We often encourage customers to record NVH issues on their Smart Phones (most don’t need any encouragement) as this provides the technician with evidence of issues that might be very intermittent in nature.
Let’s face it, the best person to capture these events are customers and most have Smart phones
Whilst Smart phone recording do not provide the definitive diagnosis they go a long way to assist with “probable causes” or raising suspicion about components.
Aux and Timing belt tensioners/idlers spring to mind here where they may whine for a very short period after start-up and then remain silent throughout the journey.
What is vitally important here will be your customer questioning as we need a few pieces of gold from them with regards to the vehicle running conditions at the time the noise was present.
For example, Engine Speed, Road speed, & Gear Position.
If we can just get Engine & Road Speed we are in with a shout.
In the example below we have a complaint of cabin noise at approx. 30 mph in 4th gear.
Combine these answers with your product knowledge of the vehicle (tyre size / gear ratio’s etc.) and we are ready to go……
Should a customer provide a video or audio recording of an NVH issue first we need to extract the audio from the video and then convert the audio file format to .wav (.wav is the only audio file format you can import into NVH)
For audio file conversion captured on my IPhone in .M4A format I used https://www.online-convert.com to convert to .WAV
Should you use the above conversion site select “Normalise Audio” (see below)
Audio File Conversion
The normalisation of audio files has the effect of increasing the volume to allow for all captured noise from the quietest to the loudest to be near audible
For extraction of audio from video there are a number of options under Google, where the extraction of audio can be carried out in a variety of media players.
Once you have the customers audio converted to .wav format we can import this file into our NVH software. To use the Import Audio Feature you will need to have your NVH licensed Scope connected to your PC
Open Pico Diagnostics and select NVH, close the Set-Up Wizard and go to the Set-Up Tab.
Enter the static RPM value where the customer complaint occurred (1950 rpm approx. in this example)
Static RPM
Next enter the Vehicle Information and include tyre size, gear and differential ratios
Vehicle Details
Now move to the Record and Analyse Tab and select Options > Advanced Options > Features and ensure “Enable Advanced Features” is selected
Enable Advanced Features
Close the Advanced Options box by selecting OK and once again click on “Options” and select “Load Audio File”
Load Audio Files
Click on the load button to locate your customers converted .wav file and select “Create Signal”
Create Signal
Create Signal.jpg (16.91 KiB) Viewed 22285 times
The customer’s audio file will now be displayed in both the Signal History and Frequency View of our NVH software
Frequency View
Given we have selected a static RPM of 1950 rpm and included the gear/differential ratios we can select the customers gear of concern (4th) and the software will display the road speed (32 mph)
Note: We cannot graph the rpm in the Signal History when importing audio. (This is characteristic of the software at this stage) Having the RPM, gear position and road speed displayed as above helps with a visual representation of the vehicle status
At this stage we are interested in the highest sustained peak within the Frequency View when driving on a smooth road surface. Notice the amplitude of signal history at 30 seconds drops considerably?
This occurs when driving from a noisy textured road surface to a quieter smooth road surface.
Using our knowledge of this 4 cylinder engine at approx. 1950 rpm the sustained peak we see below at approx. 66 Hz should be E2, a second order engine vibration. The following link will explain engine vibration orders here: post86961.html#p86961
Sustained Peak
Using a little theory here we know that Hz = RPM / 60
RPM = Approx. 1950 / 60 = 32.5 Hz (E1)
A second order engine vibration is a characteristic of a 4 cylinder engine (2 combustion events or shocks into the crankshaft per engine revolution)
E1 x 2 32.5 x 2 = 65 Hz
Our sustained peak of interest on a smooth road surface between 32 and 36 seconds in our signal history is most probably characteristic engine noise, possibly attributed to high gear, low road speed and increased engine load. (You can just hear this engine noise if you play back the audio)
Whilst this serves as an example on how to use the audio import feature, we would then need to simulate this with the customer vehicle using an Accelerometer/Mic and Mongoose lead to qualify our suspicion. With all that said, this is my car and has no faults!
I see this technique providing a form of remote diagnosis for nuisance noises.
If we know our vehicles well enough we could refer suspicious frequencies and amplitudes to our own personal data base built with years of experience and product knowledge
A word to the wise:
Smart phones are optimised for the human voice and incorporate a degree of noise cancelation at specific frequencies and so beware with low frequency noises etc.
Smart phone recordings may favour cabin noises as they are optimised for frequencies between 600 Hz and 5 kHz and have directional qualities so it may be useful to ask our customers to capture nuisance noises from within the cabin aimed at the area of suspicion.
What I am trying to say is don’t base a diagnosis on a Smart phone recording alone.
Hello,
Can be converted to waw or other audio format a .psdata file with a vibration?Today I captured some files but I am not sure if is good or not... there is some noise but I can't be sure if is a good capture or not.
Thank you.
Hello Victor thanks again for the posts and feedback
Your request to extract accelerometer data from a psdata file and convert to audio has been recognized and something we have discussed here at Pico.
For now it remains a New Feature Request whereby an option to switch between PicoScope and NVH (and vice versa) could be beneficial for a number of tests as well as NVH.
Going on the recording you made in this post topic21241.html
I have converted the audio into .wav format and imported this file into NVH
Victor NVH
Looking at the above image I have highlighted the frequencies of interest with a red rectangle as we appear to have 3 peaks that are multiples of each other (Harmonics) topic19731.html?&view=unread#unread
Given the noise appears to be low frequency, I have concentrated on all frequencies below 2000 Hz. Anything above I think we can ignore at this stage.
I have measured the first peak of interest at 47.2 Hz x 60 = 2832 rpm
This could well be E2 (2nd order engine vibration 4 cylinder engine) as 2832 rpm / 2 = 1416 rpm
E2 is often the easiest frequency to detect based on combustion events (2 per crankshaft revolution, 4 cylinder engine)
The first question then Victor is, was this the engine speed during the audio recording?
If so we have found E2, but what is interesting is how prominent the harmonics are after E2
Victor NVH 2
It would be good to see where we are with the diagnosis of this vehicle and how we may be able to relate these frequencies post fix.
Hello and thank you for your support,
Sometimes is necessary to confirm the recorded "noise/vibration" -that is the reason for my request about converting .psdata to waw or audio file .In a workshop the noise is at a high level,MEMS accelerometers mounted on the oil pan can "hear" an no/off command of solenoid(for example :A750f test) .
Can be used custom accelerometers with NVH module?I have some MEMS and IEPE accelerometers with technical specs,in SETUP only TA148,TA149 or TA259 can be selected...according to this download/file.php?id=10111&mode=view
At the first question the answer is yes,this recording was made at cold engine with engine speed 1380-1420 RPM readed in diagnose.
Best regards,
Victor
Last edited by victor2k on Wed Sep 26, 2018 10:26 am, edited 1 time in total.
Hello Victor and thank you for the feedback, sorry for the late reply.
Nice feature "listening to accelerometers" as you have described when listening/capturing solenoid operation. This removes all the ambient noise associated with the workshop whilst focusing on noise present at the point of contact. (AT sump pan)
With regards to your question can your MEMS or IEPE accelerometers function with our NVH software?
This all depends on their output characteristics.
Both devices (MEMS or IEPE) will require some form of power supply, hence the use of our TA259 interfaces with our TA143 MEMS accelerometers.
Basically the input into PicoScope NVH from the accelerometers should be maximum +- 2 V.
Therefore if you can somehow arrange the power supply to your accelerometers in such a way that their output can be connected to the BNC input of your scope, you will be able to display their output in our NVH software.
You would have to inform the software you have a TA143 connected to NVH via the relevant interface (mainly TA259) as there are no provisions to select another style of accelerometer
If you then inform the software your connected accelerometer is a microphone, post91701.html#p91701 you can listen to the audio captured by your accelerometer attached to the AT sump pan and also export the data as .wav audio files for your customer.
Sorry Victor, this is assuming your scope is licensed for NVH?
Be aware however with "accuracy" as the NVH is optimised for the TA143 accelerometer and matched to the output characteristics accordingly.
How the NVH software will respond to other accelerometers I am not sure, but if you only require audio, or a relative measurement then give it go.
Hello Steve and thank you for your support,
Until now I don't have a licensed scope for NHV,I asked to my local distribuitor for quotation but some technical infos can't be found in NVH kit description(for microphone yes,for accelerometer no) or in your posts.
The supply for IEPE and MEMS was solved,IEPE accelerometers was covered by an ancient version of PicoAutomotive for RST ...I think this option can be included in advanced setup , I have the technical specs of accelerometers.
Hello,
Here is the noise created by injectors(at 0.5E),captured by accelerometers instead of microphones.And the noise can be played and heard using some headphones to confirm it:D
Thank you