Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Sound Deadening Model 3

This site may earn commission on affiliate links.
So how can we find out what is causing Model 3 to be noisy on grooved concrete which is common hw at speed? Is it high inflation (at 42 to 45 which on all other cars I have owned was 32) or is it inadequate sound insulation?

It's most likely a combination of things. The tire tread and tire itself are probably reasonable at this point. Tire inflation pressure is high, but in my experience with my S reductions made only a minor difference. I do think Tesla's simply carry more noise from the suspension through to the chassis/body of the car. My suspicion is that they simply didn't invest a ton of time tuning this with either strategic weight placement, dampers (shocks/struts) or bushing compliance. Once that vibration/noise is present, you are stuck "chasing" it and patching around it with thinks like mass dampers. Beyond that, you have a general lack of noise barriers/blockers and less sophisticated door and window sealing. It doesn't take much of an opening for noise to travel. As an example turn on the radio or TV in a room and then start to slowly shut the door into that room. Notice how the TV noise outside the room stays almost the same right up until the door is close to being completely shut.

Finally, a lack of noise absorbers (like Thinsulate) is likely present and hard surfaces like that big glass roof do a great job of bouncing sound around (think about the difference between an empty room and one with carpets and furniture in it).

In short, no "silver bullet" typically when it comes to noise and this is why noise is part of "NVH" - Noise, Vibration and Harshness. They are all components that tie together.
 
  • Like
Reactions: chinnam3 and RBowen
Here's a mic for $22 that you plug into an Android phone and install an app for $8 that first reads the calibration curve for the mic (by serial number from their web site) and for $30 it gives you an accurate sound measurement tool.

We can then get some meaningful necessary measurements of the cabin noise, both levels and frequency distribution.

I haven't checked the Tesla cabin noise, but use this in the studio to set up monitor speakers and optimize listening position.

To illustrate one type of measurement the Audio Tool app shows, here's what I get playing pink noise from a USB stick on the "premium sound system" of the Tesla Model 3 with its equalizer controls set flat. With an ideal sound system all the bars would be about the same height. That huge bump around 70 Hz is a typical consumer audio bass boost. It's done to mask the lack of a bigger/better woofer. Some people like that, and the "premium sound system" sounds pretty decent overall, but I think that bump messes up the real bass.

Screenshot_2019-03-26-19-11-54-1.jpg



If you tweak the equalizer faders as shown you can knock down that fake bass peak and get a much more accurate (flatter) response overall. Without measurements, we'd be shooting in the dark. Same on our cabin noise, we need to know its frequency spectrum, and then we can also move the mic/phone around to try to locate the sources.

Screenshot_2019-03-28-21-54-34-1.jpg


Now, if you wanted to really enjoy those HipHop 20-30 Hz basses, you could add a little $120 active subwoofer ;-)
.
 
Last edited:
  • Informative
Reactions: JohnBinSC and KenC
Sorry about the huge images, I couldn't edit them after the fact. This sort of spectrum display, but captured from the mic without playing a pink noise source file, is what can show the frequency distribution and level of the cabin noise. Comparing saved spectra taken at repeatable fixed positions would show how any treatment would affect it. So, no, it's not a costly endeavor - the tool is cheaper and less time-consuming than any of the treatments. The Audio Tool app has lots of features to help get some meaningful answers.
.
 
Removing the dozen or so clips in a wheel well is very easy, after removing the tire first, of course.

Thank you. How did you decide where, and how much, dynamat to use?

The wife was complaining about the noise in the car last night on an hour long trip down the freeway. I observed that the car would be ideal if it had the acoustical properties of her Honda Accord Hybrid Touring model. Alas, it's far short on that score.

Any idea how much of that sound is caused by the stiff suspension and whether or not a swap in that regard is possible and would make any difference?
 
I didn't know about thinsulate. I think your result would be better than mine. Please keep us posted.
Are you going to use 3M thinsulate?
https://www.amazon.com/Thinsulate-A...1_1?keywords=thinsulate&qid=1573671601&sr=8-1

Yes. There are some less expensive rolls you get buy elsewhere. I heard you should even layer twice, for 4" thickness. The acoustic absorption will be extremely strong at 4", hopefully blocking all kinds road noise.

It appears similar material is already in the doors, so no big gains there.

Thinsulate is commonly used in van conversions to make the interiors extremely quiet.
 
Heres an article on the impact of acoustic foam inside tires.

Tested: Do Acoustically
Here's a mic for $22 that you plug into an Android phone and install an app for $8 that first reads the calibration curve for the mic (by serial number from their web site) and for $30 it gives you an accurate sound measurement tool.

We can then get some meaningful necessary measurements of the cabin noise, both levels and frequency distribution.

I haven't checked the Tesla cabin noise, but use this in the studio to set up monitor speakers and optimize listening position.

To illustrate one type of measurement the Audio Tool app shows, here's what I get playing pink noise from a USB stick on the "premium sound system" of the Tesla Model 3 with its equalizer controls set flat. With an ideal sound system all the bars would be about the same height. That huge bump around 70 Hz is a typical consumer audio bass boost. It's done to mask the lack of a bigger/better woofer. Some people like that, and the "premium sound system" sounds pretty decent overall, but I think that bump messes up the real bass.

View attachment 477148


If you tweak the equalizer faders as shown you can knock down that fake bass peak and get a much more accurate (flatter) response overall. Without measurements, we'd be shooting in the dark. Same on our cabin noise, we need to know its frequency spectrum, and then we can also move the mic/phone around to try to locate the sources.

View attachment 477149

Now, if you wanted to really enjoy those HipHop 20-30 Hz basses, you could add a little $120 active subwoofer ;-)
.
I am guessing that you really know how to use this equipment. I bought a special mic (With the serial number) about 10 years ago when I set up two home entertainment systems with carefully bought used high end Snell THX Speakers on E Bay and new high power Emotiva amps. and I ended up buying new 7 channel high end Yamaha receivers with the YPAO systems. Even my subs are Snell THX. I fully happy with these systems like I never thought I would. But that is what Professor Snell was famous for, sound measurement.
I was wondering if using the mike all around the inside while on a quiet and on a noisy road would help us figure this out. Where the sound was coming from in the inside and how best to improve it. Would love to follow you on this. Unless it’s inside reflective issues, I don’t think it’s the glass. Sitting in the car and driving around town the car seems very quiet. I never did learn how to use that mic and software.
 
Thanks for all the good info. The picture of the tire insulation looks like what I pictured, but the article I read maybe a year ago was about I think Continental tires which were going into some model of Tesla’s and the difference in noise improvement was more like 3 dB. Considering the for sure much higher cost, 1 dB or less hardly sounds worth it. 3 or 4 dB maybe. If anyone discovers more info p,ease post.
 
@Bob M it was a bit tricky to load the mic's calibration file, maybe they've made it easier now. The app has way more capabilities than we need. Maybe they've added a way to take successive snapshots. I'll try to capture some spectra in my car when I get a chance. One hard part has been annotating the captures, and in a moving car it's tough.

I'm very curious how different peoples' cars behave, maybe each car might have its own issues.
.
 
@Bob M it was a bit tricky to load the mic's calibration file, maybe they've made it easier now. The app has way more capabilities than we need. Maybe they've added a way to take successive snapshots. I'll try to capture some spectra in my car when I get a chance. One hard part has been annotating the captures, and in a moving car it's tough.

I'm very curious how different peoples' cars behave, maybe each car might have its own issues.
.

Tough to compare unless you are driving the cars on the same exact roads and comparing to the same exact spots since that is likely the single biggest variable. Someone driving predominantly on fresh black top with a high rubber content, might report their Tesla as being "silent"....
 
I need to enlist a passenger to help with capturing spectra, but here's baseline ambient noise in Park with heater off, and one I got today from recording at 70 MPH on an average California Freeway surface.

There's no higher frequency wind noise. The overwhelming impression is that it's all low frequency rumble, and that it would likely be passed from the tires through the suspension and to the body. The way any change in road surfacing is dramatically noticed in the cabin confirms that.

It's unlikely that any amount of patching with adhesive foam is going to do much. My guess would be that, in the absence of (fixable) air leaks, working on the suspension would be more effective, but I don't know anything about absorbing vibration in a suspension or acoustically decoupling it from the body. A model S, or a Lexus, probably has more expensive suspension parts.

M3RoadParked_2019-11-17-00-50-25-1x.jpg



M3RoadNoise70mph_2019-11-16-21-13-29-1x.jpg
 
  • Like
Reactions: chinnam3 and KenC
I need to enlist a passenger to help with capturing spectra, but here's baseline ambient noise in Park with heater off, and one I got today from recording at 70 MPH on an average California Freeway surface.

There's no higher frequency wind noise. The overwhelming impression is that it's all low frequency rumble, and that it would likely be passed from the tires through the suspension and to the body. The way any change in road surfacing is dramatically noticed in the cabin confirms that.

It's unlikely that any amount of patching with adhesive foam is going to do much. My guess would be that, in the absence of (fixable) air leaks, working on the suspension would be more effective, but I don't know anything about absorbing vibration in a suspension or acoustically decoupling it from the body. A model S, or a Lexus, probably has more expensive suspension parts.

View attachment 477900


View attachment 477901

Some folks have reported less noise with the aftermarket “comfort” coil overs from mountain pass (which makes logical sense to me). See post 365 here in this huge Thread: MPP Comfort Coilover reviews?

I have wondered if placing things like dynamat (the foil constrained rubber “tiles”), on surfaces closer to the shock/ strut towers might help slightly to quell the vibrations slower to the source. It’s not a lot of mass, but kind of hard to glue 50 lbs lead weights to places, so who knows...it might help...slightly.
 
  • Like
Reactions: polyphonic54
i have a Model 3 LR (32644) with the aero 18" wheels delivered June 2018 and cabin noise (mostly tire noise I believe) is my only complaint.
I don't have much more to add but am following this thread closely with great interest.
I would very much like to see comparisons of cabin noise levels between the earlier and more recent Model 3 builds to see if Tesla has made meaningful improvements.
I for one would be willing to consider paying for a retrofit that would meaningfully reduce the cabin noise. Maybe that could be an option when it is time to replace some of the suspension components?
 
Oh, I love my Model 3 and I don't find it noisy, but it's not dead quiet either, and perfection is a noble objective. It's frustrating how all the articles on quieting cars repeat the same old stories or promote some new (untested) $tuff.

There's a similar thing with electric guitars, where most everyone repeats the "obvious" truth, that adding shielding to the pickup cavities reduces interference and hum. I tried a before and after test, laboring with shielding tapes and sprays, expecting at least some improvement. But there was none. Zero. The reason being that it's low frequency electromagnetic noise that passes through (electrostatic) shields like a ghost through castle walls.

It seems intuitively true that if you'd goop down all surfaces with sticky rubber, they couldn't resonate. But are we sure how much that contributes? Should we buy $2,800 shocks on "maybe some improvement"? It's surprising that no university mechanical engineering department has rigorously tackled the issue, with measurements, for the public good : how best to inexpensively retrofit a car to minimize cabin noise. Very soft tires seems like a worthy point of attack, at some cost in energy efficiency. I mean, why add the rubber gummies after the fact?

As to time to replace suspension components, maybe they built it to last forever, at some cost in noise ;-)
.
 
Last edited:
Anyone looked into this?
https://www.amazon.com/Customized-F...574031597&sr=8-5#immersive-view_1574056203203

This for the front wheel well though. From earlier posts it seems the rear one is more important.
I wonder if that (it appears to cover the outside of the existing wheel well liners) or thinsulate on the back of the existing wheel well liners would have a bigger effect. This is cheaper than the cheapest roll of thinsulate on Amazon, though I've seen somewhere you can order less thinsulate on ebay for less than this, and I'm not sure how much you would actually need... Certainly this is more "plug and play".
 
Cotton is where in this sandwich? Cotton gets waterlogged and hosts fungi rather well.

The claimed 30 dB improvement heroically strains belief. On my spectra above, noise is 30 dB up from baseline. 30 dB is a lot. Hard to believe this contraption can mute any noise by 30 dB. That would take my 70 Mph rumble back down to my baseline level, which is unlikely when you consider that my baseline spectrum was taken in quiet ambient sound with the car idle in Park :)
.
 
  • Like
Reactions: UTMB