Yes, it is one of the primary challenges to designing a high performance floor standing speaker. It is very challenging to design a large cabinet which has a low vibration signature from 50 Hz to 2 kHz, AND meets all the other requirements for the target baffle step, low high frequency diffraction, and pleasing aesthetics. This is a challenge for both tall slim tower-type speakers and the more traditional big box shape (such as the Klipsch Cornwall)... either way, there is a lot of cabinet surface area, multiple sources of vibration energy (drivers), and multiple holes in the baffle which reduce stiffness.
Separating the woofer cabinet from the midrange cabinet is my preferred approach when striving for the highest performance. A small compact ~ 5 - 12 liter box for the midrange is much easier to design and build, knowing it will be exposed to energy in the 200 - 3000 Hz range. Designing a large woofer cabinet is easier if I know from the start that the drivers will start rolling off by 200 - 500 Hz, and especially if there is just one driver and one hole.
In the shoes of the OP I would indeed put the green sheet of damping on every side so it is sandwiched indeed. Mass spring mass. The most rigid panels should be inside as they are structural and one needs it to be stiff. Outside I would just -after the green material use 10 mm of something that damp more than the inside wood : multyply poplar, something that damps ! Of course not ton the front where the metal front stands on the green material.
I wonder how all those good material are spurced : very impressive purchasing sourcing effort aside the design. I didn't find the bulk front aluminium, expensive as well.
@augerpro have made good research about cabinets structure.
Bravo !
These large format box speakers actually have an advantage over the slimmer floor standers, in that the larger panels can be pushed into lower, more benign resonance modes outside of the critical midrange.
Mass loading the larger wall surface area and leaving some compliance on the edges can result in a highly decoupled surface, which (despite lack of rigidity compared to smaller panels) will barely react in movement with midrange frequencies. Its sort of like crossing a driver with a sharp breakup peak much lower to avoid the breakup rather than dampening it with brute force. Leaving the enclosure panel outer edges with a softer, more compliant perimeter will disrupt the transmission of higher frequency modes throughout the other enclosure panels.
In the case of the OP, the dissimilar materials used in the enclosure will allow some decoupling of side and rear panels through extra stiffness and mass provided by the front baffle and other sections of billet aluminum. Pushing the panel resonance modes very high can be just as effective as pushing them lower.
Both strategies will work, but the higher expense of using very stiff, custom machined materials is prohibitive for most diy builders. The mass decoupled panel construction technique is equally effective and will result in an equally well performing inert sounding enclosure. You may lose a little bit of LF sensitivity but its trivial, being a worthwhile compromise to end up with the best possible midrange performance.
My comments were meant to be generally applicable to typical construction. This project is anything but typical. I expect the thick aluminum baffle to have a first mode resonance at a very high frequency compared to a plywood/MDF baffle. The 10 threaded rods that connect the aluminum baffle to the rear wall will enable fine tuning of the preload, and this will allow @PSchut to tune the other cabinet resonances to the minimal signature.
@hifijim I totally understand your comments. The reason I addressed the larger woofer 3 way box concept was due to the OP thread title, building the world's best 3 way. Obviously, with the manufacturing processes the OP has access to allows him to adapt the design for the best performance given the components he's using.
In my world of large 3 way monitors, I always try to improve what's already there, adapting existing designs with reasonably cost effective solutions and pushing performance further than what was thought to be possible.
There is obviously more than one way to skin a cat (just an old school saying... I like cats, as long as they leave my speakers alone...lol). I don’t think there will be issues with HF resonances on the OPs aluminum baffle components. The way they're being clamped and fastened leaves little possibility for anything to ring or resonate. The design is very well thought out and it should perform to his expectations.
In my world of large 3 way monitors, I always try to improve what's already there, adapting existing designs with reasonably cost effective solutions and pushing performance further than what was thought to be possible.
There is obviously more than one way to skin a cat (just an old school saying... I like cats, as long as they leave my speakers alone...lol). I don’t think there will be issues with HF resonances on the OPs aluminum baffle components. The way they're being clamped and fastened leaves little possibility for anything to ring or resonate. The design is very well thought out and it should perform to his expectations.
I am not sure about that profiguy, although there was a small floor stander from Belgian manufacturer Synthese called “floating” where a Dynaudio Esotar D330 was used in a separate tube enclosure decoupled from the main enclosure. A bolt kept it in place during transport and the sound was ruined when not unscrewed. So there is something. But I think that a speaker with some moving mass should be mounted ridged to a heavy as possible enclosure. The only thing that should move is the cone.
I think the speaker we are building will weight more than 100 kg. My current larger speaker sits around 150 kg with its 65 mm thick Finsa Compac Plus baffle. That is why I decided to use 40mm and 20mm solid aluminum for front and back on this smaler enclosure to get as much mass as possible as that is to me the way to get a clean neutral and dynamic end result.
I think the speaker we are building will weight more than 100 kg. My current larger speaker sits around 150 kg with its 65 mm thick Finsa Compac Plus baffle. That is why I decided to use 40mm and 20mm solid aluminum for front and back on this smaler enclosure to get as much mass as possible as that is to me the way to get a clean neutral and dynamic end result.
One could argue about that, but I’ll leave it 😉. I remember KEF supplying the suspension stuff profiguy mentioned with their constructors series. And methinks it’s quite logical decoupling is beneficial. There is no decent car out on the roads without it.
Hi Markbakk, I think I agree with you (as my experience with the Synthese Floating) when there is a compromise on mass of the speaker. The moving mass of one woofer I use is 30 gram, so 60 gram in total. I don't know the formula but when I ran in these speakers in free air doing a mild sinde wave of 50Hz at a few volt you can feel the pulsating force in opposite direction of the cone movement, and this is a 6 kg speaker. so if this ends up in a 30kg enclosure I do not think it will totally absorb that energy. And of course mass also helps panel resonances by the air inside the enclosure.
If a car had indefinite mass, the suspension would be extremely smooth as the mass wouldn't move on bumps.
I am just applying plain logic when I design stuff. I have an electronic background (worked for Kharma, STC/Siltech/Sphinx, Sonic Frontiers, Classe Audio in a previous life) and when I designed layouts for amps and cd players next to clear electronic guidelines I used logic on what trace carried what and how ground planes were poured. I tend to do that with mechanics in these speakers as well. For example; If all else is equal I think a heaver enclosure will sound better. My friend Joost, who builds the same set together with me challenges me constant on these things and if I can't explain we adapt.
The journey is at least half the fun.
Peter
If a car had indefinite mass, the suspension would be extremely smooth as the mass wouldn't move on bumps.
I am just applying plain logic when I design stuff. I have an electronic background (worked for Kharma, STC/Siltech/Sphinx, Sonic Frontiers, Classe Audio in a previous life) and when I designed layouts for amps and cd players next to clear electronic guidelines I used logic on what trace carried what and how ground planes were poured. I tend to do that with mechanics in these speakers as well. For example; If all else is equal I think a heaver enclosure will sound better. My friend Joost, who builds the same set together with me challenges me constant on these things and if I can't explain we adapt.
The journey is at least half the fun.
Peter
Driver attahcment to the frame or box is a mystery me. Elastic with adhesive tape would decouple, but securing it with bolts or screws will screw it with extra resonances...
I use a thin window seal tape and tight screws on MFD or plywood.
I've read about KEF studies mentioned above but I don't know about the small details. From personal experience I know that sometimes resonance will be heard and more often measurend as high 3rd harmonic distortion. Panel resonance with any decent hifi speaker are impossible to measure to me (no accelerometer). Driver nearfies decay tells mostly cabinet modes.
I made extra bracing on a JBL LSR305 and recorded knuckle knocks and analyzed them with Audacity. https://www.diyaudio.com/community/threads/jbl-lsr305-tweaking.320206/
I use a thin window seal tape and tight screws on MFD or plywood.
I've read about KEF studies mentioned above but I don't know about the small details. From personal experience I know that sometimes resonance will be heard and more often measurend as high 3rd harmonic distortion. Panel resonance with any decent hifi speaker are impossible to measure to me (no accelerometer). Driver nearfies decay tells mostly cabinet modes.
I made extra bracing on a JBL LSR305 and recorded knuckle knocks and analyzed them with Audacity. https://www.diyaudio.com/community/threads/jbl-lsr305-tweaking.320206/
Midranges and tweeters aren't heavy, but woofers are 5-15kg... Saggging elastics would be a nightmare
https://audioxpress.com/article/speaker-design-driver-induced-vibrations
https://assets.kef.com/documents/tseries/t_tech_explained.pdf
https://www.audiosciencereview.com/forum/index.php?threads/kef-lsx-teardown-and-repair.55697/
https://www.linkwitzlab.com/Driver Decoupling.doc
https://patents.google.com/patent/US20140348369A1/en
https://www.rmsacoustics.nl/papers/whitepaperdistortion.pdf
https://audioxpress.com/article/speaker-design-driver-induced-vibrations
https://assets.kef.com/documents/tseries/t_tech_explained.pdf
https://www.audiosciencereview.com/forum/index.php?threads/kef-lsx-teardown-and-repair.55697/
https://www.linkwitzlab.com/Driver Decoupling.doc
https://patents.google.com/patent/US20140348369A1/en
https://www.rmsacoustics.nl/papers/whitepaperdistortion.pdf
Last edited:
Thanks for the links, interesting reading.
What was common in all tests that showed improvement was a thin sheet of material the driver was mounted on. Linkwitz used 15mm mdf that is a very thin sheet to build a speaker from.
Our inner enclosure made out of 15mm MDF heavily braced with these 13 mm panels does not sound rigid with a knuckle test. But when we add a complete layer of 13mm Finsa Compac plus it will be a big change in rigidity. But to be sure another layer is 20mm solid walnut will top it of.
This is by no means scientifically proven but it worked extremely well the last time.
The only downside is you need 4 strong man to move the speaker.
I truly believe that panel resonances will kill the the sound.
And it seems to me that driver decoupling works best on light enclosures.
What was common in all tests that showed improvement was a thin sheet of material the driver was mounted on. Linkwitz used 15mm mdf that is a very thin sheet to build a speaker from.
Our inner enclosure made out of 15mm MDF heavily braced with these 13 mm panels does not sound rigid with a knuckle test. But when we add a complete layer of 13mm Finsa Compac plus it will be a big change in rigidity. But to be sure another layer is 20mm solid walnut will top it of.
This is by no means scientifically proven but it worked extremely well the last time.
The only downside is you need 4 strong man to move the speaker.
I truly believe that panel resonances will kill the the sound.
And it seems to me that driver decoupling works best on light enclosures.
I think your choice for more mass is valid.
https://www.stereophile.com/content/monitor-audio-gold-300-6g-loudspeaker-bolt-tension-optimization
DUT has 4 bolts from woofer mounting to the back panel. Listeners found improvement in sound from tightening only 45deg instead of 90deg...
As noted in the main text, we experimented and found that a looser setting—45°, or an eighth not a full quarter turn—resulted in dramatic improvement. The excellence that the Gold 300 was hinting at was now realized. The stereo image was more convincingly grounded. The mild leanness of timbre we'd heard was supplanted by a densely textured foundation. The orchestra was now fully present in the hall, with spacious imaging and fine ambience. The soundstage was better focused, and microdetail and expressiveness were clearly improved. Image depth was rather better, and that elusive quality of natural spatiality was clearly established. Replay was flowing and imbued with richer timbre. Pop and jazz rhythms were more rewarding.
JA found worst vibrations on sidewalls. All measurements were excellent, I believe that rising treble on-axis is intentional, based on listening.
DUT has 4 bolts from woofer mounting to the back panel. Listeners found improvement in sound from tightening only 45deg instead of 90deg...
As noted in the main text, we experimented and found that a looser setting—45°, or an eighth not a full quarter turn—resulted in dramatic improvement. The excellence that the Gold 300 was hinting at was now realized. The stereo image was more convincingly grounded. The mild leanness of timbre we'd heard was supplanted by a densely textured foundation. The orchestra was now fully present in the hall, with spacious imaging and fine ambience. The soundstage was better focused, and microdetail and expressiveness were clearly improved. Image depth was rather better, and that elusive quality of natural spatiality was clearly established. Replay was flowing and imbued with richer timbre. Pop and jazz rhythms were more rewarding.
JA found worst vibrations on sidewalls. All measurements were excellent, I believe that rising treble on-axis is intentional, based on listening.
Hi Juhazi, this was a similar thing in the earlier mentioned Synthese Floating but again this is 20kg speaker. Interesting concept though.
Glueing bitumen pads on internal walls is IMO not worth it. Is OK to dampening small panels.
Polyester wool (or pillow stuffing) main quality is that it is cheap. Same with MDF.
I much prefer to use real wool. Basotect is also nice. You need a lot less than what you think.
It would be interesting to see a sim what actually happens when using thick felt vs none. I like to do the same as Gravesen. Line internal walls with felt even for bass.
Hi Rokytheman.
I agree on the bitumen pads. On the internal walls separating mid enclosure from LF enclosure I used only one layer of this Finsa sheet. That felt a bit underwhelming compared to the massive sides and back. So we added a little mdf part and on both sides bitumen pads and then these little panels felt a lot more solid (see picture above). Overkill? I don’t know but it won’t hurt.
The felt on the inside walls do a nice job in an absorbing first reflections. The long hair wool for mid and akotherm for LF will do the rest. See the akotherm graph above and it has an impressive absorption effect.
We will use D20/100
I agree on the bitumen pads. On the internal walls separating mid enclosure from LF enclosure I used only one layer of this Finsa sheet. That felt a bit underwhelming compared to the massive sides and back. So we added a little mdf part and on both sides bitumen pads and then these little panels felt a lot more solid (see picture above). Overkill? I don’t know but it won’t hurt.
The felt on the inside walls do a nice job in an absorbing first reflections. The long hair wool for mid and akotherm for LF will do the rest. See the akotherm graph above and it has an impressive absorption effect.
We will use D20/100
A small update. had only one weekend to work on it.
This is a setup of the base.
The feet stay the same (plain Stainless Steel) and the base plate will become black.
We are testing if a powder coating gives a nice result. If not anodizing is the way to go but needs additional pearl blasting.
I sanded all walls to make them even with some panels sticking a hair out as we need a flat surface for the next layer of Finsa Compac Plus.
Without the front panel there is a comfortable space to apply the felt layer. Gluing it to all surfaces and only leave the sides free of felt where it could interfere with the metal rods.
Using the metal rods as temporarily clamps while gluing the inside front panel.
Like mentioned, next the second layer of the inner enclosure.
Cheres, Peter
This is a setup of the base.
The feet stay the same (plain Stainless Steel) and the base plate will become black.
We are testing if a powder coating gives a nice result. If not anodizing is the way to go but needs additional pearl blasting.
I sanded all walls to make them even with some panels sticking a hair out as we need a flat surface for the next layer of Finsa Compac Plus.
Without the front panel there is a comfortable space to apply the felt layer. Gluing it to all surfaces and only leave the sides free of felt where it could interfere with the metal rods.
Using the metal rods as temporarily clamps while gluing the inside front panel.
Like mentioned, next the second layer of the inner enclosure.
Cheres, Peter
did you ever make a listening test with and without damping the walls in a subwoofer?
I am curious if damping in a subwoofer crossed low at maybe 100hz helps the sound or not.
I am damping, too, just to be secure but did not do thorough listening tests with and without damping.
Some say in a subwoofer damping might even be contraproductive because of unnecessary losses.