check the tandem pump. It's possible it's failing from having to do all the work without the lift pump. If it's not leaking fuel it's fine.

 

[efeballi]

Update: the black smoke was mostly gone, except after long periods in stop-go traffic; which was to be expected. However, my hunch that there was a VNT actuation issue turned out to be true, car threw a P0299 Boost Control Range Not Reached error at ~6000ft elevation and went into limp mode. Had VCDS with me, resetting didn't help, moving to the passenger seat and looking at the required/actual boost graph confirmed the issue. Boost pressure generally undershoots, but at times also overshoots the demanded value. My speculative view regarding how the graphs look is that the actuator is shot, but it could also be sticky vanes. In any case, the turbo's coming off the car, will report my findings. Might be time for a new thread

 

[efeballi]

Welp, I'm back with idle problems.
The turbo problem turned out to be a hose completely worn through, changing it fixed the turbo. I've also done an oil change recently.
Since the oil change, the car idles even worse than before, it almost dies before the ECM throttles up to keep the revs up, creating a vroom-vroom effect. This happens particularly when the engine is lukewarm i.e. left for a few hours but not enough for the oil/coolant to completely cool and drain. First cold start in the morning is usually fine, but if I go for lunch at noon it's almost certain the vroom-vroom idle happens. This does not go away when I touch the throttle to bump the revs up. The car is completely fine to drive, other than very slow maneuvering sometimes stalls the engine, despite being an auto. The idle is also completely fine at the end of the drive.
Some googling revealed that oil contamination in fuel can cause this, dipstick shows no oil leaks but I can't be 100% sure. Will check VAGCOM for injector balance, but I don't think that's an issue. People have commented that the fuel pump may also be the culprit. What do you reckon?

 

[tdi_my live]

Can you do a wot log Blocks 3-8-11

For rough running, have you checked the alternator freewheel?

 

[efeballi]

Update: after changing the filters, I wanted to take care of some loose wiring in the engine bay, by the EGR flap. Turns out that was the fuel temperature sensor, and I got a CEL the morning after, which VCDS revealed to be Fuel Temperature Sensor A - Open or Short to Plus. I logged the temperature when going to work, and the jumps and dropouts in the sensor reading was very obvious. I got a new sensor and plugged it in, and logged the temperature drop for an hour while the engine was cooling down. No jumps or dropouts in an hour. Will drive it around for a bit to see if the problem resurfaces. If it does, something is wrong with the cables between the ECM and the sensor.
Car now idles and drives much better, it used to lurch a little bit in stop&go traffic as I pressed and let go of the throttle. I feel like performance suffered a little bit, but that might also be down to the fact that it's now smoother and less lurchy/jumpy.

Also, with the advise of the above post, I ran the logs and saw nothing untoward. EGR's right on the money, boost generally undershoots and there are symptoms of sticky VNT, smoke limitation never poses a problem.

 

[efeballi]

2 year update: turns out there's a 14 year old TSB about rough idling, which perfectly describes the case I have.
https://vwts.ru/articles/tpi/tpi-2021604_2.html (google translate is your friend)
I've also increased the idle RPM to 870 through VCDS and that glosses over most of the idle problems anyway.
Now the problem is to find someone to do the ECU SW update.
In-tank fuel pump recently died, engine quit and left me stranded I had to have the car towed after rolling to a safe stop. Mechanic ID'ed the problem to clogged LPFP not being able to tolerate fouling. At the time I was unable to find an OEM pump so went with an aftermarket, that's probably the reason. OEM Pierburg pump now in, no problems in ~6 months.
Lurchy behavior turned out to be turbo VNT vanes, now resolved.
Not many other components are there to explain the clicking noise other than the DMF. It is a fairly big job that requires the removal of the DSG so I'm inclined to postpone it till the clutch pack needs replacing.