There are situations where all diagnostic processes taught by mainstream training courses won’t help or lead to wrong direction, I have one such story to share, hopefully it will be inspirational diagnostic process.

Customer complaint:
Engine warning light appears suddenly while driving, engine low power until ignition OFF/ON

The background, collecting information:
Porsche Cayenne Hybrid V6, regularly serviced history.
Multiple return repairs, once a week the warning light is back.
Diagnostic tester always reads the same DTC, one or more cylinders misfire and sometimes all cylinders misfire DTC with very high misfire counter.
In the past repairs, (there was many return repairs) all related parts were replaced, all coils, spark plugs, injectors, some sensors, fuel delivery etc. Each time the test drive passed, and vehicle was returned to customer but came back as return repair few days or weeks later with the same DTC. (there was in plan engine ECU and wiring harness also)

First diagnostic approach with different strategy:
Given the history of new parts installed, it was clear that a different strategy is needed, the question was where to start? As always, the best is to delete all the history and information given as it is rather misleading. The previous technicians had good experience and it is good enough reason to assume they were misled by some tricky malfunction, most likely not related to already replaced parts.
First step is to personally experience the warning light at test drive, easy to say but few test drives did not log any DTC (no wonder it always passed QC). Trying to simulate customers style, non-aggressive driving style, more like cruising smooth style and suddenly warning light appeared. At low RPM (1200) at speed around 90km/h, there was felt engine misfire and car got high frequency vibration for about 3 seconds (felt like driving on fine, high frequent slow-down road strips).
All we got was the same, few cylinders DTC, cleared the fault memory and drive again, trying to learn how to easily replicate while having the Picoscope NVH in mind as next step.
After several time cleared DTC and repeated the same again trying different speeds and gears, we have learned the vibrations were related to almost zero torque around 1100-1200 RPM at any gear from second up, so the NHV would not reveal anything new.

New diagnostic plan:
There is a Porsche TPI document describing vibration caused by transfer box (additional gearbox responsible for 4x4 drive, central diff), but this was already processed at previous repairs.
However this is the direction most likely to go, vibration from transmission, E-machine or elsewhere, as it is clearly related to RPM and torque.
So, given this is a hybrid vehicle, anything on powertrain can be responsible and easiest way is to monitor the RPM of individual parts of the powertrain. Also, there was a question, could it be caused by the E-machine or the DC/AC converter? (Hybrid electric motor-generator)

Picoscope connection and recording:
Let’s connect to cylinder 4 ignition coil driver signal for sync (easiest to reach), then engine RPM on crankshaft sensor (at ECU socket), one of the hall sensors of the E-machine (wire harness) and on the last channel the HV current at orange cable from HV battery to DC/AC inverter using current clamp. Now we know how to replicate the malfunction, all we need to do is to test drive and record. Set the Pico software to “action” where once the software fills up the buffer, it will save the file to HDD, then start recording again. All this will run automatically without any input until button stop is pressed. This allows us going for test drive without second technician who would have to manually save the recording each time the event happens.

First conclusion after data captured by Picoscope and analysed:
First picture below shows the point when the vibration started. On second picture is visible the vibration in one engine revolution, both the engine and E-machine rise and lower the RPM approximately two times between spark command, so double frequency in relation to engine cycle and similar to engine RPM. (using measurement, we get frequency around 15 Hz) This explains the misfires DTC that are pointing to random cylinders or all cylinders with too many counts. The engine ECU identifies a misfire by analysing the crankshaft speed acceleration caused by combustion pressure which accelerating the crankshaft, such vibrations totally “confuse” the ECU algorithm and therefore the software logic identifies it as misfires.
OK, we can replicate it and record as waveform in Picoscope, now just identify what part of the powertrain causes the powertrain 15Hz frequency vibration.
Below is simplified diagram of the powertrain, it contains these parts in order and possible speed monitoring points. Engine (crank sensor), Hybrid clutch (no sensor), E-Machine (3x Hall sensor), Gearbox (1x input and 1x output sensor), Transfer box (I did not check, unknown).

Record additional data support the deeper analysis:
The question is, does the vibration come from the gearbox/transfer box side? or engine? or is it the 3-phase electric motor/generator or the DC/AC converter? Let’s go for more test drives and record the vibration while also monitoring the gearbox input and output shafts speeds via the corresponding sensors.

Determine the vibration origin:
Here we have engine speed increase and decrease in frequency around 1x engine revolution or 15Hz exactly, the same identical frequency is at E-machine and both transmission speed sensors.
However, while the engine speed goes up when vibration is present, all the 3 other speed signals going down, exactly opposite. So, we can make statement the vibration differs by opposite amplitude between engine and the rest of the powertrain. The frequency is consistent in all 4 speed sensors, about 1x per engine revolution and at 15Hz regardless of the vehicle speed or gear engaged. Another words, the split or phase is between engine and E-machine where the Hybrid clutch is located.

Let’s have a look on the Hybrid clutch:
By looking at the drawing from parts catalogue, the clutch plate looks like a damper to me, something similar to double mass flywheel in term of function. So, there must be some kind of spring and some shock damping design in the clutch plate. Thinking of that, assuming the damping is non-functional, the clutch plate would be just a spring and each spring has own frequency, right? We can observe the vibration is always in one range of engine 1200 RPM or 15Hz frequency. It is not related to gearbox output so therefore it is not caused by gearbox or transfer box. The most logical diagnostic conclusion based on the data would be the Hybrid clutch plate malfunction.
(15Hz multiplied by 60 = 900, so the powertrain vibrates 900 times per minute and RPM is 1200, it is close so the RPM could cause the clutch plate resonance)
It means order new hybrid clutch and hand the car over to workshop for clutch removal.

Clutch plate physical inspection:
The conclusion was correct, and the hybrid clutch plate damping is gone, it acts as a spring only. The new clutch rectified the customer complaint and vehicle has been returned to customer.
See the short video below. (This NOTE is added 20/02/2024. Below are some notes explaining the dificulties why it was not easy to replicate the complaint, also it is good to mention the vibration was mostly possible to replicate after longer drive only. My explanation is the clutch had to get hot and therefore "softer" which allowed the "spring" effect. All the attached pictures and .psdata files are before the clutch has been replaced. There was no opportunity to record data on fixed vehicle after new clutch has been installed)

Here is the picture when no vibration is present, all combustions are equal.
And here the pico captured the vibration even it was not noticed by driver or DTC or warning light appeared. The following capture was a test drive when no vibration appeared, I have added it here just to demonstrate how easy can the technician get misleaded. The problem can be captured and analysed only if the recording time matching the event of interest, in this case the customer complaint. Thank you for reading this long article, I hope it inspires other Pico users to explore new Picoscope capabilities or help with decision to invest and add the Picoscope set to workshop tool list. So there will be more Pico users in this forum and more articles to read and learn from each other.
Regards,
Roman

You are a saviour to the folk who own that car.

Hello and thank you again Roman, I have forwarded your case study onto the auto team along with topic23410.html

Could I check you would be OK with links to these case studies being featured in our Newsletter?
https://www.picoauto.com/library/newsletter

Your case study raises an essential point about getting tuned to the symptom “Trying to simulate customers style, non-aggressive driving style, more like cruising smooth style and suddenly warning light appeared.”

Unfortunately, due to time constraints and working pressures this fundamental step in diagnosis is over-looked (Road-testing with the customer is an option to consider if the situation allows)

I remember suggesting this to a team of technicians in Canada who laughed and went onto to explain how some vehicles arrive at their service department on a train, where they off-loaded, service & repaired then and loaded back onto the train without ever seeing the customer.

Moving on

The conditions for the appearance of the vibration were challenging by themselves @ 1200 rpm with Zero torque. Knowing now the fault, our hybrid clutch would be allowed to chatter/vibrate (hind-sight being a wonderful thing)

I can visualize the clutch momentarily loading and off-loading the engine due to drivetrain/vehicle load against minimal engine load (zero torque)

We can see this activity for over 4 seconds in your 2 Vibration appears, detail capture

Whilst the engine is running at approx. 1205 rpm, the undulating frequency is approx. 16.6 Hz
For the engine speed to deviate repeatedly by 174 rpm at a frequency of the 16.6 Hz, the vibration must have awful and I can see how this influenced the misfire counter.

I find the phase shift between Engine rpm and E Machine fascinating as this remains present after fix, however the vibration time is reduced to approx. 1.6 seconds and the rpm undulation to approx. 128 rpm
I assume the engine ECU would allow for this characteristic at these engine speeds / loads and momentarily turns off the misfire counter. (i.e. this it can ignore but not the prolonged vibration before fix)

Rest assured Roman you are inspiring others which is a gift that will keep giving

Thank you

Take care……Steve

Hello Steve, Thank you for your comments and additional analysis which I value so much.
Regarding the Newsletter, we would be very happy if Pico adds our case studies in, it will definitely motivate the technicians who were working on these cars, diagnosing the problems with Picoscope. As well as many others who may start with installing free Picoscope 7 to laptop and practice with files from this forum or elsewhere before they get hands on the hardware itself.
I have written the last part of this case study little bit confusing way, the last pictures and .psdata files appears like the post-repair data while I wanted to add few extra as demonstration why the problem replication was not so easy. (I have edited, added extra note in brackets at the end just to explain this) So I apologise for misleading here.
However, as you mentioned, the Picoscope has captured the vibration, but nobody noticed it at this test drive, no DTC or warning light. I think as the road also provides vibration and as you said the ECU ignores or filters out such short period events as problem. I have discovered this short vibration only after test drive, it demonstrates the advantage to record problem even before it is noticeable. It is also good to mention the frequency was between 14 Hz and 17 Hz but never 10 Hz or 20 Hz as this would be outside of the clutch spring frequency and basically avoid the resonance.
Regards,
Roman

Thank you Roman, and thank you for the edit as the phase shift both fascinated and niggled me at the same time.

What a real shame we did not get the opportunity to capture post fix data. This often happens as the customers we support rectify the issue and release the vehicle

This I understand as payment is required and the vehicle under test is consuming precious workshop space

If there is ever an opportunity to run the test again on a known good vehicle that would be awesome.

I will also try to replicate this test with such a hybrid vehicle as the drive plate/clutch damper is the weakest link in the drive train.

If we can develop a test plan it will become incredibly useful as the phase shift and offending frequency appear to be the key to diagnosis

Thank you again, take care......Steve

Hi Steve, I will get the good know car waveform when the same vehicle is available, it just take some time to have car and time together. Once I have it recorded, I will upload the file here and also to waveform library, good and bad one.
Regards,
Roman

No worries Roman, if you can that would be great but we also have to be realistic as post fix captures are like Rocking Horse Poo!

Setting up another car is going to take considerable time and effort which is better spent on a vehicle with a problem

I will search for something similar too, but like everything else, other priorities take over

What we have in your case study is GOLD already

Take care......Steve

Hi All Picoscope fans,
After few months of this car on the road, I have customer feedback as solid confirmation this vehicle had never problems since the repair.
Such good job the Picoscope has done!
Regards,
Roman