A4 1.8 Stumbles Under Moderate Acceleration Before Turbo Kicks In

[Kubelwagen]

I've owned my 2003 A4 for about 3 years now and it's always been a little lurchy while accelerating until the turbo comes in. Once it's spun up there isn't any noticeable hesitation. Sometimes it kicks in like a switch instead of smooth like it should. No doubt because of whatever is causing the hesitation or lurchy throttle behavior.

Through typical maintenance over the years I've replaced the spark plugs (NGK Iridiums, as I recall), coil packs, turbo, diverter valve, timing belt and water pump, fuel pump, replaced the upstream O2 sensor and recently cleaned the MAF with MAF cleaner and did a throttle adaptation, just for kicks. Replaced a few deteriorated check valves as well. None of the pertinent maintenance has had any noticeable affect on the lurchy acceleration.

I've monitored the MAF and IAT and they appear to be working properly. I've had no CEL's ever except for a recent catalyst issue. I had considered replacing the MAF with a new one just to see if that helped but about soiled myself at the prices of new Bosch units. Obviously there are generic MAF's available for dirt cheap but I haven't heard good things about using generic parts for something as critical as an MAF. I have a 2000 Passat with the 1.8 that runs great so maybe I should "borrow" that MAF, if they're the same part number, and try it on the A4.

I have VCDS Lite and was wondering it there is something specific I should be monitoring? Or if there's information somewhere that would be specific to monitoring performance and what values should be expected? I use Torque for monitoring standard sensors as well. Thanks in advance for any direction.

[old guy]

Possibly a sticking N75 boost control valve. You could log the N75 duty cycle along with the actual and requested boost. If it's operating properly you will see the duty cycle increase and the actual boost track relatively close to the requested boost. When the N75 sticks you will see the duty cycle go up, the requested boost go up but the actual boost will lag behind and then spike as the N75 finally opens up.

[Kubelwagen]

Excellent idea. I'll do some testing. With the high mileage on the car, it would probably be wise to just replace it.

[gmudan]

Excellent idea. I'll do some testing. With the high mileage on the car, it would probably be wise to just replace it.

Dont just throw money/parts at it, thats how people get fed up when it doesn't get fixed. Do some testing first like OG mentioned and it'll be better for your wallet if you pinpoint the problem.

[Goose The Moose]

Possibly a sticking N75 boost control valve. You could log the N75 duty cycle along with the actual and requested boost. If it's operating properly you will see the duty cycle increase and the actual boost track relatively close to the requested boost. When the N75 sticks you will see the duty cycle go up, the requested boost go up but the actual boost will lag behind and then spike as the N75 finally opens up.

I may have possibly pinned down my intermittent misfire to a faulty n75, seeing that I only get misfires around 2600-2800rpms (which is when the stock turbo kicks in, I believe).
Oh Great One, can you clarify exactly which blocks to log in VCDS lite since there are disagreements on the forum on just such matters?

[KHARDa4]

Possibly a sticking N75 boost control valve. You could log the N75 duty cycle along with the actual and requested boost. If it's operating properly you will see the duty cycle increase and the actual boost track relatively close to the requested boost. When the N75 sticks you will see the duty cycle go up, the requested boost go up but the actual boost will lag behind and then spike as the N75 finally opens up.

While looking at n75 duty cycle what will show a bad valve? what data am I looking at to make sure the n75 is functioning properly.

[old guy]

I may have possibly pinned down my intermittent misfire to a faulty n75, seeing that I only get misfires around 2600-2800rpms (which is when the stock turbo kicks in, I believe).
Oh Great One, can you clarify exactly which blocks to log in VCDS lite since there are disagreements on the forum on just such matters?

Block 115 will give you actual / requested. Blocks 118 or 119 will give you the duty cycle: Clicky click

[Kubelwagen]

I connected my VCDS-Lite and this is what I get for blocks 115, 118 and 119.



I just took a picture of the screen because that was easier at the moment. Some of the columns are different than what was listed in the link provided and even then Lite doesn't have complete labels. Many just read "Load" or "Duty Cycle." Columns 3 and 4 in the link for block 115 show "specified" and "actual" boost. Both on Lite read "Absolute Pressure." One could assume they're the same as what's listed in the link. But then where the link shows block 119 columns 2 and 3 as "adaptation boost pressure" and "duty cycle boost pressure control valve," Lite shows "Lambda" and "Load."

I went through every active block and couldn't find any with more descriptive labels. As a matter of fact, blocks 62 and 63 have 5 columns all labeled "Duty Cycle" and there's a different value in every one.

Couldn't find anything online or at Ross Tech that was more helpful, either. If anyone has info on what the measuring blocks are actually showing in Lite, please share.

I'll use VC-Scope and log what's in block 115 and display it later. If someone can relate what the measuring block and column is for the control valve duty cycle, I'll graph that as well. Thanks.

[old guy]

Block 115-3 and 115-4 are showing you the specified and actual boost pressures. Since the car is at idle in your screen shot you are looking at readings of less than 1 atmosphere (1 ATM = 1013 millibar). You will need to log this block at 100% engine load. Take a look at block 114-4 and see if that will give you the N75 duty cycle. If it does then log blocks 114 and 115.

[Kubelwagen]

Here's what 114 looks like-



Here are all the active blocks I could find labeled "Duty Cycle": 60-1, 60-2, 62-1, 62-2, 62-3, 62-4, 63-1. Here's an example-



Once the correct duty cycle block is identified, am I to understand that you need the graph to show 100% engine load? As in WOT? I can certainly do that. :)

Here's one of several VC-Scope graphs I captured. 115-2 appears to be engine load. It's green on the graph. Red is RPM, yellow should be requested boost, blue should be actual boost.

Full size image link-
[img=http://s16.postimg.org/c2rm5wp2d/333a.jpg]



By the way, thank you for your continued help. It's greatly appreciated.

[Kubelwagen]

Dont just throw money/parts at it, thats how people get fed up when it doesn't get fixed. Do some testing first like OG mentioned and it'll be better for your wallet if you pinpoint the problem.

This is true in most cases. Since I do all my own work, except for clutch replacements, simply because I don't like manhandling the transmission, taking an educated guess and throwing a reasonably priced part or two at the car doesn't bother me a bit. I don't get fed up because I understand that the car is just a machine and like all machines, you just have to figure out how they work, what's wrong when they become ill and what they'll need to operate properly again. This A4 is my baby so I certainly don't mind installing a few new parts, especially when they're replacing parts that have 180k miles on them.

Sometimes testing can only conclusively diagnose a bad part if the part has actually failed. One that's intermittent may still cause problems, but not necessarily show as defective during testing. So once the N75 is replaced, it will either fix the problem or indicate that the problem is something else. At least I'll know for sure.

[old guy]

Although it is labeled load instead of duty cycle I'm pretty sure block 114-4 is monitoring your N75 duty cycle. Engine load ranges fro 10% to 150%. It doesn't have to be at the maximum reading. As long as you are trying to boost the N75 duty cycle will activate.

What I was looking for would be an active duty cycle calling for boost while comparing the actual versus requested boost. If the N75 was sticking when you were graphing you would see the N75 cycle go up, the requested boost go up but a noticeable lag before in the actual boost. Once the N75 opened up the actual boost would then rapidly climb. Really you don't even have to look at the duty cycle. Just look for a significant lag between the requested/actual boost numbers.

Another possibility is a sticky DV. This typically won't happen with a diaphragm DV but it can happen with a piston type DV. When you are steady state driving the DV is usually partially open. As soon as you start to accelerate the DV closes and you start to build boost pressure. If the DV sticks there will be a lag between the requested and actual boost since the stuck DV will dump the building boost pressure back into the TIP. Once the boost pressure on the DV control line gets high enough the DV closes and presto! you have boost making it come on like a "switch" that you described. A cracked or split DV control line from the manifold to the N248 or from the N248 to the DV could cause the same symptoms.

[Kubelwagen]

I had suspected the Diverter Valve could also be a possible cause of the choppy acceleration and replaced it when I replaced the turbo a few months ago. I don't remember there being any issues with the lines to and from the DV but obviously something is causing the issue so it won't hurt to do a thorough inspection of the lines. Whatever the issue is, I'll find it.

I'll see if 114-4 behaves like we'd expect and make a new graph. Now that the laptop is in the car with cables and inverter to power it and such, I'm going to check several other things as well such as MAF flow, fuel pressure, and whatever else seems logical.

What doesn't seem logical is why the VCDS-Lite program doesn't have more descriptive labels for the measuring blocks. Maybe it's a marketing attempt by Ross Tech to get people to upgrade to the full package but someplace there has to be a legend of what the blocks are actually measuring. I'll keep looking into that myself as well but again, if anyone knows or has a source that lists these labels, please share. I'm sure I'm not the only one who would benefit from the information.

Actually, I'm going to start a new thread for that specifically and see what happens.

[Kubelwagen]

It occurred to me this morning that the issue could actually be something in the drive-by-wire system as well. Glitchy TPS or other related parts. Some of those sensors are labeled in VCDS-Lite but once I know what measuring blocks are actually what, I'll graph everything relevant and share what I find.

[Charles.waite]

Subbed. I have a similar issue that I've been dealing with.

I'm curious, if you turn off ESP, does the issue go away? Mine seems to. Also when I clear codes and "reset" the ECU is also goes away for a day or two. I'll admit I haven't done much diagnosis on this, but its a minor issue thats been bugging me for a while.

I'm APR Stage 1+. And this issue has persisted for a few years now through a new Fuel pump, fuel filter, new injectors (TT386 injectors when I went to 1+ from stage 1), New clutch, new PCV and check valves, and quite a bit of other things. Turning off ESP seemed like a really strange thing to affect it but who knows. I though maybe I have a glitchy wheel speed sensor somewhere but I haven't had the time or inclination to bother diagnosing it.

[slark]

In my experience a bad n75 will cause surging acceleration under boost. Seems more like a boost leak somewhere.

[Kubelwagen]

In my experience a bad n75 will cause surging acceleration under boost. Seems more like a boost leak somewhere.

My surging or lurching issue is most noticeable before boost ramps up at around 2,500 to 2,800 rpm. So while the N75 is still certainly a possibility, I had also thought that something in the throttle control system could be intermittent. Today I monitored the TPS with Torque and noticed that sometimes the TPS voltage would drop after reaching a certain point, even when the throttle was held steady. Not enough to cause lurching, though. Other times it stayed steady. Maybe the system does that purposely depending on conditions. But the ESP suggestion could be valid until it's proven it isn't. Something is causing the throttle to drop out and it very well could be electrical or sensor related and not boost related. I'll keep at it until it's figured out.

[Goose The Moose]

Throttle basically controls how much air (re-circulated and fresh) is going into the system, this is measured in g/cm3 (i.e.: volume) in block 003. Throttle is electronically controlled, so if there is an issue with it, it would be evidenced through a change in volume under load. Graph volume vs load(rpm), it should increase as the load increases.
I would also do a Throttle Body adjustment, just as a preventative.

[Kubelwagen]

Throttle basically controls how much air (re-circulated and fresh) is going into the system, this is measured in g/cm3 (i.e.: volume) in block 003. Throttle is electronically controlled, so if there is an issue with it, it would be evidenced through a change in volume under load. Graph volume vs load(rpm), it should increase as the load increases.
I would also do a Throttle Body adjustment, just as a preventative.

Another good idea. I had already performed a throttle adaptation just in case there might have been an issue there. It was in spec and there was no change in throttle response.

Since throttle position not only controls air volume but injector duty cycle, it seemed logical that if the TPS was having issues, it could be dropping the duty cycle of the injectors and producing the same lurching issue. I'm not convinced that's really the problem but it's prudent to investigate everything that could be a culprit. That's why I'd like to find out what the generic measuring block labels really are in VCDS-Lite so I can diagnose more areas.

I'll VC-Scope some new graphs shortly and put them up. The new N75, along with a new cat, should be showing up this week. I'll get them installed and we'll see if there are any changes for the better.

[Kubelwagen]

Just thought I'd update that the VCDS-Lite label issue has to do with the program not finding the correct label file for the engine control module. It's generating a 'redirect fail' instead of loading the correct file number. I knew about the label files but had forgotten how they work. It was obvious even in the pictures I posted earlier of the measuring block screens. In the upper left you can see "Redir Fail!" There should be a file number there corresponding to the engine controller number which you can see in this scan:

Address 01: Engine Labels: Redir Fail!
Controller: 8E0 909 518 AF
Component: 1.8L R4/5VT G 0002
Coding: 0016501
Shop #: WSC 63351
VCID: 749B93C957A3

The VWVortex forum has a Vag-Com sub-forum for VW/Audi and I've posted the issue there. There's probably a missing file somewhere I'll have to download into the labels directory of VCDS-Lite.

My wife said she read somewhere that we have to interact with our grandchildren at least once a week to help stave off the onset of Alzheimer's. Apparently it's the same thing with using Vag-Com.

[Kubelwagen]

Subbed. I have a similar issue that I've been dealing with.

I'm curious, if you turn off ESP, does the issue go away? Mine seems to.

I forgot to respond to this inquiry. I turned off the ESP and the lurchy throttle remained. It's my experience with the ESP that it rolls off the throttle smoothly when activated. My problem seems more like a drop off or mild oscillation. But that doesn't mean your issue couldn't be related to that system.

I should point out that my issue with the lumpy acceleration isn't severe. It's just annoying. It wasn't until I drove the perfectly smooth Passat I gave to my daughter that I remembered how the Audi should be running.

[Charles.waite]

No worries. Mine is more of an annoyance and if people were in the car with me they would probably think I drive like a cabbie or something as its really hard to roll into the throttle smoothly. I've been driving with ESP off a bit recently and it seems to help but I think the issue might still remain, I'm not really certain anymore...

As far as VCDS, maybe try fully uninstalling then reinstalling? Might help.

[Kubelwagen]

The VCDS issue has been resolved. The program loads label files (descriptions for the related sensors) based on the controller module number it scans. There was a file in the Ross Tech label directory for the AMB engine code but the label file was 06B-909-518-AMB. My controller number scans as 8E0-909-518-AF. After renaming the old file to the actual controller number, the label file loaded and the measuring block descriptions are now correct. One down.

The new cat and N75 showed up today so we'll know pretty soon if it makes any difference.

[Kubelwagen]

New N75 and cat have been installed. The cat I purchased was a poor choice and will likely be replaced with one from 034 shortly. That's a different story.

The N75 seems to have cured the lurchy acceleration. The turbo still comes in fairly abruptly sometimes but I suspect that's due more to setup issues between the K04 and possibly a minor adjustment still needed with the N75. But overall, acceleration is much smoother than it was before so I'm calling this a success.

Old Guy gets the kudos for suggesting the N75 might be the problem. I had considered replacing it years ago but forgot about it and it went out of sight out of mind until this thread. Thank you.

[Goose The Moose]

I replaced the n75 with no improvements.
On the misfire counter I'm getting a lot of misfires on cylinder 2, way more than on any other cylinder. However, they still happen on 1, 3 and 4, but that's 1 misfire compared to 30 on cylinder 2. I think misfires on other cylinders happen because they have to pick up the slack from the malfunctioning cylinder 2.

I'm planning on switching places on the ignition coils, if that don't work then re-wiring the connectors that go into the coils [can someone suggest a DIY for that] and down the road replacement of the upper breather hoses and PCV plumbing under the intake manifold... I love my car I love my car I love my car ad infinitum