As most of the regulars know on here, my 110TDi has had turbo problems for a long time. In mid-Novermber last year I eventually got around to taking my turbo off and cleaning it. The vanes were pretty much jammed, so I thought that was the end of it - till yesterday, when I gave the car some stick up a hill and limp occured. It then occured 3 more times on the way home last night.

Now, I am hoping I can rule out sticky vanes in such a short time (2 ish months), so therefore that leads me think slightly differently:

- how does the car detect / determine an overboost has occured?

Is that detected by the MAP, in the pipework after the intercooler?

Or is there someother clever detecion magic in play?

If the vanes are ok, and overboost seems to be caused most when giving the car some stick up a hill, the causes could be:

- MAP

- n75 (£40)

- MAF

- leaking vacuum pipework (I replaced the hose from the vac ball, and the control pipe to the turbo 2 months ago)

- faulty vacuum pump? (sometimes my brakes feel much more sensitive than what I would class as "normal")

If I am honest, the car has felt a bit slow for the last month, but I had wondered if it was me getting used to it after I had cleaned it.

Either tomorrow or this weekend, I will check the other vacuum hoses. The direct feed/return from the n75 to the airbox is attached and looks to be fine. I'll also see if I can get vagcom to talk to the car while it is running (currently my laptop & serial card seems to drop the connection when the car starts ), and try and log something from the MAP side of things.

Sorry for dragging this one out folks :'(

My understainding is that "overboost", as the term is used here, is shorthand for "MAP sensor return has exceeded the programmed upper limit for the throttle setting and engine revs". This might, in itself, explain why it occurs when going up hill, since that's when you're most likely to be using high throttle at low(er) revs, and maintaining a fairly constant speed, so that the MAP can build to its actual peak value, rather than being pegged back by increasing engine speed.

How do we achieve a high MAP? I can think of 2 main ways of doing this with a turbo engine:-

1) Over-speeding the turbo wrt the anticipated design speed, so that it pumps more air than than the engine can cope with, causing the MAP to rise.

2) Holding the turbo design speed, and introducing an obstruction between the MAP sensor and the combustion chambers, so that the MAP builds due to a lack of airflow.

Since you say that the car feels slow, I'd incline towards (2), and suggest detaching the intercooler bottom hose and seeing what (if anything) runs out as a first step.

Are you thinking that maybe my intercooler is a bit blocked with oil...... I was planning on giving that a flush when I did my cambelt in a few thousand miles time, as I need to disconnect some of the same hoses / remove a headlight, etc.

The car has limped when I have been accelerating harder (ie. such as when overtaking) rather than normal crusing. 2 of the limps occured when I was overtaking vehicles on a uphill section of road, and was in 3rd gear around 3.5-4k revs (I would guess).

no the map sensor is after the intercooler so a blockage there would register as low boost not high. basically the ecu allows for 2100 mb of pressure (requested 1900) and anything over 2100 (and over 3K rpm) for more than a few seconds will register as over boost and cut to limp mode.

1st check is that the actuator is moving smoothly and that vac is being sent to it (remember the vac pulls the boost up and loss of vac SHOULD drop boost to min).

seen a few faulty map sensors too that would register high mb at idle etc.

in vag com check measuring blocks 10-11 which should give requested boost V actual boost and airflow from the maf.

remember you cannot see a large rise in boost without seeing a corrosponding rise in airflow

should see around 998-1040 mb at idle

Ok, an obstruction in the intercooler won't cause excess MAP then, but the base logic that you can achieve high pressure,low flow by blocking the intake somewhere is still true unfortunately.

From what F16v says, I'd agree that checking the MAP sensor makes sense, before dismantling the intake!

Just trying to sort out a suitable laptop :-) I have a HP laptop with a serial port... just need to source a HDD carrier for it.

In the mean time I have been out and sucked on the vacuum hose for the turbo and it moved, smoothly and did not stick, so that is a good start.... :-)

daft as it sounds are you using supermarket fuel ie tesco,asda,morrisons or are you using premiunm stuff like shell or bp when i had my golf tdi it used to have sticky vanse when i used the supermarket stuff as it has less detergent in it thus creating more soot

Up to a few weeks ago I have been a Shell customer, but now the Shell stations at both home and work are now BP (and added an extra 3p/litre) I have now switched to either Waitrose (2 tanks), Tesco (1 tank). I checked the vanes/ vacuum movement at lunchtime and it seemed to be free enough. I am half way through my 3rd tank of supermarket fuel.

I got another limp when going up a hill on the way home tonight, 3rd gear about 4000rpm, strangely enough I got to the red line in 2nd gear moments beforehand though.

I got the impression that when I was queuing in traffic and then booted it pulled really well.... but then after that it felt a little slower again (almost like a built up a reserve and then it was "used") - I know that must sound really odd.

its prob not holding boost long enough flat out in 2nd to register as overboost. sounding like it could be your n75. do you still have the manifold anti shudder flap by the egr valve and is it still connected properly?

its prob not holding boost long enough flat out in 2nd to register as overboost. sounding like it could be your n75. do you still have the manifold anti shudder flap by the egr valve and is it still connected properly?

From reading the posts above, the anti-shudder flap might be the place to start looking. If this is partially shut it would represent the restriction / blockage which would give the additional pressure and could also reduce the performance.

Anti-shudder flap is still there, and pipework is connected (or certainly should be connected!). As far as I know it is still working as it should, but I'll check the pipework is sound, and there is full movement on it tomorrow.

I might order in a new n75 from the dealer as well.....

Thanks guys... :-)

Overboost is simply the excess boost between what the car is asking for (requested) and what the turbo is delivering (actual) - The MAP sensor (manifold absolute pressure) measures this and correlates it to requested boost, and will invoke limp mode if actual is above requested by too much - as this is dangerous.

Could be N75. If the MAP sensor itself were broke, there is normally a fault code for this. Has the car been fault code read? I would check that pipe to the actuator again as I burned a hole in mine when it came unclipped from the bulkhead clips, and rested against the exhaust manifold.

i heard somewhere that you can bypass the n75? google search? i dont know if bypasing it will cause damage etc?

Truly dodgy MAP sensor will cause the car to cut out at idle (on a N/A petrol, at least). The old 'does it run any worse with it disconnected?' test should rule it out if VAG-COM doesn't pick anything up.

i heard somewhere that you can bypass the n75? google search? i dont know if bypasing it will cause damage etc?

What would you hope to achieve in bypassing the N75?

The turbo boost is a closed loop system. The N75 duty cycle is set by the ECU to achieve the requested boost pressure as measured by the MAP sensor. A faulty MAP sensor can cause overboost. IIRC, this is detected when a high duty cycle and high MAF reading does not result in a MAP reading within the expected range.

What would you hope to achieve in bypassing the N75?

from what i could gather, it would stop the car going into limp mode? :S

Truly dodgy MAP sensor will cause the car to cut out at idle (on a N/A petrol, at least). The old 'does it run any worse with it disconnected?' test should rule it out if VAG-COM doesn't pick anything up.

Just seen my MAP, it is covered in bits of straw! Looks like a bit of a pain to get at - headlight out job me thinks.....

The hoses to the anti-shudder look ok. As the engine was cool, I had a good suck on the vacuum control pipe to the turbo and that moved well (I had my fingers down on the turbo).

Wil try and get some vagcom logs over the weekend and see what the MAP is reporting.

Log RPM, N75 duty cycle, requested pressure and actual pressure at WOT (through 3000rpm). Plot the duty cycle and pressures against RPM.

If the thought of an obstruction within the intake system is a probable cause, what state is your intake manifold and/or EGR in? When the turbo was changed on my car, I cleaned out the intake manifold and it was stuffed full of crud, and you've got about 50% more miles on your car than me! That's assuming you've not cleaned out the system at some point anyway.

P.S. if you do clean out the intake manifold, budget on a gallon of break cleaner.

Just seen my MAP, it is covered in bits of straw! Looks like a bit of a pain to get at - headlight out job me thinks.....

The hoses to the anti-shudder look ok. As the engine was cool, I had a good suck on the vacuum control pipe to the turbo and that moved well (I had my fingers down on the turbo).

Wil try and get some vagcom logs over the weekend and see what the MAP is reporting.

I had to remove the headlight and washer bottle to get enough room to get at the MAP

I had to remove the headlight and washer bottle to get enough room to get at the MAP

I seem to have more cabling down there... guess that might be my windscreen wash fluid level sensor.

Removing the headlight on the offside is a PITA as it appears that part of the locking assembly is broken (the bit you press your finger/thumb against). It beat me when I did my cambelt change a year or so ago, but I did manage to undo it while ago as an experiment, I might have to venture out in the cold and have another try. Still awaiting for my mate to find the hard drive carrier for the HP laptop I have so I can get some logs from the car before I fork out on some new bits.

That jubilee clip does not look Skoda standard fit to me....

fair play, that map sensor looks a right sod to get to :S

seen a few faulty map sensors too that would register high mb at idle etc.

in vag com check measuring blocks 10-11 which should give requested boost V actual boost and airflow from the maf.

remember you cannot see a large rise in boost without seeing a corrosponding rise in airflow

should see around 998-1040 mb at idle

Right then, my bag-o-spanners reported at idle 1122-1132mb

I have a couple of logs from the car this morning and towards the end of the 2nd log it had definitely go into limp mode, enough though I wasn't aware, until I turned into my side road and found there was no power low down in the revs (but plenty up from 3000 rpm). Almost as if the n75 was just doing what ever it felt like!

I shall try and process the logs in a bit, as they are rather long. I was single handed, so just started driving, unfortunately the traffic in Weymouth was bad :-(

The logs are here if anyone wants to have a look

Edited 22 January, 201115 yr by mbames

Right then, my bag-o-spanners reported at idle 1122-1132mb

I have a couple of logs from the car this morning and towards the end of the 2nd log it had definitely go into limp mode, enough though I wasn't aware, until I turned into my side road and found there was no power low down in the revs (but plenty up from 3000 rpm). Almost as if the n75 was just doing what ever it felt like!

I shall try and process the logs in a bit, as they are rather long. I was single handed, so just started driving, unfortunately the traffic in Weymouth was bad :-(

The logs are here if anyone wants to have a look

you need to seperate the different logs really and run them thru the graph on vagcom. starting at 485.84 you appear to have a reading of 1095 mg airflow (which is high) and 2377mb of boost and continuing to 496.78 which is what has set the limp off.

it is reasonable to say it is not the map sensor playing up as you are seeing a corrosponding spike in the airflow drawn in.

have you checked the vac pipe that goes to the airbox as that dumps the vac from the actuators and if blocked will make the actuator work slowly