- Pico Staff Member
- Posts: 1567
- Joined: Sun Aug 25, 2013 7:22 am
Every once in a while, a “problem car” comes along that defies all diagnostic knowledge and evades every possible test sequence so leaving you stranded!
Below is one such vehicle that drove me insane with endless nights studying and numerous conversations with colleagues only to draw a complete blank.
• Vehicle details: Audi A4 3.0 V6 TDI
• Engine code: CRTC
• Year: 2017
• Symptom: Severe lack of power from rest
• Mileage 32820 miles
Vehicle fails to accelerate rapidly when pulling away, e.g., pulling away at traffic lights or pulling out of a road junction
Verifying the customer complaint is an essential step in the diagnostic process but quite often a time-consuming task with little success. On this occasion, the fault was clearly apparent and I would describe the vehicle as lacking the required turbo boost to enable the desired acceleration. Multiple warning lights were illuminated and performance beyond 2500 rpm appeared normal (i.e., there was a clear increase in torque thanks to an increase in boost)
Another note to add, the awful lack of performance was bordering on dangerous for a driver not aware of the symptom whilst driving during typical traffic conditions. In addition to the above, the driver activated “Launch Control” feature was inoperative. (There was no getting around the inherent lack of power from rest)
With the customer's complaint verified the Vehicle’s ID and Specification were confirmed. Confirmation of vehicle specification is of the utmost importance when it comes to diagnosis as there is often a temptation for customers to modify their car with fashionable accessories that lack the fundamental quality control and engineering that was intended for the vehicle. (This now extends to software with regards to “performance” upgrades etc.) A brief evaluation of the vehicle did not reveal any concerns in the form of aftermarket accessories.
The Customer Interview utilised the four targeted “open” question approach to establish facts from fiction:
1. How long has the problem been evident?
At least 12 months
2. When did you first notice the problem?
The customer could not be 100% sure on this question but feels as though a tolerable lack of power may have started 6 months prior to the 12 months mentioned above and become progressively worse with time
3. Has any work been carried out on the vehicle recently?
In a nutshell, YES. Replaced items include.
Alternator (due to engine oil contamination)
EGR & Cooler
Intake manifold & swirl flap unit
Fuel rail pressure sensor
Camshafts (Part of an authorised modification)
Intercooler and hoses
4. When do you experience the problem?
When pulling away from rest, regardless of engine temperature
The Basic Inspection confirmed no fluid leakages, no visible signs of damage to hoses, connections/ wiring harnesses and no sign of my favourite discovery; accident repair. Known fixes/bulletins had also been checked and confirmed as either actioned or not applicable
Before I continue, the above vehicle was recovered to Pico given it was neither taxed or insured for the road. The reason I mention this is that to reproduce the symptom, you only need to pull away from rest which could be actioned in our very own car park. This style of road testing will come back to haunt me later on in the study!
A vehicle scan of all onboard control units revealed the relevant code below
Boost pressure regulation
P0299 00  – Control range not reached
BINGO, this DTC was exactly what I was expecting based on the symptom and how the fault code could be generated during car park testing. I.e., if the vehicle was driven gently from rest, all appeared near normal, however hard acceleration from rest (gas pedal to the floor) resulted in the appearance of the above fault code.
• Intake air leak or restriction
• Exhaust leak or restriction
• EGR held open
• Turbocharger control
• Turbocharger failure (New OE turbo charger installed)
• Engine mechanical integrity & valve timing
The action plan is predominately governed by accessibility, probability and cost. Based on the acquired fault code and symptom, the following items became the targets for scrutiny
• Air flow measurements/evaluation
• Turbocharger actuator control
• EGR operation
• Engine mechanical integrity tests
As ever, starting with serial data (VCDS) makes perfect sense to obtain an overview of parameters under the fault condition with minimal intrusion.
Below we plot the target and actual MAF, Boost / Fuel pressure and turbo actuator position in relation to WOT when pulling away from rest
As can be seen above, target and actual MAF and Boost pressure are most certainly low. A word to the wise about the Target MAF within Image 1 above, note how the value peaks at over 300 gm/sec!
This of course is not the case; it was noted that when the gas pedal is floored from rest, MAF serial data simply jumps to 2000 mg/stroke rather than plotting the change in target MAF. (mg/stroke is converted to gm/sec for the purpose of displaying the same units in our graph for target and actual MAF)
At this stage, please focus on the expected trend (target) MAF in Image 1 when the gas pedal is pressed to the floor
Note in image 2 how our actual boost pressure is way below target, yet the target and actual turbo actuator control signals (Image 4) are perfectly matched!
How can the turbo actuator control signal be correct yet the turbo boost pressure be too low?
An examination of the turbo control actuator and pushrod confirmed no sticking and seamless operation/travel during acceleration from rest (I captured this event using a webcam under the bonnet)
Enough of serial data, it was time to verify the above with PicoScope
Below we have captured Engine speed & MAF using the Crank & Frequency math channels respectively, along with manifold and fuel pressure (Manifold pressure captured using WPS500x)
Given the gas pedal position sensor utilises the SENT protocol, for now, we will refer to the initial increase in MAF (CH D) at 1.253 seconds as an approximation of “pedal to the metal”
Note the huge flat spot of approx. 3.5 seconds where fuel pressure (CH B) & boost pressure (CH C) appear to plateau where I would expect both to increase dramatically. Focus now on Channel D and in particular the apparent disproportionate air flow increases for very little gain in engine performance. (Could this be an intake air leak?) Note our peak boost pressure at approx. 4100 rpm = 1.287 bar gauge (2.287 bar absolute)
Let’s take a step back here and think about the typical order of events that take place during acceleration from rest.
1. Fuel pressure should initially increase in relation to gas pedal position (regardless of engine speed) I.e., Fuel pressure often mirrors gas pedal position. Fuel pressure should then continue to increase with torque and load requirements
2. MAF and boost pressure increase in relation to turbo actuator position and engine speed
Based on the above statements, do we therefore have low fuel pressure? I think we can agree from the above data we are making progress but not drawing any diagnostic conclusions
Whilst there are numerous other tests we can carry out, I decided to find a donor vehicle to obtain objective data on which to base further intrusive diagnosis
Now, finding an exact donor car is like finding an honest politician and sometimes you have to settle for what you get. Below we have a back-to-back test of MAF, boost and fuel pressure using an Audi Q7 but with the identical CRTC engine
I think the above data speaks volumes and demonstrates the correct interaction between MAF, boost and fuel pressure. The Q7 and the A4 are like chalk and cheese, note the behaviour of boost pressure at WOT and the peak pressure of 1.907 bar gauge (2.907 bar absolute)
Time to recap:
• Both serial and scope data confirm low boost pressure, however the scope delivers far superior resolution accompanied with the instantaneous interaction of MAF, and fuel pressure during the flat spot
• PCM returns fault code P0299 00  Boost pressure regulation – Control range not reached
• Peak boost pressure is low: 1.907 – 1.287 = 0.62 bar (Images 5 and 6)
• Target and actual boost control actuator position values are correct (image 4)
Based on the above, should I be replacing the new turbocharger for another new turbocharger; absolutely not!
At this point I have chosen to manually verify fuel and boost pressures along with MAF using the techniques below:
Fuel pressure test
Thanks to Frank Massey’s awesome video here https://youtu.be/tXEkQt9ZODQ we can not only proof test the high-pressure diesel pump but also the integrity of the priming system and injectors for leakage. Below we confirm our rail pressure to be safely held at 4.614 volts (Rail pressure sensor voltage) with a rise time of 172.4 ms. This is indeed an efficient and leak free high pressure diesel system
When evaluating MAF it helps to know the expected airflow if this engine were 100% volumetrically efficient (100% VE) We have discussed this topic in depth here topic21311-20.html#p99396
Based on data acquired with VCDS at 4360 rpm we obtained a MAF value of 239.86 gm/sec.
The engine power output of our CRTC unit = 275.77 PS which would consume 275.77 gm/sec if 100% VE. The typical volumetric efficiency of a modern-day diesel is around 88%
Therefore, to calculate the VE of this engine we have our acquired MAF @ 239.86 / 275.77 * 100 = 87% VE
I am happy at this stage our peak airflow is at an expected level when the engine performance is near normal above 3000 rpm (Remember our faut condition is from rest to around 2500 rpm)
Boost Pressure/MAF test
Now, this test is not recommended as the consequences are huge with engine failure being the extreme outcome. Removing the pushrod between the turbo actuator and the turbo vane housing allowed for the vanes to be fixed in the upper most position so generating maximum boost and removing PCM control of the turbocharger. Pulling away from rest remained poor with the huge flat spot still present, however, our boost pressure increased to 3.487 bar absolute with an airflow of 316.7 gm/sec.
This proves the integrity of the turbocharger when the vanes are fixed in the position required for take-off from rest and also qualifies the air intake, exhaust and EGR system as our engine is able to inhale and exhale at a rate of 316.7 gm/sec.
Knowing what we now know from the above tests, why does our boost pressure remain too low during take-off from rest? An air leak would result in low boost pressure but high airflow, in our scenario we have low boost pressure and low airflow! If we had a physical opposition to airflow, we could not achieve the peak airflow figures (316.7 gm/sec) we have captured above. A physical opposition to MAF at low rpm would have the potential to produce effective boost pressure as opposition to flow = increase in pressure!
Could we therefore have a mechanical issue with our engine? (The final item to inspect in our “Action plan”)
A relative compression did not reveal any concerns and so camshaft and crankshaft correlation were next up for inspection
Below we have the “problem” Audi A4 cam and crank correlation results at 74.33° using the missing tooth of the crankshaft as our “Zero-degree point” and the falling edge of the smallest pulse of the RH intake camshaft adjacent to the injection event of cylinder 1 (engine at idle speed) Please see here for more information on this measurement technique topic18471.html & topic21754-10.html
Below we have a donor Audi A8 utilising the TDI V6 engine but with a different engine code! Luckily, they are a near perfect match from a cam and crank correlation point of view at 74.77° (engine at idle)
Comparing the cam and crank correlation during acceleration and deceleration confirmed a phase shift of approximately 2.16° for both engines. Whilst these tests inspire confidence regarding cam and crank corelation, they only relate to a single camshaft (only the RH intake cam incorporates a cam sensor) and therefore we cannot qualify the dynamic in-cylinder valve timing.
Thanks to the WPS500x we can evaluate the characteristic pressure activity in each cylinder to determine valve open and close events. Overlaying these test results will enable qualification of in-cylinder valve timing
Below we have all 6 cylinders overlayed so as to ensure the key points of interest (EVO, EVC, IVO & IVC) align. Note how the second compression towers are all misaligned compared to the first! This is simply due to changes in cylinder speed during the 4-stroke cycle. Note also how peak compression is low and the waveforms resemble a conventional gasoline engine at idle. This is because the air intake was purposely restricted in order to exaggerate and clearly define EVO, EVC, IVO, & IVC along with intake and expansion pockets that are not normally visible with a diesel engine (I.e., Diesel throttle plate is typically wide open and so very little intake manifold vacuum exists)
Additional tests confirmed peak cylinder pressures to be correct at WOT during cranking without restriction in the intake manifold. All cylinders returned approximately 25 bar where I have included cylinder 3 below at 25.58 bar as a typical example. Note: When using WPS500x for peak cylinder pressure measurements, remember to compensate for the compression hose and dummy glow plug volume, please see the following forum post topic16131.html
So where to now? We appear to have exhausted the “Action plan” and yet still we have a huge lack of power when pulling away accompanied with fault code “P0299 00  Boost pressure regulation – Control range not reached”
My thoughts return now to the turbocharger, could this be the incorrect part installed as the replacement part number did not match the original part number!
Original A4 Turbocharger:
Garrett number 839077
OE part number 059 145 873
Replacement A4 Turbocharger:
Garrett number 888580
OE part number 059 145 873 FA
Listed for C7, Q7 and Amarok via the Garrett website
After much deliberation with Garrett and VAG, it appears OE superseded part numbers may not initially keep track with part numbers utilized by Tier 1 suppliers (like Garrett) for the aftermarket. This may occasionally result in a “disconnect” when cross referencing new OE part numbers with part numbers used for the aftermarket. What I am trying to say here is, the turbo installed (whilst different to the original part number) is correct for this vehicle.
Like there are not enough variables to consider when diagnosing!
Time to get back to basics and consider the operation of a turbocharger, the primary function of which is to generate flow. Flow is dependent upon compressor speed and so if we can measure this speed and compare with a known good donor vehicle, we may have a breakthrough in our diagnosis if a difference is found
So how can we measure and plot turbo compressor speed when pulling away from rest? The answer is “with great difficulty”!
I made numerous attempts using the Pico optical sensor https://www.picoauto.com/products/nvh-a ... sensor-kit aimed at reflective tape adhered to a single compressor vane. Below we have captured the turbo compressor speed with the VNT lever manually shifted from rest to its uppermost position (max boost) and back to rest with the engine at idle speed
Note above, we use a frequency math channel to denote compressor speed which varies from 6479 rpm to 27,600 rpm. Here we conclude the turbocharger VNT to be functioning correctly and the desired increase in compressor speed achieved.
Moving onto capturing compressor vane speed when pulling away from rest (PCM controlled) To say this was challenging would be an understatement. Unfortunately, whilst the initial increase in vane speed was captured, the tape would become detached from the compressor vane and ingested by the engine! (Not for the faint hearted) See below and note the behaviour of all waveforms at 5.187 seconds
Note above how compressor vane speed, manifold pressure and engine speed are slow to initially respond to WOT then begin to increase dramatically from 5.187 seconds onwards at approx. 1827 rpm. Here we capture a considerable and sudden increase in compressor vane speed which reinforces the original symptom described early in the “Technical Description” where an increase in torque could be felt once the engine rpm achieved approx. 2500 rpm. At a compressor vane speed of approx. 120,000 rpm, the sudden drop to zero does not reflect the compressor vane speed but the point at which our reflective tape was lost forever (ingested by the engine)
We are now starting to build a picture of what is affecting the performance of this engine from rest and that would be insufficient air flow as a result of low compressor vane speed from idle to approx. 1827 rpm
Up to this point in the diagnosis, the vehicle had not left our carpark and so it made perfect sense to investigate how the vehicle drove on the open road and if any other symptoms would come to light and assist with further diagnosis. After acquiring Trade Plates, we could now adventure onto the highway to carry out a full assessment of the vehicle performance under varying load conditions. (Something I should have actioned at the very start of this study!)
It soon became apparent that under light throttle application when cruising at approx. 60 mph (on a gentle incline) the vehicle exhibited a vibration which could be felt in the floor pan. Further analysis of this vibration confirmed a combustion anomaly which manifests itself as an E0.5 vibration order (one disturbance for every 2 crankshaft revolutions) More information on engine vibration orders can be found here viewtopic.php?p=86961#p86961
Given we know the mechanical integrity of the engine to be good, the focus of a potential combustion anomaly (or dare I say misfire) required further analysis of injector quantity deviation values using VCDS. Below we can see cylinders 1 and 6 are of a real concern!
Have we found something new here and totally unrelated to our lack of boost pressure or, could they somehow be related?
To reinforce the NVH and VCDS data above, the crank signal was graphed using PicoScope and cylinder 6 was most certainly underperforming during the load conditions described above
Following on from the above, increasing the filtering & scaling applied to Channel E (Fuel pressure) we can see below, fuel appears to be injected based on the repetitive cyclic nature of the fuel pressure signal which typically falls as crankshaft speed rises (The rise in crankshaft speed is attributed to combustion thanks to fuel injection)
A review of DTCs after the above road-tests revealed additional fault codes in the following occurrence order which suggests the effects of incomplete combustion has been detected by emission control sensors
All emission related fault codes above refer to Bank 1
1. Lambda sensor is the first to detect a fault 16:14:16 @ 52895 km
2. Next: NOx S2 B1 16:15:46 @ 52897 2 km later (at similar load and speed to Lambda)
3. Next: NOx S1 B1 16:16:33 @ 52899 2 km later (Slightly increased engine load and speed)
4. Next: Boost regulation 16:23:57 @ 52908 9 km later (high load at 86% ACC pedal position)
• Codes 1, 2 & 3 above are all set under the same driving conditions across 2 minutes and 4 km
• These codes are related to our incomplete combustion (E0.5)
• Code 4 is a stand-alone code and not linked to emission codes (1, 2 & 3)
• Once again, the target and actual position of the turbo charger actuator were correct when the fault occurred. Boost pressure however was too low by 515 hPa (515 mbar or 7.5 psi)
At this stage “enough is enough” and we cannot pursue further diagnosis until our combustion anomaly is resolved. The decision was taken to install 6 injectors and then continue diagnosing the low boost pressure thereafter. The vehicle was handed back to the workshop in question and a few weeks passed before I received a phone to say the car is cured! I initially interpreted this call as the “misfire” is cured but after repeated questioning of the technician it soon transpired the low boost pressure was also resolved. With complete and utter disbelief, I returned to the vehicle to carry out post fix tests and evaluate boost pressure when pulling away from rest.
Note below we have engine speed, fuel pressure and MAF only as a WPS was not installed to the intake manifold. Please focus on the dramatic rise in engine speed from idle within the 1.857 second time span between the time rulers
Given the vehicle was now reassembled and ready for hand over to the customer, I grabbed as much data as possible from the onboard sensors to enable full evaluation of boost pressure in relation to gas pedal position. Both the MAP and ACC Pedal position sensors utilise the SENT protocol in order to transmit such data to the PCM. To graph this data we can use the Pico SENT decoder as Ben has described here https://www.picoauto.com/library/traini ... -picoscope and please also refer to the forum post here topic22061.html
Whilst the above links will describe in detail how to decode SENT, the following link viewtopic.php?p=102933#p102933 will then demonstrate how to graph this data alongside your analogue data such as MAF, fuel pressure and engine speed. (In the link above I have used Image 19 which will help with interpretation of image 20 below)
Note above we have now included the decoded MAP and APP sensor data (SENT) alongside MAF, fuel pressure and engine speed in order to demonstrate the “response” and relationship between all inputs. I hope we can agree the MAP sensor (boost pressure) increases exponentially soon after the gas pedal is pressed to the floor.
Now for the Million Dollar question, “Why did the boost pressure increase after replacing fuel injectors?”
After a lengthy phone with Andy Crook at GBR https://gotboost.co.uk/ it transpired that turbo compressor speed is not only dependent upon the Kinetic energy (dissipated into the turbine) but more so Heat Energy! The following link has this covered https://van.physics.illinois.edu/qa/lis ... d-turbines and worthy of a read because it simply reinforces why this vehicle lacked turbo boost from rest. There was simply insufficient heat generated as a result of poor combustion under load.
Please note that this vehicle did not misfire, smoke, judder or rattle when pulling away from rest, it simply exhibited a huge flat spot! Yet another example of just how difficult conclusive injector diagnosis can be which we discuss here https://www.youtube.com/watch?v=7jFf32L ... Automotive and here viewtopic.php?p=103210#p103210
So, why was combustion temperature too low?
The answer here lies in the bench testing of these diesel injectors and a huge thanks goes out to “Colchester Fuel Injection” https://www.colfuel.co.uk/ for allowing Pico access to their test procedures and Bosch rig (See below)
Each of the original 6 piezo injectors removed from the Audi A4 were subjected to 12 tests as follows:
Leak test at 30 MPa: Pass
Leak test at 160 MPa: Pass
Start test at 30 MPa: Fail (Low delivery rate)
Back pressure: Pass
Start up: Pass
VL at 199.9 MPa: Fail (Low delivery rate)
EM at 79.9 MPa: Fail (Low delivery rate)
VE at 120 MPa: Fail (Low delivery rate)
LL at 25.1 MPa: Pass
IMA calculation: Fail
As can be seen above and below, the “Start test”, VL, EM & VE tests all fail due to poor flow rates (quantity) under varying load conditions. Noting the test pressures, we can derive VL must be full load” and LL “light load” (EM and VE are therefore varying loads in-between light and full)
OK, so what caused the low flow rate of these injectors and why were all affected in the same fashion?
Dismantling injector number 6 hopefully goes someway to revealing why.
Below we have a number of images demonstrating a clear passage for pressurized diesel from the HP injector pipe inlet through to the nozzle valve and needle (Image 26) However, the nozzle cap of the injector within the combustion chamber appears to have been exposed to damage in the form of possible detonation (hammering) which has closed or reduced the intricately machined spray holes (Images 28, 29 & 30)
As we can see, the spray holes are most certainly deformed (if not closed) which of course will affect delivery rate /quantity and more so “spray pattern” which is paramount for complete combustion. Given all injectors failed in a similar fashion, we can only assume the cause of such damage would be misfuelling! (A discussion for another time)
So, there you have it, what a journey and thank you for hanging on in there. I hope this helps someone, somewhere when you believe you have tested absolutely everything you possibly can, think outside the box, talk to colleagues, research and share your journey
- Posts: 65
- Joined: Mon Dec 14, 2020 3:38 pm
this will give me at least a day of reading and learning from the linked articles
great post! thanks!
- Posts: 9
- Joined: Tue Mar 20, 2018 8:41 pm
Well done Steve, what a great post. It proves persistence pays off!
I have seen poor quality fuel cause similar damage to injector nozzles before on 4 cylinder diesels but never on the V6's. When I have seen such damage on the past it has exhibited different symptoms to this too, namely a soft misfire and diesel knock under load due to a severe build up of carbon on the piston rings and damage to valve seats causing a lack of compression. Of course having DPF's fitted disguises a lot these days!
- Posts: 24
- Joined: Sun May 02, 2010 12:38 am
Great write up Steve. In North America the 2003 to 2007 Ford 6.0L diesel is notorious for setting a P0299 code for any low power concern. This can be turbocharger related, boost or exhaust leaks, engine misfire, or even low injector voltage. I was surprised at the beginning of the post to see that a late model vehicle would not set other codes with this issue but then you ended up showing us the importance of driving a vehicle for other code monitors to run.
I especially enjoyed the failure analysis that goes deeper that just 'bad part'. Roughly 13 years ago, I had enough of guessing and relying on pattern failures to diagnose vehicles. That's when I purchased a Pico and never looked back. Specializing in diesel fuel injection, I always dig deeper to the root cause of the failure and don't settle for 'bad part'.
As far as the nozzle damage, we can only guess at this point but one possibility could be related to the camshaft replacement. The injectors would have been removed for this procedure. If a technician had decided the clean the nozzles incorrectly, say with a wire wheel, the nozzle holes could have been distorted as pictured above.
- Pico Staff Member
- Posts: 1567
- Joined: Sun Aug 25, 2013 7:22 am
Hello and thank you for the feedback which is always appreciated
It's intriguing to know that a fault code (P0299) could be flagged to cover a multitude of possible causes yet the description was very specific in the case of the Audi (Boost pressure regulation – Control range not reached)
It does leave you a little bit between The Devil and the Deep Blue Sea when you have a symptom, that matches the fault code and accompanying scope data but still no fundamental fault to link them all together
The importance of the full road test will haunt me forever but we live and learn together
The nozzle damage will remain a mystery for sure and it has bugged me in case I have missed something like sabotaged fuel! (I would hate to think something was lurking in the bottom of the tank!)
With that said, the vehicle is back in service without any performance issues so all appears well.