Audi A4 case study

Audi 2.5 TDI 2004
Mileage 142,000
Vehicle is in generally good all round condition.

Fault detail
Check Engine ( MIL ) light ON, engine lacks power and will not rev above 2500rpm.
The owner of the vehicle reported that the vehicle has been driving fine, but it slowly lost more and more power, with the Check engine light being triggered more and more frequently until eventually the vehicle became undrivable and (quote) “stuck in ‘limp home’ mode all the time.”

Recorded DTCs
P0299 Turbo/Super Charger UnderBoost
P0102 MAF or VAF A Circuit Low Input
P0108 MAP/BARO Circuit High Input

Check Engine Light Triggering sequence
Start Condition – Check Engine light Off, all fault codes cleared.
Start engine and start to drive vehicle – Check Engine light is triggered after a period of driving and appears to be link to engine rpm. The DTC trigger threshold appears to be linked to engine rpm. The Check Engine light is off at low rpm but is set as engine speed goes above 2250 rpm.
If the vehicle is driven carefully and the engine revs are kept below 2250 rpm it is possible to drive the vehicle without triggering the Check Engine light.
Driving Video 1 – 1st and 2nd gear. In 2nd the throttle is fully depressed (WOT(wide open throttle)). Note how slowly the revs build, unable rev above 2500rpm.

Initial observations and analysis
1. The 3 triggered fault codes are all related to the induction air flow and air pressure.
MAF – Manifold Absolute Flow (Induction system air flow)
MAP – Manifold Absolute Pressure (Induction system pressure)
Turbo/Super Charger UnderBoost – Referenced from ‘ECU Turbo Pressure Map Chart ’.

2. Although both DTCs P0102 and P0108 refer the circuit Inputs (Hi and low) it is unlikely to be a circuit fault as the Check Engine light is reliably triggered by the engine rpm reaching the 2250rpm threshold. If the fault codes were wiring related then the fault code would be triggered in a different manner. For example, if the fault was a permanent electrical fault then the DTC would be triggered at the Ign ON, if the wiring fault is intermittent then DTC triggering would be in a much more random way and not engine rpm related.

3. From driver feel, the vehicle feels correct at idle and very low rpm. As rpm is increased the engine lacks power, with no noticeable pick up at the point the turbo should start to work (above 2000rpm). The engine feels ‘strangled’ and the first impression from the driving experience is that there is a problem with  air induction or exhaust gas expulsion.

System Examination
From the recorded DTCs and the driver experience it appears that the air induction and exhaust gas expulsion systems are the likely problem areas so the vehicle investigation will start here. Below is a pictorial image of the induction and exhaust system of the A4 2.5 TDi V6.
Air enters the air intake at the intake scoop at the top of the front grill and is drawn through the air filter and air box and into the turbo intake port via the airflow meter.
The air is then pressurized by the turbo impeller and expelled from the turbo higher in both pressure and temperature (due to pressurization).
The high pressure high temperature air then passes through 2 air to air interaircoolers where the air temperature is reduced. Manifold air pressure is monitored in the second intercooler. The air then passes through the throttle body and into the combustion chamber via the inlet manifold and intake valves.
Audi A4 Induction and exhaust

This vehicle has a direct injection system, the diesel fuel is injected directly into the engine combustion chamber where it mixes with the air to form the combustion charge.
The burnt exhaust gas is expelled from the combustion chamber into the exhaust downpipe via the turbo exhaust impeller. The exhaust gas then passes through the front and rear catalytic converters and finally exist the exhaust system via the rear exhaust silencer and tail pipe.

From the DTCs, Manifold Absolute Flow, Manifold Absolute Pressure and Turbo Charger Underboost it is tempting to just consider the induction system, however it should be noted that there are no sensors in the exhaust system. So a blocked or restricted catalytic converter for example, will reduce the total gas flow through the ‘System’ as much as a blockage or restriction in the induction system.

Brainstorming and ‘easy wins’
As with all problems a simple process of listing all the possible causes is a good starting. With the list generated another good step is to try to find ‘easy wins’. An easy win being a quick and simple test that will allow that item to be eliminated from the list or indeed offer data that means further testing in this area should be considered.

List of possible cause
Blockage or foreign object in the induction system – reduced gas flow through engine
Blockage or restriction of the exhaust – reduced gas flow through engine
Turbo unable to generate boost pressure – due to induction or exhaust system faults, blockage etc
Mechanically damaged or worn Turbo – turbo not capable of generating boost.
Leaking induction system after turbo – low induction charge pressure
Turbo control failure – turbo not being ‘requested’ to generate boost
Base engine problems – valve timing or injection pump timing incorrect

Possible tests
Blockage or foreign object in the induction system. This is fairly simple but access to the whole induction system can be limited. There is value in inspecting the easy access areas to get a ‘feel’ for the condition of the system.
Blockage or restriction of the exhaust. This is harder to do as exhaust parts do not come apart easily. There is value in inspecting the internals of the exhaust but it is not an easy task.
Turbo unable to generate boost pressure. This is tricky to deal with. It is possible to measure actual boost pressure via the OBD live data, but if there is no or low boost it is much harder to understand why this is the case. Will the turbo develop boost if given the right conditions, or is the turbo unable to develop boost even if all the conditions are correct. Confirming that the engine assembly can delivery all conditions to allow a turbo to delivery boost will require base engine mechanical check (valve timing, compressions etc) fuel delivery checks (quantity of fuel, timing of injection, injection spray pattens etc) induction and exhaust system integrity, control system integrity.
Mechanically damaged or worn Turbo. Mechanical parts can not be directly monitored by the electronic diagnostic systems. However the electronics are monitoring ‘the system’, the electronic’s switch on an actuator and monitors the system response. If the mechanical device, in this case the turbo, can not deliver the correct response the electronics may well trigger a fault code based on the system failure, and as part of the diagnostic path the mechanical device will at some point need to be inspected.
Leaking induction system after turbo. There is value in testing the induction system for leaks given that the DTC for low boost has been triggered. The turbo might be developing boost correctly but the boost pressure could be lost due to an air leak in the induction system. A very simple test is to listen for air leaks when the turbo should be working. Small air leaks will not be audible, however a small air leak, although undesirable, will not cause the total lack of boost this vehicle is experiencing. An air leak capable of  wasting all available boost (MAP sensor reading at the outlet port of second intercooler shows zero boost) will create an audible noise.
Turbo control failure. On this vehicle this is very easy to check. The turbo is high in the engine bay just at the back of the engine, the turbo vane control actuator is easy to access and observe.
Base engine problems. Valve timing can be hard to check and is critical. Injection pump timing nearly always require specialist equipment as it is done dynamical. In the case of this vehicle, it is unlikely that these setting are incorrect for two reasons. 1. both have been checked by previous technicians who tried to fix this vehicle before it was passed to us. 2. Valve and pump timing if incorrect would have an effect on engine starting and low rpm running. This vehicle starts perfectly and runs smooth at idle and low rpm, if the injection or valve timing were incorrect it is unlikely that this would be the cause.

Easy wins
From the above list of possible causes there is a lot of work to do, a base engine problem which is not directly monitored by the electronics could cause conditions which stop the turbo from creating boost. The lack of boost will trigger DTCs. The DTC will most probable be linked to a sensor, but in this case the sensor is not at fault but simply giving true feedback of a failed mechanical system.
What is required is to look at the system and find easy tests that will confirm whole areas of the system are working correctly. An example of this is the turbo boost pressure actuator. The total electronic and mechanical control of the turbo on this vehicle can be reduced to a single movement of the actuator rod on the turbo. The Engine Control Unit (ECU) monitors many engine condition inputs (engine speed, airflow, coolant temperature, ambient temperature, induction manifold pressure etc),  the ECU then compares this data to the internal programmed calibration data and produces a desire boost pressure for those conditions. The output of the ECU is a variable electrical signal to a vacuum solenoid which in turn, controls the level of vacuum applied to the diaphragm of the boost control actuator on the turbo. As the vacuum is applied, the diaphragm moves which in turn moves the internal vanes of the turbo, this changes the airflow characteristics of the turbo and regulates the levels of boost pressure created.
It is fair to say that when monitored,  if the boost control actuator rod moves from the engine off de-energised position to any other position when the engine is running then the boost control system has the capability of requesting a change in boost pressure. I.e, the ECU has monitored all inputs, has calculated a desired boost pressure for those conditions, has indicated that desire boost to the vacuum solenoid and the vacuum solenoid has delivered vacuum to the boost actuator diaphragm, the actuator rod then moves in accordance to the vacuum applied. One simple test has confirmed the ECU has all the inputs required to create a boost calculation. The ECU is capable of making a boost calculation and delivering that calculation electronically to the vacuum solenoid via the vehicle harness. That the vacuum solenoid is itself in tact and has a vacuum supply and the pipe connecting the vacuum solenoid to the boost actuator is holding vacuum and is not split. The boost actuator is also functioning and moving the external lever on the turbo. So wiring harness, sensors, vacuum system, processing capacity and mechanical linkages in a simple observation test. A very easy win.
This observation test of course, does not tell us that the boost calculation is correct, or that boost is actually achieved, or indeed that the vacuum solenoid is not mechanically suck in the vacuum applied condition; sensor input could be incorrect or the internals of the turbo faulty, but we do know that ‘the system’ can operate the boost control actuator rod, all be it, potentially at the incorrect times.

East Win Test 1 – Observe Turbo Actuator rod.
Engine off, note turbo actuator rod position.
Start engine and allow to idle.
Actuator rod moves to the upwards position.
End of test. We now know that the system can operate the turbo actuator rod. if maybe operated at the incorrect time, or operating the rod may have no internal effect on the turbo, but the control system can change actuator position.

Consider fault codes, vehicle condition and test results.
Having carried out the above observation test and indeed after each test, whether successfully, unsuccessfully or inconclusive, it is important to go back to the fault data and the vehicle condition and consider the next step. It is very easy to get  into an incorrect thought process or diagnostic path based on gut feeling and not on facts.There is of course, times when gut instinct needs to be followed but only when the gut instinct and the facts align.
From the observation test we know that the boost actuator can go into a boost required condition. We also know that boost is not being measured at the manifold absolute  pressure sensor (MAP sensor) or being felt when driving the vehicle. At the moment we do not know whether boost is being created by the turbo but lost before it reaches the MAP sensor  and engine inlet ports, or whether the turbo is simply not creating boost.
As the vehicle condition shows no boost it is fair to conclude that the vacuum solenoid stuck in the fully vacuum applied condition is not the fault as the vacuum solenoid stuck in this state would give boost all the time and therefore over boost conditions.
Again as the vehicle condition is no boost at any time, it is unlikely, but not impossible, that the route cause is poor sensor input causing the ECU to calculate no boost required. This theory is also backed up by the fact a DTC is triggered; if the ECU had calculated, due to incorrect input data, that no boost was required then the correct output as measured by the MAP sensor would be zero boost pressure. The fact that the DTC is triggered tells us that the ECU calculated data and the actual measured data misalign.

From the results of the observation test we had 4 possible areas requiring further investigation, ECU calculated boost = zero, Vacuum solenoid stuck open, boost pressure created by turbo but lost before MAP sensor and turbo not creating boost. From simple re-examining of the data and vehicle condition we can rule out the first 2 options, ECU calculated boost = zero, Vacuum solenoid stuck open.
There are now 2 system areas which match the vehicle condition and test results, a) boost pressure being created by turbo but lost before MAP sensor and b) turbo not creating boost

Boost pressure created by turbo but lost before MAP sensor.
If boost is created but is being lost then there must be an air leak between the turbo and MAP sensor.

 We know boost pressure is not reaching the MAP sensor as we can read the MAP sensor data via the diagnostic data. We also know the MAP sensor is giving reasonably accurate data readings at with engine off ignition off the MAP reads atmospheric pressure, and with the engine running this drops to slightly below atmospheric pressure.  It is a reasonable assumption to say that either the boost pressure is not being created or it is being lost between the turbo and the MAP sensor.


The required action is to pressure test and or visually inspect the intercoolers and interconnecting hoses and piping. In the case of this vehicle the intercoolers are below the headlamps and require some level of vehicle dismantling. A good test would be to pressure test the induction system from the turbo to the throttle body / inlet manifold intake.

Turbo not creating boost.
The reason that a turbo does not create boost can be multiple, turbo itself faulty, restriction to airflow through the turbo, incorrect engine fuelling, mechanical fault with base engine, exhaust system failing, EGR (exhaust gas recirculation) failing, etc. This list of possible causes can be broken down into 2 section; those requiring visual inspection and the those requiring specialist test equipment. A level of reasoning is now applied as to the order of testing; incorrect engine fuelling will require tests to measure the injection timing and the quantity of fuel delivered, not easy on a diesel fuel injection system. Mechanical base engine faults will require cylinder compression test, valve timing tests etc. Although these test would undoubtedly give quality data about the base engine and the state of the fuelling system we know that the engine starts easily and runs smoothly at idle and up to 1500 – 1700 rpm. If there were faults with these systems serious enough to stop the engine revving above 2500 rpm, then it is likely that the idle and low engine performance would also be effected. So for now fuelling test and base engine tests have been put to one side and visual inspection on turbo, EGR value, and exhaust intake systems are to be carried out.

On the Audi A4 2.5 Tdi V6 engine the turbo is located high in the engine bay at the rear of the ‘V’ (between cylinder 3 and 6). This location is easy to access and therefore the next action to be carried out is the removal and visual inspection of the turbo. This will also give access to the turbo boost pressure outlet hose and piping so the induction inspection and test can also be started.

2.5 TDI Turbo location

Audi A4 Turbo
East Win Test 2 – Turbo Removal for Inspection.
As the turbo was removed it was noted that there was oil residue in the boost pressure outlet hose and piping. It should be noted that it is common and indeed almost expected to see a level of oil residue in the induction system due to crankcase ventilation and small oil lost from the turbo bearing. Excess oil could point to the turbo bearing being worn and expelling oil. If and when the intercooler and induction system is inspected the system should be ‘opened’ at a convenient point to monitor any oil levels in the induction track.

Turbo Inspection
With the turbo removed it was noted that there was excess play in the turbo shaft bearing. With the vacuum actuator link rod removed from the turbo internal vane mechanism it was noticed that the internally vane mechanism was very easy to move. It is difficult to quantify the exact amount of resistance require to move the internal vanes, but as there are significant amounts of moving parts connected to this lever it is fair to expect an amount of effort required to move the lever, on this turbo the effort level was considered to be too low and therefore an indication of internal problems.

A4 turbo tests

A4 turbo tests

With the turbo removed from the engine, it was is also possible to noted that there was damage to the exhaust impeller. With the exhaust ‘snail shell’ casing removed (not recommended) there is mechanical damage to both exhaust turbine impeller and the turbo vanes.

turbo impeller damage

With the damage to the turbo exhaust impeller, the worn bearing and the very loose vane control mechanism it would appear that either the turbo shaft has enough lateral movement to allow the turbo impeller to move outside of it’s correct rotation and come into contract with the turbo vanes, or the vanes mechanism has worn and become sufficiently loose allowing them move into the rotating impeller. The third possibility is that a foreign object has passed through the exhaust manifold and into the turbo exhaust impeller causing damage to internal components. Whatever the route cause the turbo is damaged and needs replacement. If a foreign object is considered to be route cause then further investigation is required to understand where the foreign object has come from and action taken to ensure this does not happen again.
In the case of this vehicle, the excessive wear in the turbo shaft bearing is considered route cause, excess oil in the induction shaft also points towards worn turbo bearings and seals, so the turbo is to be replaced.

Road test after Turbo replacement.
After the turbo had been replaced the vehicle was given a road test and to our great surprise and disappointment the vehicle driving condition was completely unchanged. The engine still would not rev about 2500 rpm, it lack power and there was no turbo boost either felt by the driver or measured on the OBD data. In short the new fully tested and proven turbo had made no difference to the vehicle.
It is of course, easy to think that the diagnostic process has been unsuccessfully, but the diagnostic process is simply unfinished. The turbo was badly damaged and required replacing. At the moment it is still unclear whether the turbo was  ‘root cause’ and the failed turbo has caused further engine problem or whether the turbo has failed due to another element failure. Maybe the vehicle had run low on engine oil at sometime, causing the turbo to be starved of oil and prematurely fail the turbo shaft bearing. Running the engine low on oil could have damaged base engine components so the lack of power maybe due to an engine fault. This however this is unlikely to be the case as it does not fit the original DTCs and low engine rpm running conditions.

Continued Diagnostics.
During the visual inspection of the turbo, excess oil residue was noted in the induction hoses. Loss of turbo boost due to a air leak in the induction piping was on the list of possible root causes, so the investigation will now turn to the induction system.
Audi A4 intercoolers

The intercoolers are located behind the headlamps, there is a steel pipe linking the turbo to the first intercooler. The two intercoolers are linked with a steel pipe, in the case of the Audi A4 this is integrated into the cooling pack assembly. While it is possible and indeed preferable to pressure test the complete system as a whole (blank one end of the induction system and apply and measure a small pressure to the other end) we are investigating excess oil build up as well as possible air leak so we will split the system and investigate.
When the rubber hoses connecting the intercooler lower ports to the steel interconnecting pipe were removed engine oil was also found; it literately ran out of the intercoolers and piping causing a puddle approx 20cm in diameter under each intercooler. For this reason we decided to remove both intercoolers to be cleaned and presure tested.

Intercooler pressure test

Pressure testing the intercooler
With the intercooler removed and cleaned, it can be pressure tested.
When it comes to testing, the most important thing is that the test and test equipment is functional. The equipment does not need to be fancy or impressive, just satisfy the test. In this case we need to understand whether the intercooler will hold a pressure. We do not need to know whether the intercooler will hold full boost pressure, as our fault is no boost pressure, so the test pressure can be much lower 0.3 – 0.5 Bar (5 – 10 psi) is fine for this test, this means the equipment can be a little less robust. If our vehicle fault was ok at low boost but loss of high boost pressure, then we would need to test at a high pressure and need more sophisticated test equipment.
As we have the intercoolers removed from the vehicle, another simple test which is possible is to pressurise the intercooler to a known pressure say 0.4Bar, and then reconnect the MAP sensor. With the ignition on the pressure measured by the MAP sensor can be read via OBD data. The MAP reading be it should be the same as the pressure set on the pressure test equipment – note unless the pressure gauge is a calibrated unit there is likely to be a small difference between gauge reading and MAP sensor reading.

Checking MAP sensor
Checking MAP sensor

When carrying out this type of test, it is likely that the Check Engine and DTC will be triggered, so it is important to be able to clear any DTCs created.
In the case of this vehicle all the intercooler hoses, piping and intercoolers themselves tested OK, but there was noticeable levels of engine oil throughout the induction system. As no fault was found with the induction system between turbo and the throttle body and because of the levels of engine oil in the induction system the decision was taken to remove the alloy casting which links the throttle body to the 2 inlet manifolds.

Audi inlet manifold

Inlet Manifold inspection
With the front section of the inlet manifold removed the second major problem with the vehicle was found (the first fault was the failed turbo), a large build up of engine oil sludge. A sooty lining of the inlet track is to be expected as the exhaust gases do flow through the inlet manifold during the EGR cycle (Exhaust Gas Recycle) but in this case it is clear that the excessive engine oil found in the intercoolers has also been sucked up into the inlet track and inlets ports, but unlike the cold intercooler piping and hoses on the hot alloy inlet manifold it has harden to form a sticky sludge.
Due to the build up of the oil sludge in the front inlet manifold both the right hand and left hand manifolds were removed to allow inspection into the inlet ports themselves.
The oil sludge build up was extreme as can be seen from the images below.

oil build up Audi A4

The oil sludge in the right hand inlet ports is wet and sticky where as the build up in the left hand inlet ports is dryer and baked. In both cases the sludge thickness is at least 3 – 5 mm thick but in some places up to 10 mm in depth. It is likely that the RH bank, being closer to the throttle body, has ‘pulled’ more oil mist into the ports than the LH bank, so the oil sludge in the RH remains wetter.

Audi A4 oil build up

With oil sludge build up of this level in the upper part of the inlet ports, it is likely to be of a similar level around the inlet valves themselves. As the valve opening gap; the gap between valve seat and valve head is only measure in mm’s then any oil sludge build up here will restrict the amount of air charge that can pass into the combustion chamber.
The diameter of the induction ports and the valve opening aperture has been reduced with the effect of restricting the air flow into the engine.

Cleaning the Induction system.
The alloy manifold casings and all 12 inlet ports were chemically cleaned to remove all trace of the oil build up.

Audi inlet port

Road Test
With the induction system cleaned and reassembled the vehicle is ready for a road test.

Driving Video 2 vehicle before investigation and rectifying action – Acceleration in 1st and 2nd gear. In 2nd the throttle is fully depressed. Note how slowly the revs build, unable rev above 2500 rpm.
Driving Video 3 vehicle after port and manifold cleaning – Acceleration from stand still in 1st and 2nd gear

The problems experienced on this vehicle where:
A) Engine lacks power and will not rev about 2500 rpm.
B) Setting fault codes i) P0299 Turbo/Super Charger UnderBoost ii) P0102 MAF or VAF A Circuit Low Input iii) P0108 MAP/BARO Circuit High Input

The root cause was the turbo, which had a worn shaft bearing allowing oil to leak into the induction system.
Over time the oil leaked into the inducting system, when the oil reached the hot engine it is formed a oil sludge. Slowly the oil sludge builds in depth. The reduction of the port diameter and the sludge build up at the back of the inlet valve reduces the air mass allowed into the combustion chamber.
As the oil sludge builds deep in the inlet port, the air charge is further reduced until a noticeable drop in engine power is felt. The more the sludge builds up, the lower the engine power. As the turbo shaft bearing is worn further more oil is leaked and the process accelerates.
At some point in the process the air flow into the engine, as measured by the air flow meter, will fail to match the ‘air flow map’ in the engine ECU. This will then trigger a DTC.
As the inlets ports are further restricted so the triggering of DTCs will become more frequent and more DTCs will be triggered, turbo boost pressure incorrect, volume of air incorrect, pressure in the induction system incorrect.

As the turbo continues to wear, the movement in the turbo shaft will eventually become so great that the impeller comes into contact with the boost control vanes inside the turbo, shattering the impeller vane and damaging the vanes and vane control mechanism. At this point, the turbo which was still creating boost, will fail to create boost and there will be a further reduction in engine power. It is likely that it is at this point that the owner stopped using the vehicle and considered it undrivable and stuck in ‘limp home’ mode all the time.

Note: It is likely that the vehicle would have smoked and used engine oil. These symptoms were not mentioned but it is highly likely to have been the case.
It is possible that the oil consumption was in the order of 1 ltr between services so it might not have been noticed or indeed low engine oil may have been the trigger to book the vehicle in for a service.
The exhaust on this vehicle expels the exhaust gases under the rear of the vehicle and the relatively high rear glass may have meant exhaust smoke was hard to detect from the driver’s seat.