Booming noise on odd vehicle

dustymills · Post by **dustymills** » Thu Jan 04, 2024 9:25 pm

Hello all, I am attempting to diagnose a vibration/booming noise on a custom powertrain installed on a pickup truck. This powertrain consists of an electric motor, mounted to a gear reduction (1.9:1) box, through a short driveshaft, through a transfer case (2HI is 1:1, directly coupled), through a longer propeller shaft, and finally into the (3.76:1) differential. I've attempted to set up the vehicle as discussed here: topic23103.html?&p=103941&hilit=optical#p103941 . I'm using a hall effect sensor measuring driveshaft speed for my RPM input, and I have my 'add vibrations' calculated from that.

In the capture I'm sharing, I have a 3axis NVH sensor mounted to the bottom of the steel dashboard, vertically, with the X axis facing FWD. Unfortunately, because of the modifications made to the vehicle, there are very few things I can take for granted during my diagnosis, so I'm trying to develop some sort of plan to gather XYZ data from different components, then measure the difference my changes make.

I'm here to ask for confirmation that my setup is valid, and any guidance on a test plan is appreciated also!

dustymills · Post by **dustymills** » Mon Jan 08, 2024 10:59 pm

I've continued to attempt to diagnose this vibration, and while it's still as strong as ever, I have some additional information to consider. First, I transposed the axle ratios from another project, they are in fact 4.56 and I have updated my pico settings to match. I also calculated the tire speeds with that correct information and am able to estimate the T1-T3 more accurately. Second, this vibration does clear up on VERY light cruise/coast, it is definitely induced by load. Third, I have stiffened the gearbox crossmember to confirm it was not being excited, I felt no change and confirmed it flexed less on video.

In addition to the measurements I have made, I have test driven the vehicle in 2HI, 4HI, 4LO, and without any rear driveshaft. The vibration stays the same, although the overall noise does go up a bit riding in 4WD. This has led me to consider the transfer case as the source for this vibration. Using the same test drive pattern & route, I made two recordings with the XYZ sensor mounted to two different locations: The motor/gear reduction assembly, and the transfer case. In both scenarios, the sensor was mounted vertically with the X screw threads facing forward. To me, it seems that the transfer case is experiencing much more vibration than the motor/gear reduction assembly.

Does this sound like a reasonable conclusion? After initially feeling this vibration, I did inspect and overhaul the transfer case with new bearings and seals, and found no damage.

Post by **Steve Smith** » Fri Jan 12, 2024 9:16 am

Good morning and thank you so much for the post which sounds like an interesting challenge and unique too

I am so sorry for my delayed reply, I will study your configuration and data ASAP and rest assured I will feedback

Take care.....Steve

Post by **Steve Smith** » Sun Jan 14, 2024 4:56 pm

Hello and thank you for the posts, images and pddata files, sorry for my late reply

This sounds like an intriguing case and given we have an EV power source, even more so

What I have noticed in all you captures is the “clipping” in the signal history (I.e. the signal is being “cut off”) which can introduce a variable when looking at peak vibration

Given the vibration levels are high (below we have 240104_VibrationAnalyser_003 good capture) measured at the steel dashboard we need to switch our accelerometer location to “Engine compartment”

: 1

Above I have used the right click feature on the frequency view to select Single Display mode and we can see the vehicle is being offended in the X axis (Fore-Aft direction) Hopefully this provides another clue to the vibration source

Above we have 187 mg during what appears to be progressive acceleration with D2C being identified as the offender. Just to play devil’s advocate, the accelerometer on the steel dashboard does concern me given the reduced mass. Could the accelerometer be moved to somewhere like the seat bolt through to the chassis frame (Once again to remove variables)

Moving onto your custom vibrations (thank you for including this link topic23103.html) and the calculations, could I confirm our optical pick up is connected to the driveshaft/propshaft between the transmission (1.9:1) and the transfer case?

Our optical sensor is now our fundamental speed signal by which all other speeds are calculated

This is normally E1 (engine speed via OBD) but here we are using the optical sensor aimed at the driveshaft/propshaft. Therefore, the signal history blue line graph is displaying the speed of the driveshaft/propshaft (D1C custom vibration)

If the above is correct then motor speed is calculated as follows:
E1 x 1.0 = D1C (Shaft at which optical sensor is aimed)

M1 (Motor speed) = D1C x 1.9 (Gear reduction connected to motor)

Our other custom vibration orders are as follows:
D2C = D1 x 2
M3 = M1 x 3 (3- Phase Motor, phase frequency)
T1C = E1 (or D1C) x 0.219 (Using diff ratio 4.56:1)

Based on the above, our displayed orders are correct and we have a 2nd order driveshaft/propshaft vibration (D2C) in the Fore and Aft direction
2 shocks per one revolution of the driveshaft/propshaft
On a conventional drive train this would be P2 which can be attributed to Universal joint (UJ) wear or alignment. Given this vibration increases with load, this would also point towards UJ’s

Below (Display mode “Single”) we can see D2C fore and aft prominent vibration level @ 184 mg but note that all axes are high when measured at the Transfer under load

: 2

Below we have the Vector sum display mode of all three axes

: 3

Below we can see how removing the load (decelerating) the vibration (D2C) falls dramatically

: 4

Below we have a 3D view of the above vibration which includes a 3rd Time axis for viewing the rise and fall of vibration during acceleration and declaration

: 5

Below we can see a lower level of vibration at the Motor by comparison to the Vector sum at the transfer. We are therefore correct in focusing on the shafts to the transfer, at connection to the transfer. (Front shaft to differential only relevant when transfer at 1:1)

: 6

Just a note on the graphed road speed, I assume on road test, gear 1 was selected? I ask because we have road speed displayed but neutral selected in the screen shots above and below.

To summarize

• Given the vibration remains with the rear shaft removed, we can focus on the front shafts
• Can the front shaft (to the front differential) be removed with the rear installed? (Only relevant when Transfer in 1:1)
• Reset your accelerometer setting to “Engine Compartment”
• Inspect shaft alignment
• “Food for thought” add optical tape to your shaft for a speed signal rather than a magnet as this may introduce a potential imbalance (I noticed this in your second image)
• A USB webcam under the vehicle may help locate component response as we have here in this case study https://www.picoauto.com/library/case-s ... celeration here is the video https://youtu.be/JUCy-bZwOmE Keep your eye on the transmission casing
• Using the Slow-Motion feature of your Smart phone is another option to consider

I hope this helps, take care……Steve

dustymills · Post by **dustymills** » Mon Jan 22, 2024 4:30 pm

Steve, I wanted to first thank you for your detailed analysis and suggestions. They have given me the confidence that I'm spending my diagnostic time efficiently. Unfortunately, I am having a small struggle to get good data to share with you, and the weather has prevented me from driving this truck much. That said, I wanted to let everyone interested know that I am hard at work getting this sorted!

Based on your suggestions, I've updated my accelerometer settings, and the physical sensor position to the passenger seat frame bolt. I've tried to get a photoelectric setup running with what I have on hand (pico photo-sensor setup now on order!), but for now, my best results have been with two magnets spaced evenly on the driveshaft, and I've updated my pulse settings. I do plan on improving this in the near term. Unfortunately, my most recent captures have all had issues with either my RPM signal dropping out, or one axis of the sensor being 'pinned' to zero (attached). I'm working hard to sort this out and get good data. I still struggle with remembering to shift the vehicle into 1st on the software. Last, I have been able to get some useful information from a GoPro camera running in slow-mo under the vehicle. I'm currently manual synchronizing the start points, but I may look into a USB cable to run this in split screen.

Most importantly, I have modified the motor and gear reduction mounting setups to lower the motor in relation to the transfer case. My focus had been on making sure the components were parallel, but the culprit seems to have been the working angle of the joints themselves. The data taken with the NVH hardware is what has given me the confidence to pursue this. Modifying my motor mounts has reduced the vibration greatly.

Now that I have results moving in a positive direction, I'll redesign these mounts to place my drivetrain in a more ideal position. We might even get some newer tires installed, I suspect these have flat spotted to some degree.

Of course, I'll update when this vehicle is buttoned up, or if I have any interesting updates.

Again, I sincerely appreciate your insight and welcome any ideas from the community.

PicoScopes
& kits

For vehicle manufacturers

EV kits and
accessories

Noise, Vibration and Balancing

Milliohm tester

Pressure sensors

Current clamps

Breakout
leads/boxes

Accessories

Automotive Forum