The Limp Mode that would not die.

I've actually got TWO problems that just won't go away.

They're both long-standing problems, and I've just let them go for several months due to lack of funds and time.

First problem was a bad glow plug. I tested them with an ohmmeter - found #2 was bad, and I replaced that - and I'm still getting bad glow plug codes:

17056 - Cylinder 2 Glow Plug Circuit (Q11) Electrical Fault
P0672 - 35-10 - - - Intermittent

If I clear the code, next time, it comes back as #4 bad, or #1. Different each time. I tested them with the ohmmeter, and they're fine. I tested the harness, per the Glow Plugs 101 article - and I got 12+v from each socket. I had checked my relay in the past, and it was good - I'm going to check it again tomorrow, but I assume it's good.

The other problem is, limp mode.
Drives fine; then up around 3000 rpm in 3rd gear, it varies - the power cuts out. If I switch-off the ignition, and switch it back on, power's back.

I replaced all the vacuum lines.
I swapped the EGR solenoid for the N75, and that seemed to work. So I got a new N75, and I tested the VNT; seems to actuate across the full range, all the way to the stop - and it operates freely.

After I put the N75 in, it seemed to go away for a while. But it came back. Some days, it happens ANY time I demand any power from the car. Other days, it doesn't pop out at all. I feel like if I could just keep the dang glow-plug code clear, I could at least see the CEL come on for the limp mode - but that CEL is on all the time.

But here's the code I get on limp mode:
16618 - Boost Pressure Regulation: Limit Exceeded (Overboost Condition)
P0234 - 35-10 - - - Intermittent

I cleaned out the MAF - the screen was kind of dirty. I performed the Cage mod. Didn't make any difference in performance at all; nor limp mode.

Is there a better way to diagnose if it's the MAF without unplugging? Unplugging the MAF causes limp mode immediately when I start it up. So there is obviously a difference - it's doing something.

But it seems with both this limp mode problem, and the glow plug problem, I've got a couple of gremlins partying it up under my hood. I'd really like to evict them.

Is the intake manifold clean?

Maybe you could log 003 and 011 measuring blocks to see what's up.

You might need a new glow plug harness, or at least a good cleaning at both ends.


Your '03 ALH is stock, right?

(also - when I took my old N75 apart, there was a tear in the diaphragm. The old N75 was definitely broken. So I can't understand why I'm still going into limp mode. . . )

My ALH is stock - but I put in a VR6 clutch and flywheel, when my stock D/M flywheel burnt out.

What is meant by logging 003 and 011 measuring blocks.
I've got the 409.1 VAG COM sw, and an ebay cable. Will that software do this? Or do I need the full version?

Where is the other end of the glow-plug harness?

I don't know - I have a Ross-tech cable and the latest 80x.x release. It's a basic function, so it should be able to at least read one block at a time. Check their website for limitations, or just try it. Their FAQ says anything from 000-025 is accessible.

The glow plug harness has a connector in the protective channel that goes from under the battery to the plenum under the windshield wipers. You should be able to undo the snaps and get the cover opened enough to reach the connector.

Okay - here's my plot for blocks 11 and 03. If I did this right. I was headed up a hill, and right towards the end of 3rd, it threw the overboost code and I went into limp. Blah. But I caught it.:


What does it mean?
If the chart is fubar, I have the raw data. I also have a logfile from a prior drive where I didn't go into limp.

The intake manifold is either restricted, the vane actuator is incorrectly adjusted, or it doesn't have full range of motion. Make sure the actuator can return to the full up position. Use basic settings 011 to help with visually checking the motion range.

Okay - thanks.

The one thing I haven't done yet is clean the intake manifold. . .

I'm pretty sure that the VNT is free; it goes down to the stop, and clamps down on the feeler gauge. Whether it goes all the way back open, I don't really know; but I can feel that the spring, or whatever supplies the return force is very firm. But it slides open and shut without any binding. It's nice and smooth.

Well something is certainly causing it to build up too much pressure, either a severe air blockage or an actuator problem. Can you link to the logs themselves?

You're overboosting signifcantly... (DOH!) You can see your boost far exceeding the 1950mB requested. Your N75, vacuum, lines, sticky vanes or actuator are likely suspects. If you have a Mighty-Vac, hook directly up to the turbo actuator and a) see if it holds vacuum b) start and end of movement, should start moving around 2 and quit around 17 inches of vacuum. If not, a spread of 15 is normal full range. If you're throwing a bunch of fuel at your turbo, it will have a tendancy to overboost but, your system is just either failing to or not able to bring it back down. Boost valve?

Okay; so when the purple line goes up over the yellow line, that's bad?
(I figured. . . but I haven't seen any other of these plots, and my GOOD plot does this too, and didn't limp).

I borrowed my friend's Mighty-Vac, and the actuator starts around 2-3 inches of Hg, and quits around 15-18 or so. It holds that 18" value for. . . 10, 15 minutes, or longer. It's not leaking.

So I'm going to clean my intake, I guess.

Since these boost values are based on the MAF's signal, then can I assume the air-blockage is AFTER the MAF? (ie - intake manifold, rather than snow-screen)

Any blockage would have to be post turbo, but it takes a lot to cause overboost. Make sure in basic settings mode the arm returns to the top when N75 shuts off.

The limp mode trigger is >400mbar deviation from requested.

Can't tell for sure, your picture is too small, but it appears your MAF is toast. Looks to be about 1000+ around 3k RPM. It should be around 800-900.

You said you cleaned it, what version is it? Brand? BOSCH is on rev. C. Try updating to the new sensor.

I cleaned the intake.

OMFG - what a pain. No not taking off the intake or cleaning it, but that damn exhaust cooler, and the flexi-exhaust pipes, I couldn't get the damn things to line up. Took me way more than the advertised 4-5 hours.

I run my jetta on B99; so here's hoping that I never have to do that terrible awful task again!

I don't want to replace my MAF. I'll run another test tomorrow, and see what the numbers look like.

well how much of an opening was there?

it was about 80% clear.

I've seen much worse clogs in pictures on this site.

Okay, sounds like NB TDI was right.



(I screwed up the labels on this graph: should read MAP Spec and MAP Actual. And I switched the colors from my last graph, where purple was my Actual, and yellow was my Spec)

so. . . .
new vacuum lines = still throwing p0234
new N75 = still throwing p0234
Checked out actuator = still throwing p0234
"cleaned" MAF = still throwing p0234
Clean intake = still throwing P0234
(my VNT actuator does move like this: http://pics.tdiclub.com/data/500/2006_0603_175432AB.AVI - but like I said, the spring that returns it seems to be pretty firm)



I guess I'm saving up for a new MAF.

Question: I'm trying to understand, I guess, how the VNT/Turbo works.
More vacuum = more boost?
Or does the turbo boost it's max when the VNT is fully "relaxed" - and boost less when vacuum is applied and the lever pulled down to the stop?

I guess from this animation:
http://www.youtube.com/watch?v=Gc2awh0O0Bc
The vanes are most closed when more vacuum is applied.

I guess I'm trying to wrap my brain around how I should troubleshoot this.
So; my turbo seems to return from high-vac/vanes closed/high-boost, mechanically, just fine. That spring is VERY firm, it provides pretty good resistance to return to the "open" position when I push it down with my finger. Doesn't bind at all.

So going on the theory that I've got a bad MAF (or electrical gremlins?); MAF is telling us there's less airflow than there actually is? (or the signal is intermittent/flaky)? - thus, the ECU is sending bogus data to the N75, keeping my vanes closed longer than they should be?

(I like the "flaky signal" theory - because that works with my "bad glow plug harness" story)

I would think that if the vanes were binding on crap; they would not return to the "open" state in a smooth way; you'd notice that when you move it with your finger, or when you watch the mighty-vac actuate it, and the return line on the graph would be "chunky" (not a smooth curve). You'd be seeing steps or flat spots in the Actual MAP, as boost went back down.

If the VNT leaked, or if there was a vacuum leak anywhere, you'd see an underboost, because the vanes wouldn't be actuated all the way.

If the N75 was not opening as the ECU commands (or do I have the N75's operation backwards?), then we'd have the same problem: not enough vacuum, not enough pull-down on the actuator, and the vanes wouldn't close down, they'd remain open, and you'd have under-boost.

can you for giggles remove or tighten the boost valve all the way and rerun these logs? do you have a boost gage? my guess is that your boost valve is set too high. looks like it is set to cut boost off at 20psi..

you installed the boost valve before you changed the N75 right?

your boost valve is obviously not doing anything for you because you are overboosting and you can see its not clipping it off like it should. (This is a 3rd gear pull BTW)


you can see in this photo that my boost valve is clipping my spike off nicely and keeping hte boost right near requested.. (I have RC3 18psi tune so My car requests higher than yours) I think I need to loosen my boost valve just a touch because its having a hard time actually reaching the requested boost level after it pops open to clip the spike off.

And here is a log of my MAF Actuals vs RPM... (5th gear, 60-100)


you can see Im running 1100mg/r at around 3000 rpm there. Granted, I have no idea if this is a good graph or not, I know my boost log looks good, but the MAF I am not sure of.

here is the boost log of the same run above. you can see its struggling to reach actual. Ill be adjusting my boost valve just a touch soon and relogging.

after rereading the thread, it sounds like you dont have a boost valve yet.. you need one.

will allow you to clip off that overboost and keep it in check.

eh? What's a boost-valve?

no, seriously. . .

Is this a stock part that somehow went missing through all my misadventures under the hood?

www.boostvalve.com

make sure you get the one specific for your TDI. Item #TDI-BV

use checkout code tdiclub to save 10%.

Doesn't my graph on page 1 of this thread show that the MAF is actually working?



I see MAF Actual curving way up above MAF Spec. (up to 1050, for a spec of around 750-800. wish I knew how to attach the whole csv logfile to this posting) -

On the MAF 101 FAQ, it says that a bad MAF will have an Actual that is below Spec.

This post:
http://forums.tdiclub.com/showpost.php?p=2022780&postcount=76
Seems to be saying that it could be a clogged exhaust, (oh noes) or overfuelling - (how does that happen?).

So what does this boost-valve do? does it sit in line between the Actuator and the N75? Why would I need an aftermarket part if all my stock parts are working as they should? (I know - something is NOT working as it should; but I don't know what that is just yet).

One thing I would try is just lengthen the turbo actuator rod a turn or so (lower characteristic response).

It could just be off from the computerized map, causing overboost.

The MAF should be fine - notice how the higher than requested reading corresponds with the overboost condition.

No, no, no... gees guys. His MAF is fine. MAF and MAP readings will follow each other. Low MAF readings even while producing high boost is the warning sign that the MAF is bad. The MAF is not showing any signs of problems. You are overboosting causing limp mode aka WarpFieldCollapse. You can see it by following Specified MAP vs Actual MAP. Specified is asking for 1950mBar (std for stock ECU) yet you're producing around 2500 mBar for the duration of that gear until.... 3rd gear it's had enough and your boost drops to 1300-1400 mBar. Not surprisingly, your MAF numbers drop too mimicking exactly what is going on w/your boost. A boost valve is made to stop spiking only, I wouldn't rely on it to control continuous overboost like your showing. It will solve the symptom only, not the problem. Here's some "normal" logs to compare to. Here I'm adjusting the boost valve to clip the spiking.





In this case, I'm adjusting the boost valve up to allow boost to meet specified. Here's a another 'normal' boost log w/o a boost valve at all...



As you can see, I do exceed specified MAP but only for a split second, not sustained like your log shows. You are overboosting for the entire duration. Ignore your MAF for now and find out why you are not pulling back excessive boost. If your actuator is good as you have verified, it's time to look at sticky vanes, hoses, N75 and vacuum. A normal 'spike' will look like the last log above. In that run, the stock boost control system pulled things back w/o the help of a boost valve. Your system should be able to do the same but it's not.

Thank you, FlyTDIGuy.

You put my mind at ease wrt the MAF.

I suppose if the causes are limited to:
Bad N75,
Vacuum Leaks (hoses),
Sticky Actuator,
Sticky Vanes.

I've got a new N75. (I suppose I should probably re-test the lines with the mighty-vac).
I've replaced all of my vacuum hoses (except the brake boost; if that's the thicker one that the thin-one connects into).
I ran the Basic Settings on my actuator, and it slides up and down like the clockwork that it is.

So. . . if that control lever goes up and down, the vanes may not be following? They're stuck inside the housing?

What are the chances this could be caused by a clogged exhaust?

Lengthen the actuator rod a hair.

1/4 turn at a time? 1/2 turn? full turn?
And - do I need to lock-down that locking nut each time? or just when I find a final setting that works?

One thing at a time... based on this...

borrowed my friend's Mighty-Vac, and the actuator starts around 2-3 inches of Hg, and quits around 15-18 or so. It holds that 18" value for. . . 10, 15 minutes, or longer. It's not leaking.

Click to expand...

I'd say you're fine and don't adjust your actuator rod. The diaphragm starts to move at the right pressure and has the correct spread in travel. If the rod is working freely and the vane lever moves, the vanes are working. Excessive wear and slopiness in the internal vane parts will cause stickiness and overboost but is rare and usually on very worn turbos. I'd be looking real hard at your N75, its plumbing (make sure it's correct) and relief/overflow hose (to the airbox). I would unhook the actuator from the vane lever and check the lever for easy operation and gritty/sticky tendancies also. As for a boost valve...



It's a simple device that works well for controlling spikes. It's good insurance but most prefer to be able to run without it. A well programmed and properly controlled turbo shouldn't need it.

Okay - great, thanks.

Also - I posted a new thread on this but didn't get a response.

After I put everything back together from the Intake Cleaning procedure, I now get an audible whistle as the turbo spools up. I had one before, but it was nowhere near as loud as it is now.

Is it possible that the intake gunk was muffling this sound somewhat, and now that it's all clean in there, I'm now able to hear the turbo like the car would have sounded "new"?

Or (more likely) - did I not bolt something down tightly enough when I put the intake back together, and it's now leaking?

If the latter is the case, I would assume that if the leak is on the intake side, I would have an oily mess, because the pressure would be shooting EGR gunk out. (at least downstream of the EGR) - and if the leak is on the exhaust side (by the EGR intake or outlet - both of which I had one HELL of a time getting lined up and bolted back on, by the way), I wouldn't think that would necessarily whistle, because that's between the header and the turbo - and so that exhaust stream is pulsed. So that's why I'm scratching my head.

Also: there's no real change in performance or engine codes or anything since I cleaned the intake (other than, seems a bit peppier, more go-pedal-response) - and I would assume that a leak would cause an actual performance problem.

So maybe I'm getting the turbo sound I'm supposed to have, and did not prior to the intake cleaning? wish I could get a recording and post it, but I'm not sure how to get a recording without ambient noises (road noise, wind) damping the turbo whistle.