Custom speaker enclosure, as innert as possible

[mike7877]

Back around 1990, I bought a pair of Tannoy System 12 DMT monitors, which I still have. (I also specified and used the 8-inch model at work in a television station.) Tannoy appears to have done considerable research ahead of this design, which features top, bottom and side cabinet panels which are “floated” within damping material as a means of attaching them to an MDF space frame which connects a solid MDF baffle and rear panel.
View attachment 386519

What's an MDF space frame?

 

[DVDdoug]

"It's not impossible to judge a change without measurements."
I enabled the tone control and turned treble to +5.
"I can hear there's more treble now."
"No, you can't! Not without measurements!"

True, you don't always need measurements. But on the other hand, we tend to over-estimate our hearing ability and our auditory memory.

And blind ABX tests are often humbling.

Probably every mixing engineer can tell you of a time when they were making small EQ adjustments thinking they were making a change/improvement, only later to discover that the EQ wasn't in the signal path!

 

[mike7877]

Yes, completely agree. And measurements I have posted in the past support what you are saying.
Many of these alleged 'panel resonances' are actually resonances due to internal dimensions. Internal waves are mitigated by much different methods than bracing panels. I posted measurements, completely consistent with your experience. And also agree with you that fancy internal diffusion can mitigate these internal resonances:
View attachment 386479
It's a PITA, and this construction tames only one of the internal cabinet modes. That's a ton of work compared to just putting some polyfill and taming all three modes! And if not done right those diffusers will rattle and cause new problems that are potentially quite audible. Polyfill is very effective at taming these internal midrange resonances as I show, and much easier and less messy. That said, it is really disappointing to many people who want to believe that these exotic efforts are critical to good speaker building. They are not, a few good slabs of Baltic birch or MDF and moderate quantities of (very unsexy) polyfill are what is needed

We have an entire fifth-column of people on ASR obsessing on fancy cabinets (like Magico's enclosure absurdity). No measurements are ever provided. It's quite counter to ASR's methods and goals, and further confuses and misdirects people as to what is really important in speakers.

Comments like this need to be supported with actual measurements of the sound. Using ASR as a lunch pad for unscientific speculation, claims of exceptionally low distortion, etc. should be supported by evidence.

I've said in my OP that the improvement from bracing this cabinet is much bigger than the improvement I've gotten from bracing other cabinets. Much bigger. Most of the time it's a small change... In line with your experience. The box could have had problems to begin with, and the improvement made more obvious by the use of extremely quiet electronics, and drivers with extremely low THD.

I don't want to take apart the bracing of these speakers, parts are under compression, and I didn't use a torque wrench, so I may not be able to get the same result easily. I'm sure you'd agree that if I use different forces in different places, I may not attain the same effect I enjoy now?

 

[Salt]

You know my preference for exotic things, and so why not slate?:
https://fischer-fischer.de/en/

 

[mike7877]

True, you don't always need measurements. But on the other hand, we tend to over-estimate our hearing ability and our auditory memory.

And blind ABX tests are often humbling.

Probably every mixing engineer can tell you of a time when they were making small EQ adjustments thinking they were making a change/improvement, only later to discover that the EQ wasn't in the signal path!

Yes, I agree. Does me saying all the other bracings I've done have barely made a change not count for anything?
These boxes could've been particularly problematic. See my post 23 for why I don't want to dismantle

 

[MAB]

"It's not impossible to judge a change without measurements."
I enabled the tone control and turned treble to +5.
"I can hear there's more treble now."
"No, you can't! Not without measurements!"

This is a silly argument. Of course you can hear +5dB of treble. I didn't accuse you of being 5dB deaf Instead I am pointing out we all fooled by a 0.05dB change in actual performance, or even less. Who knows in your case? Nobody, since you didn't take measurements. Measuring the potential effect of things like cabinet bracing using your ears is useless. An impedance trace of the speaker is massively illuminating. Measurements of audible phenomena in speakers is what is done at ASR (and precious few other places). We do seem to get sidetracked in measuring inaudible phenomena (most of the DAC reviews for example).

I didn't make measurements before the bracing, and I'm certainly not going to remove it to prove something I already know to be the case...

Yes, it's called 'getting the carriage in front of the horses'. And I'm certainly not asking you to go back to square 1 and repeat.

If bracing didn't make an audible difference, cabinets wouldn't be braced.

Good grief, even more silly, belly rolling in laughter here. Same reason people make $50k cables, $10k cable lifters, and put $1k cedar boxes filled with dirt on their speakers. In a world free of measurements, manufacturers are free to spin BS about "proprietary cabinet bracing", and all sorts of exotic BS. This is how the industry works. The point made over and over at ASR. Do you read this website? Also, many manufactures use bracing since it is cheaper than making a massively strong and heavy cabinet. Good manufactures measure, figure out where to brace based on measurements so they can deliver an economically efficient design.

All I'm saying in my post is that all the parts are state of the art, and when something is near perfect, problems are more obvious

Particularly odd since you read datasheets, publish threads on them wondering ad infinitum what the measurements mean. Yet you don't do the simplest thing, measure something you have to see if there is a problem and if you actually changed the problem.

Does me saying all the other bracings I've done have barely made a change not count for anything?

It really counts for nothing. The expected changes based on measurements is small, actually tiny. Which is why it would have been nice to see some measurements.

 

[Salt]

Where, why and at what amount would I see enclosure-wall- ringing in impedance measurements?

 

[Salt]

I think we all can agree that bracing reduces wall ringing of enclosures and increases frequency of them, though attenuated. So not bad at all.
B&W marketed their "Matrix" in '80 and '90 as a revolutionary progress , what does not mean it's useless .

A long time ago I wanted a curved enclosure like Wilson's with a dead resonance, and it ended in double-slit-MDF-GFK-sandwich filled with micro iron spheres.
You can knock on them till the knuckles get bloody ...

 

[MAB]

Where, why and at what amount would I see enclosure-wall- ringing in impedance measurements?

In Erin's review of the JBL M2, in the section on woofer impedance measurements:

JBL M2 Master Reference Monitor

JBL M2 Master Reference Monitor

www.erinsaudiocorner.com

I performed the same when I built a pair of M2 Clones.

JBL M2 Clone

I hope you pull this off... I also hope you go the hardwood route or at least veneer the boxes. I positively love this aesthetic! Thanks for the positive vibes! Prolly omit the midranges given what I am learning. If hardwood, it would be oak-ply with solid end-pieces. Appreciated!

www.audiosciencereview.com

These resonances are quite small, but real. And inaudible.

I also looked at breakup mode of the surround, evident in the impedance trace, the distortion measurement, and the off-axis behavior of the sound field

JBL M2 Clone

I recently measured a JBL D2430K compression driver and M2 lens. My goal is to build an M2 clone. The speakers are going to go into individual boxes allowing me to stack the woofer and D2 into an M2, or even include a midrange. I got a pair of B&C 5NDL38 for cheap, and if they don't work I...

www.audiosciencereview.com

This was courtesy of some advice from one of the more knowledgeable ASR members. If I didn't measure, I wouldn't know.

JBL M2 Clone

I recently measured a JBL D2430K compression driver and M2 lens. My goal is to build an M2 clone. The speakers are going to go into individual boxes allowing me to stack the woofer and D2 into an M2, or even include a midrange. I got a pair of B&C 5NDL38 for cheap, and if they don't work I...

www.audiosciencereview.com

This resonance does affect the sound in a measurable way. Is fundamental to all drivers, and present to a greater or lesser extent on each element.

Erin used a DATS to measure the speaker resonance of the Tekton Troubadour (red arrow below), in order to show the effect (or lack of effect) of plugging the threaded holes for the feet in the cabinet, and prevent getting sued by a manufacturer having a meltdown (or whatever was going on).

And of course, the largest resonance in a loudspeaker is going to be the resonance of the system (circled above, two peaks since this is a ported alignment). All this is easily measured with impedance. In fact, when I build ported or passive radiator enclosures, impedance is the quickest and most accurate way to check the system tuning.

The point is, an impedance measurement can resolve cabinet resonances (and holes), resonances in the drivers like surround oscillations and cone-breakup modes, and even be used to check for discrepant drivers (voice coil rub, suspension differences between drivers, drivers with incorrect number of turns on the voice coil (Peavy drivers we were using had QC issues). And it can do this for resonances that mare way below human audibility (like these supposed cabinet resonance OP is claiming are audible).

 

[ThatSoundsGood]

I always do Constrained layer damping with mass loaded vinyl in between plywood and MDF along with bracing. My speakers are really heavy but they are really inert. I haven't used an accelerometer to actually test it but the difference is quite obvious.

 

[Mr. Widget]

I always do Constrained layer damping with mass loaded vinyl in between plywood and MDF along with bracing. My speakers are really heavy but they are really inert. I haven't used an accelerometer to actually test it but the difference is quite obvious.

I have built a number of medium to large sized speakers with 3/4" MDF as the core and then laminated 3/4" Baltic birch plywood on the outside with ample internal panel and cross bracing. For the bracing I use single or double thichness birch plywood. These cabinets are very inert. The MDF adds mass and is a fairly lossy material. The birch plywood is very stiff keeping the MDF from flexing. Birch plywood like concrete can ring on its own.

Adding a layer of mass loaded vinyl between them if done correctly so there is no chance of air trapping or delamination would be possibly even better.
A well braced stiff box of birch plywood with a significant amount of internal mass loaded vinyl would be my choice for a smaller enclosure.

 

[Salt]

In Erin's review of the JBL M2, in the section on woofer impedance measurements:

JBL M2 Master Reference Monitor

JBL M2 Master Reference Monitor

www.erinsaudiocorner.com

I performed the same when I built a pair of M2 Clones.

JBL M2 Clone

I hope you pull this off... I also hope you go the hardwood route or at least veneer the boxes. I positively love this aesthetic! Thanks for the positive vibes! Prolly omit the midranges given what I am learning. If hardwood, it would be oak-ply with solid end-pieces. Appreciated!

www.audiosciencereview.com

These resonances are quite small, but real. And inaudible.

I also looked at breakup mode of the surround, evident in the impedance trace, the distortion measurement, and the off-axis behavior of the sound field

JBL M2 Clone

I recently measured a JBL D2430K compression driver and M2 lens. My goal is to build an M2 clone. The speakers are going to go into individual boxes allowing me to stack the woofer and D2 into an M2, or even include a midrange. I got a pair of B&C 5NDL38 for cheap, and if they don't work I...

www.audiosciencereview.com

This was courtesy of some advice from one of the more knowledgeable ASR members. If I didn't measure, I wouldn't know.

JBL M2 Clone

I recently measured a JBL D2430K compression driver and M2 lens. My goal is to build an M2 clone. The speakers are going to go into individual boxes allowing me to stack the woofer and D2 into an M2, or even include a midrange. I got a pair of B&C 5NDL38 for cheap, and if they don't work I...

www.audiosciencereview.com

This resonance does affect the sound in a measurable way. Is fundamental to all drivers, and present to a greater or lesser extent on each element.

Erin used a DATS to measure the speaker resonance of the Tekton Troubadour (red arrow below), in order to show the effect (or lack of effect) of plugging the threaded holes for the feet in the cabinet, and prevent getting sued by a manufacturer having a meltdown (or whatever was going on).
View attachment 387036
And of course, the largest resonance in a loudspeaker is going to be the resonance of the system (circled above, two peaks since this is a ported alignment). All this is easily measured with impedance. In fact, when I build ported or passive radiator enclosures, impedance is the quickest and most accurate way to check the system tuning.

The point is, an impedance measurement can resolve cabinet resonances (and holes), resonances in the drivers like surround oscillations and cone-breakup modes, and even be used to check for discrepant drivers (voice coil rub, suspension differences between drivers, drivers with incorrect number of turns on the voice coil (Peavy drivers we were using had QC issues). And it can do this for resonances that mare way below human audibility (like these supposed cabinet resonance OP is claiming are audible).

I don't think there is a reliable proof any mini peak in impedance is audible (as proofed by Erin, btw.), but resonance is of course a target, an braces may help

 

[MAB]

I don't think there is a reliable proof any mini peak in impedance is audible (as proofed by Erin, btw.), but resonance is of course a target, an braces may help

Agreed.
Erin concluded it was inaudible. I agree based on my measurements and experience. Which is my point all along.

I used to construct massively overbuilt cabinets, one set was HDF/MDF sandwich with polyurethane epoxy. Interior bracing was also overbuilt. They measured and sounded no different compared to the cabinets they replaced, except for one artifact in the impedance measurement which turned out to be an internal reflection off of one of the braces. Not that it was audible, but I cut the brace away just to get rid of the resonance.

 

[Mr. Widget]

I don't think there is a reliable proof any mini peak in impedance is audible (as proofed by Erin, btw.), but resonance is of course a target, an braces may help

I think every case is different. The plot below was an impedance plot I made while trying to track down an audible and annoying bass resonance in a DIY project. The damped plot (red curve) was when I placed damping on the cabinet walls. The sound was improved, but still not acceptable. I ultimately fundamentally changed the design.

 

[MAB]

I think every case is different. The plot below was an impedance plot I made while trying to track down an audible and annoying bass resonance in a DIY project. The damped plot (red curve) was when I placed damping on the cabinet walls. The sound was improved, but still not acceptable. I ultimately fundamentally changed the design.


View attachment 387046

That's a very large impedance peak, I agree with you that will potentially be audible. And thanks for posting measurements so that we have something concrete to talk about! I also agree that each case needs study (measurements) to determine if audible, what course of action to take, and to confirm that the resonance is actually tamed. Like the resonant peak due to periodic breakup modes of a surround, that is an audible phenomenon. Or the holes in the Tekton Troubadour that the designer claimed to produce audible hypersonic resonances, which was totally made up.

 

[Salt]

I think every case is different. The plot below was an impedance plot I made while trying to track down an audible and annoying bass resonance in a DIY project. The damped plot (red curve) was when I placed damping on the cabinet walls. The sound was improved, but still not acceptable. I ultimately fundamentally changed the design.


View attachment 387046

Could you work out it was an internal resonance or vibrating of the enclosure?

 

[Mr. Widget]

Could you work out it was an internal resonance or vibrating of the enclosure?

This was almost 20 years ago so I don't remember all of the details, but I performed a series of tests and experiments and determined that the enclosure walls were the problem. I re-designed them using much denser and thicker material and that solved the problem.

 

[DavidMcRoy]

What's an MDF space frame?

From Oxford Languages: A space frame is "a three-dimensional structural framework which is designed to behave as an integral unit and to withstand loads applied at any point."

Here is a link to a discussion with photos of the interior construction of a Tannoy DMT cabinet:

Tannoy System 15DMT II

I've been a fan of the 15" tulip Dual Concentric since building a pair of 215 clones 10 years ago. It is an excellent driver and I have wanted to try some Tannoy built MkII System 15's for a while now. The thread is more about taking a peek inside and seeing how they are put together. I've...

pinkfishmedia.net

A discussion of Tannoy DMT can be found at this link, under "Cabinet" beginning at the bottom of Page 12:

https://www.hilberink.nl/tannoy2/dmtseries/dmtseries1.pdf

 

[Holmz]

…
In that vein, I've been thinking... instead of a rubber CLD, how well would a 1/4" thick rubber enclosure contained within a 1.5" thick concrete box with" sand as a buffer in between work? Like 0.5 to 1" sand. And for when the woofer moves outward, the vacuum created within is countered (on the rubber) by a lattice of thick steel wire spaced horizontally and vertically, equidistally 1.5" (apart), and present wherever possible (ie. When it can cross from one side to the other, both sides rubber only - if the driver or crossover is in the way on one or both sides, not there!)
…

What makes you think that 1/4 is better than 0.5mm?
The pupropse of the rubber to be in shear, and when it is thick then the gradient is lower than when it is thin.

 

[kemmler3D]

Good grief, even more silly, belly rolling in laughter here. Same reason people make $50k cables, $10k cable lifters, and put $1k cedar boxes filled with dirt on their speakers. In a world free of measurements, manufacturers are free to spin BS about "proprietary cabinet bracing", and all sorts of exotic BS.

I'm just going to say that in my time on DIYAudio.com, I've seen a fair number of before / after measurements of bracing and it tends to be pretty visible in measurements. Anecdotally a lot of people claim the changes are audible. And it's very uncontroversial in the speaker-building world that bracing is necessary to prevent unwanted resonances.

I don't think I've ever seen blind tests of braced vs. unbraced speaker cabinets... but at the end of the day the measured differences are often big enough to be clearly audible in theory. I don't think it's silly to do that kind of work on cabinets "just in case", certainly more sensible than upgrading from a 90 to 110db SINAD DAC.