Quieting the Baratza Sette - 4 reversible mods tried & tested

Grinders are one of the keys to exceptional espresso. Discuss them here.
Posts: 3
Joined: 3 years ago

#1: Post by davee5 »

-- TL;DR --
I made a few reversible modifications to my Baratza Sette and considerably knocked the edge off the noise. It's still pretty loud, definitely not quiet, but the noise is much more tolerable.

I also tried to record my process and test results so ya'll can do it too. The modifications are:
- seal the vents
- install hard drive mounting+damping grommets under the grinder/motor unit
- partially enclose the grinder/motor unit in mass-loaded-vinyl (MLV) backed with foam (CCF)
- mount Constrained Layer Damper (CLD) tiles on the enclosure walls

Total cost $30-40, shipping inclusive + a few hours of hacking.
All mods are reversible and should not damage your machine.

Results video here:
Spectrographs showing the mod effects on frequencies and intensities are shown under "Results" at the end and in follow-up post #3. (I had them inline but the images were tooooo big.)

-- Introduction --
Hello HB forum! Long time lurker / grateful learner, first time poster! Here's hoping my first material contribution proveds useful to the community that's long been a solid resource for me.

I recently picked up a Baratza Sette 270Wi refurb for a song, both to upgrade my aging Baratza Virtuoso and in anticipation of a significant upgrade to a prosumer espresso machine. I had read everything I could and it came down to the Sette or the Mignon Specialita for me. I really wanted to like the Sette because I think the design is extremely elegant -- low retention, ease of cleaning, grind speed, quick and simple adjustment -- and because the built in Acaia scale is rad. I also have kept my Virtuoso running with several rounds of parts and service bulletins from Baratza. Some people think having to repair a unit is an indictment of quality but I personally love their focus on repairability, enabling end customers to maintain their units indefinitely. Stuff breaks, its great that these guys are ready to support you. Plus snagging a refurb makes the already low-cost Sette half the cost of the Mignon, so the only real elephant remaining in the room was the noise issue.

Ah, the Sette noise issue. There has been so much noise around the noise that my partner and I ordered the refurb and were fully prepared to send it back within the 30-day return window. We're on week two and it is indeed quite loud, but it's not as earsplitting as we had braced for. Still, surely something can be done to improve the noise issues, right?

There's got to be a better way! And here I am at home with the kids during a pandemic, and I have these engineering degrees just laying around, so... let's get to modding!

-- Objectives --
To my ear the Sette has 2 main sonic issues to address:
- it's too loud in overall volume (sound pressure / decibels)
- the timbre of the grinder is harsh and high

So I set out to see if I could address these issues with a few additional project goals:
a. material costs should be very low (under $50)
b. grinder appearance and operation should remain "stock"
c. all modifications should be reversible

-- Research and Plan --
By day I am a mechanical engineer and industrial designer with 20 years of experience developing consumer electronic gadgets. I already know a few things about how to design and assemble products, how to reduce noise issues, how to manage heat, etc. Yet research is always warranted when starting a project, so I also supplemented my knowledge by reading up on noise reduction in hi-fi audio forums, especially the weird world of competitive car audio. There were 2 resources in particular that were helpful:
- Sound Deadener Showdown, a now defunct website on how to make your car interior whisper quiet)
- DIYMobileAudio, a forum that includes enthusiasts who rigorously test sound deadening materials and techniques.

In the end I identified 4 steps I thought would help tame the Sette noise while meeting the core objectives of cheap, internal, and reversible.
1. Block off the big air vents in the case.
2. Improve the mount damping between the grinder and the case
3. Block and absorb sound with internal baffles
4. Damp case vibrations with sound deadening tiles

-- Process & Testing --
To determine if each step is helpful I wanted to measure results against my objectives. Ideally I'd be using quality test equipment that allowed me to make a change and then measure the impacts on sound level and quality. Unfortunately I don't have a sound pressure meter (decibel reader) so I can't provide hard numbers on how much each step changes things. Sorry. But I do have a half-decent external microphone (Blue Yeti) and some basic sound processing software (Sonic Visualizer) so I could assess changes in frequencies and intensities with each change.

After each change I photographed each step of installation, reassembled the unit, recorded the sound of the grinder running open (no beans were harmed during the making of these tests 'cuz we are running low), processed the spectrographs, and took notes to share with the general grinder public. Changes were made sequentially, so the results are not strict A:B, A:C, A:D tests versus stock; more like A:B, A:B+C, A:B+C+D.

The following steps are outlined in order of reversibility. If you want to try this at home and you're concerned about being able to go back to the stock configuration easily, go roughly in this order. Only the last step here will be a pain to undo.

Step 0 - Gather your Supplies and set your Baseline
Material sources are listed at each step below.

Step 1 - Block the air gaps.
The Sette has 2 sets of air vents, one on bottom and one in back of the device. I'm personally wagering that these were intended to function as heat vents but are basically useless in practice.

In my professional experience the installation of vents on any machines that do not use true active cooling (i.e. using fans to force air through a heat sink) is basically window dressing and wishful thinking. If you're not in a forced air environment usually what happens is the entire enclosure heats up slowly and heat dissipation becomes a matter of surface area radiating heat away from hot things and into the environment.

The Sette vent design in particular is unlikely to be useful. The probability that air is going to (a) enter under the foot, then (b) effectively pass over the board and motor (the hot things) and then, (c) vent out the rear where the opening is facing down instead of up with the now hopefully warm and buoyant air is basically nil on all 3 counts, let alone together.

In practice the Sette heat path starts at the motor and grinder -> the air inside the unit -> the housing / enclosure -> the ambient environment. Steady-state will take a long time to reach, and passive, wishful air venting is unlikely to substantially change that process. Since the Sette is meant for intermittent use anyhow, it's exceedingly unlikely that you're going to cause damage here, the heat was just going to stay in the motor body anyhow and it still will.

However, despite being unlikely to keep your grinder internals cool, these vents are *excellent* at letting sound out. So if you cover these up with something relatively heavy and thick and you'll immediately notice an improvement.

- Gaffer tape or equivalent, ~$0.05
- [or CLD tiles, see Step 4]

Time required: 30 seconds.
Just tape over the vents! I found that I had the best results with a few layers of tape, 4-5+ of gaffer tape sounded better than 1-2. Covering the bottom vent matters less than the top back vent, but just do both and be done with it.

5 layers of gaffer tape over all vents makes a notable reduction in mid-frequency sound, it is clearly more pleasant and less harsh. Spectrum analysis indicates a sharp reductions at 530Hz, 310Hz, and light overall reduction. It does seem to slightly increase intensity of to 900-100Hz range, but it doesn't sound any more high pitched. This is also the only mod you can A:B test instantaneously with your own ears.

Step 2 - Vibrationally isolate the grinding unit.
For all the flak Baratza's engineering team gets around using too much plastic for cost reduction, the Sette is cleverly designed (with Etzinger?) to keep all the moving parts as close together as possible. The motor/grinder/adjustment assembly is functionally one single monolithic unit, keeping everything nice and close so it's reasonably rigid without resorting to huge metal enclosures. That business end is mounted to the case by 4 screws on the back of the motor mount. Those mounts, however, leave a lot to be desired in terms of vibration isolation.

The motor mounts have thin rubber bushings surrounding a plastic screw barrel, which do have some give but don't seem especially compliant. In my unit, refurbished a few weeks ago, there were also 2 loops of felt under the grind unit that I presume are intended to act as dampers between the grinder and the case. I have not seen these parts in any of the Sette service bulletins or part diagrams, but I did see them in user generated teardown videos starting around mid-2020. That leads me to believe the felt loops are band-aid fixes added in later to try to reduce noise. Felt does have noise absorbing and vibration isolating uses, but I suspect it's pretty fussy in practice.

My professional experience is with more engineered damping materials, notably viscoelastic damping elastomers like those used in shock mounts for optics and hard drives. Luckily for us, the screw sizes, boss sizes, and offsets for the motor mounts are pretty close to those used in hard drive snubber mounts and grommets. Ideally use name brand isolation discs with crush ribs from a company like E-A-R Isodamp as that's the good stuff. Here I admit to just sourcing for convenience instead of pure performance. People who like spec sheets and part sourcing can wade through the Isodamp lineup and MOQ management can do so via their online supplier. I'd suggest going for the softest stuff you can get, which are the blue E-A-R grommets. The source below on Amazon is purported to be the real deal, (part number G-410-1).

- Akust HDD Anti Vibration Screw kit (6-32 size) 8 pcs blue. $6 from amazon.
- The full lineup of Isodamp grommets. $0.28 per piece, MOQs and shipping notwithstanding.

Time required: 10 minutes.
Open up the case per Baratza instructions and remove the grind motor unit. Take the 8 mounts in the kit and cut 2 in half bagel-style. I placed the 6 intact dampers loosely between grinder body and housing, 3 on each side inside of the felt ring. Place the half dampers under each screw mount, inside the counterbore. Reassemble.

The screw thread engagement was reduced to 2.5-3 turns, but that should be acceptable for thread engagement. The grind unit also sits a touch higher and is mounted noticeably wobblier, making the adjustment cone slightly fussier to install, but it's not problematic. That wobbliness also indicates that the motor unit is no longer not coupled as rigidly, which is good. Thre is no impact on hopper clearance and fit.

(Note there's a chance than using fewer grommets under the grinder / body could work better as it might reduce the vibration transfer I could also convince myself that more damping surface area is better too. But I bet that's in the weeds, this isn't magic.)

Damping snubbers knocked down 200Hz significantly, general reduction from 300Hz-400Hz as well as slight reduction from 900-1000Hz, created a new resonance at 490Hz but overall the noise is more concentrated. Imperceptible reduction in volume, more of a decrease in harshness.

Step 3 - Block and absorb sound inside the unit.
The car audio folks all say the best thing you can do to quiet a vehicle and isolate road noise is to install MLV. Mass Loaded Vinyl is limp rubber that has been filled with metal powder, making it both flexible to absorb and not reflect sound, but also heavy enough to eat up some of that energy. To work properly it needs to move freely, so it can't be in hard, constraining contact with vibrating surfaces. Car folks use closed cell foam sheets (think yoga mat) to allow the MLV to move freely at acoustic levels of vibration, while also being mountable at human scales. The foam decouples the MLV from whatever its mounted to. Note that mass and thickness matter here, car sound proofing enthusiasts say "1lb" rated MLV or heavier.

- Noise Grabber 1lb mass loaded vinyl 1'x1' sheets at Amazon - $7.
- thin closed cell foam sheet (I had some Outdoor Wilderness Fabrics FY20 1/8" hanging around) - $2 or less

Time required: 1-4 hours, depending on your enclosure design intricacy and tuning.
The more coverage the better, so the goal here is to build an acoustically an isolated MLV sound cage between the grinder and the case. Granted there's not a ton of room in there, but after cutting a handful of prototypes in construction paper I ended up making half a box with a hole in the "lid" for the grinder / hopper port, as well as a tunnel for the motor. Be careful around the hopper detection switch. Reassembly was a bit of a chore as my MLV+foam are in interference contact with the case.

Spectrogram indicates that most major resonances have been eliminated. Significant intensity reductions across the board, notably in the mid and low ranges. The sound is definitely quieter and more diffuse. This was a big step forward.

I will note here, without supporting diagrams, that I first tried MLV without foam, thinking it would be fine. It was not. MLV without foam protection at the contact points merely transferred vibration and created new harmonics. MLV alone, without CCF, turns out to be a step backwards. However, once I redid it with foam wrapping the MLV sheet it worked as intended.

Step 4 - Reduce case vibration and resonance.
The other material that car audio folks love are CLD tiles. Constrained Layer Damper tiles are basically viscoelastic materials bonded to a thin sheet of metal; the best are super soft butyl rubber bonded to dead-soft aluminum sheet. The butyl is adhered to the walls in question enabling the rubber to damp vibrations between the vibrating panels and the dead weight mass.

The car audio forums really, really love to test this stuff so it's easy to find recommendations discussing the finer spectral response curves between Dynamat, Second Skin, etc. But the standout for sourcing purposes is Kno Knoise Kolossus by KnuKoncepts.
- most recommended product on DIYmobileaudio.com, following extensive testing
- superior damping, good heat resistance, and easily obtainable.
- can be bought in small sheets for small dollars, rather than in sizes made for complete cars for many hundreds.

However, the butyl in this stuff is an effective damper because it is super gooey and sticky. Whatever plasticizer or compound they're using is also a bit stinky, not noxious, but a whole sheet has a smell for sure. Perhaps CLD tiles are not great for aromatically sensitive applications, like coffee grinders. I pinged KnuKoncepts and the company claims it has no MSDS for review, so they punted with a lawyer-approved "we do not recommend our products for use in food processing equipment." Fine, I get it.

Regardless, butyl cleans up OK after application so this is technically still reversible in the case that it causes perceptible issues with flavor.

- Kno Knoise Kolossus Edition single sheet, 1.75 sqft. $9.50, $16 shipped from KnuKoncepts

Time required: 10 minutes.
Minimum suggested coverage from car folks is 30%, and any additional coverage will continue to help with damping. This stuff is cheap enough that any easily accessible spot where it will fit should be covered. Still, I decided to keep it away from the coffee-end of the machine for now.

I used beater scissors to cut tiles that would fit between ribs in the case. I put CLD on most of the rear housing from the vent down, using one large tile to act as my vent cover. I also put some CLD tiles on the side walls of the housing, flanking the power board and scale / portafilter holder internals. I generally left enough room around each tile to provide access to tools for removal putty knife, guitar pick, etc) in case I want to undo this mod. I also put foam over the tile the covers the vents to ensure the aluminum sheet would never make contact with the motor wiring. Note that this uses about 10% of the sheet, you just can't reach that much area easily.

Strips out a bunch of the mid-highs and highs and pushes the dominant note to the low end at 200Hz. Huge reductions from 300-800Hz. Reduction in overall volume, but not a ton. The spectrogram indicates that it concentrates the high end, but it seems muted and not very harsh. In bang-for-your-buck in time and effort this seems like a clear winner -- excepting those pesky questions around butyl...

Clear reduction in noise, both less harsh and less of volume. Mild success!
That said, this grinder is still nowhere near "quiet." This is not a series of experiments that results in "wow, that's quiet!" or even "hey, that's pretty reasonable." It's now merely "quite tolerable." Conversations will not happen over this thing, but your occasional hangover headache will definitely appreciate the reduction in noise and harsh high-end. It sounds like a better built machine and it is less maddening, but it will still wake up your baby. Still, for $30 out of pocket and a few hours of tinkering, this was a highly satisfying result.

Here's a video that walks through the mods and shows the spectrographs on screen. Mind you I'm a product designer, not a video editor or sound guy, but I hope its helpful.


The step-wise changes in spectrographs are as follows:


+ Vents sealed

+ HDD damping snubbers

+ MLV/CCF sound cage

+CLD Tile panel resonance dampers

Be well and do good!

:: Dave

Posts: 1115
Joined: 14 years ago

#2: Post by boren »

Amazing stuff and a lot to digest. Thanks so much for investing the time to document and share this. I'm very tempted to try one of these hacks.

davee5 (original poster)
Posts: 3
Joined: 3 years ago

#3: Post by davee5 (original poster) »

Thanks! I'm indebted for all the hard work everyone else has put in on this forum, so I figured it's time for me to contribute in kind.

I realized that the spectrographs are hard to compare as posted (first time posting, so my image formatting skills are marginal). I put together a side-by-side to show the sequential changes in sound characteristics.

The frequency range here runs from 140Hz to 1200Hz, which is about the full width of the real sound coming off the Sette in my recordings. There are some transients at higher frequencies, but not much of the noise is that high.

Draw your own conclusions about what these changes in signature mean, but the net result is clearly a general push towards the low end and a roll-off of the harsher, higher frequencies. I find it much more pleasant.

User avatar
Team HB
Posts: 5012
Joined: 18 years ago

#4: Post by RapidCoffee »

Amazing first post! Keep 'em coming. 8)

User avatar
Team HB
Posts: 6890
Joined: 19 years ago

#5: Post by Jeff »

Always a refreshing change to see objective data to support perceptions