Dual tuned panel resonator?

[ToBeFrank]

I am planning to build a 17.25" tall panel resonator. It will be built as a box with all sides enclosed and will be placed on the floor against a wall. I have two different panels picked out for the face. One is 1/8" hardboard with .71 PSF. The second is 3/16" hardboard with .9 PSF. These calculate to resonant frequencies of 49Hz and 43Hz respectively. I also plan to put fiberglass insulation inside the box. Since this will be placed on the floor, do I still need to have a stiff back panel or can I use the 1/8" on one side and the 3/16" on the other side? This would allow me to change the resonant frequency by turning the box over. Is there a reason I shouldn't do this?

[Ethan Winer]

I can't tell what you're describing, but if it's only 17 inches I don't think it will do very much.

--Ethan

[ToBeFrank]

I'm planning to build a panel resonator (panel absorber, membrane absorber, diaphragmatic absorber... it appears to have several names) with depth 17.25" and panel with surface density of either .71 or .9 pounds per square foot (PSF). Everything I've read indicates one of these is very effective at it's tuned frequency and my calculations show that to be 48Hz or 43Hz respectively. Why do you think it wouldn't be effective?

[PaulP]

Originally Posted by ToBeFrank

Everything I've read indicates one of these is very effective at it's tuned frequency and my calculations show that to be 48Hz or 43Hz respectively.

Care to share your sources with us ?

Paul P

[SAC]

I think we are at logical impasse here!

Typically a tuned enclosure for such a low frequency is larger.
But if you are sure of the design, why are you asking? As if anyone dares express any question we are told your calculations say its correct.

And if the design supports turning it around (for which I am not sure of the advantages...)...OK. But typically panel resonators have a front and back - but there is certainly no hard requirement stating this.

Which then begs the question itself, if it is correct, why are you asking?

The fact is, we have little to no information on which to base any opinion, except your assertion that 'it works'.

OK.........

[ToBeFrank]

Originally Posted by PaulP

Care to share your sources with us ?

This is the main one: Master Handbook of Acoustics. I've also read about people having success with panel absorbers on this forum and on other forums.

I sense some hostility and I'm not sure why. I really would like to know why this wouldn't work. I used the formula from Everest's book to calculate the resonant frequency. Ethan's comment seems to indicate he thinks 17" isn't deep enough, but most of what I've read is people building with much smaller depths (around 4") with subsequently higher resonant frequencies. Ethan's own design uses a much smaller depth than mine. Will it not work because the depth is so large?

[ToBeFrank]

Originally Posted by SAC

Typically a tuned enclosure for such a low frequency is larger.
But if you are sure of the design, why are you asking? As if anyone dares express any question we are told your calculations say its correct.

Which then begs the question itself, if it is correct, why are you asking?

The fact is, we have little to no information on which to base any opinion, except your assertion that 'it works'.

I am not sure it is correct and I have not asserted that it works. What I have asserted is that I ran the numbers correctly through Everest's (or whomever's formula it actually is) formula. The formula is (using "lb/sq ft" and "inches"):

F = 170 / sqrt(m * d)

Is this not the correct formula to use or does it not work out in the real world?

And if the design supports turning it around (for which I am not sue of the advantages...)...OK.

Since Everest's formula is the theoretical resonant frequency, it's possible (likely?) that I won't hit my intended resonant frequency of 43Hz. If I could flip the box and get a slightly higher or lower resonant frequency, I may be able to get closer to my intended resonant frequency.

But typically panel resonators have a front and back - but there is certainly no hard requirement stating this.

The BBC report on their low frequency panel absorbers indicated that the box didn't absorb low frequencies as well when the back panel was not sufficiently thick. Since my box will be placed on the floor, I was wondering if the floor would be acting like a thick back panel and thus I could get away with the thinner back panel.

Should I be asking these questions on some other forum?

[PaulP]

Originally Posted by ToBeFrank

This is the main one: Master Handbook of Acoustics. I've also read about people having success with panel absorbers on this forum and on other forums.

I sense some hostility and I'm not sure why. I really would like to know why this wouldn't work. I used the formula from Everest's book to calculate the resonant frequency. Ethan's comment seems to indicate he thinks 17" isn't deep enough, but most of what I've read is people building with much smaller depths (around 4") with subsequently higher resonant frequencies. Ethan's own design uses a much smaller depth than mine. Will it not work because the depth is so large?

No hostility intended, and welcome to this forum. I do question the bit about
people having success with panel absorbers. I've been studying the subject
for a while now and have seen very few attempts and little in the way of test
results indicating how well they work.

Ethan probably has the most experience, and is the source for the most
comprehensive document I've seen (written quite a while ago), but he's usually
fairly quiet on the subject .

Two forum members did build some recently, see :

• Somebody Explain to me: "freestanding" traps vs. "sealed" traps...
• build and test of resonator traps
• whats the relationship between air and absorbtion

They both stopped posting about their efforts so we don't know how well
things turned out.

If you have additional sources of implementations I'd be interested in seeing
them as I'm also interested.

By the way, the BBC appears to have abondoned the idea of panel absorbers
in favour of ones containing deep light insulation for absorption of very low
frequencies.

Paul P

[johndykstra]

what is the surface area of this 17" deep trap. I think Ethan thought you meant surface, not depth.

Are you certain where your 40ish-hz problem is coming from? If it's not a height mode, I reckon on the floor is of little value, unless of course it's placed in a high pressure area between front and back wall, but that's assuming that it's a length mode.

Panel traps being meant to be sealed, I'd worry about the elasticity of a resonant back panel, as it will likely affect the tuning of the front panel.

shootin in the dark here, but these are the questions I'd be asking.

[DanDan]

tobefrank, you are welcome here.

I sense some hostility and I'm not sure why.

Your senses are acute but it is a singular phenomenon I assure you.

Panel and other resonant absorbers get discussed quite frequently.
What seems to be consistently missing is some way of evaluating the potency of them vs other LF designs. Clearly they will have an effect at resonance, but perhaps little at all other frequencies as they are sealed. How does one compare that to a broadband unit like a SuperChunk or the in-between Corner Straddling panel trap?
I reckon you would need a very large amount of this type of treatment to achieve a significant effect. e.g. in concert halls, sometimes too much nice looking wood panelling sucks the bass out of the room.
You might find my experiment here interesting Acoustics Forum • View topic - Helmholtz formula, is this incorrect?
DD

[SAC]

I would suggest a bit of empirical testing.

The Q and resonant frequency are dependent upon more than simple mass and surface area. Membrane compliance, enclosure volume, additional absorptive material will all modify the results.

And while, contrary to the opinion of a few, resonant traps do indeed work (although high Q); membrane traps, of all of the types of resonant traps, are by far the most complex as there are a multitude of components with far ranging compliance values available for the membrane.

And while D'Antonnio & Cox offer myriad information regarding the design of various other types of resonant traps, even they admit that membrane resonators are "still designed by trial and error" (p.158 Acoustic Absorbers and Diffusers).

On the other hand, porous and slot type resonators are covered in exceptional detail in the same text.

[jhbrandt]

Frank,

Check this out. I hope it helps - or confirms what you are doing.

--->> Porous Absorber Calculator V1.56

Cheers,
John

[jayfrigo]

Originally Posted by ToBeFrank

The BBC report on their low frequency panel absorbers indicated that the box didn't absorb low frequencies as well when the back panel was not sufficiently thick. Since my box will be placed on the floor, I was wondering if the floor would be acting like a thick back panel and thus I could get away with the thinner back panel.

Should I be asking these questions on some other forum?

Welcome. Don't worry, you're in the right place. As for the location and back panel, unless the box is firmly fixed in place (strongly coupled to the mounting surface), or the box is extremely heavy with the back on the floor (unlikely given the quoted size, plus a back-to-the-floor orientation may not be best), it likely will still be able to vibrate and flex enough to relieve pressure in back. Therefore the effectiveness will probably be somewhat diminished by keeping a thin back, based on what I can surmise from your description.

You could certainly build it that way and test it, but I think you're trying to avoid wasted time and materials by getting as close as possible with predictions before building, so I'd say you're safer keeping to a back with more mass, unless it is possible in your application to firmly fix it in place, where the back will be coupled to the mounting surface (floor or wall, depending on orientation; I wasn't 100% clear about your intent from the description). Obviously, once built, you still may require some trial and error to fine tune it, but your careful investigation up front improves your chances of being close to your design goal.

I assume you plan to build several? A single device of that size is unlikely to change the room's response very much. The more area you can cover, the greater the impact will be.

[Ethan Winer]

Originally Posted by ToBeFrank

Ethan's comment seems to indicate he thinks 17" isn't deep enough

In the OP he wrote "I am planning to build a 17.25" tall panel resonator." So my comment was that it's too small to work well.

--Ethan

[ToBeFrank]

Originally Posted by PaulP

They both stopped posting about their efforts so we don't know how well
things turned out.

This seems to be the common theme. It's either "I'm going to try this" and then no results or "I built some and they worked great" but no details on how they were built and placed.

Originally Posted by dykstraster@gmai

what is the surface area of this 17" deep trap.

The first experiment will be 8 square feet (2' x 4').

Are you certain where your 40ish-hz problem is coming from? If it's not a height mode, I reckon on the floor is of little value, unless of course it's placed in a high pressure area between front and back wall, but that's assuming that it's a length mode.

Good point. I've done so much moving and measuring of my sub I thought I remembered it as a height mode. I tested it last night and indeed the 43Hz mode is a tangential mode. I've already identified the wall next to my superchunk'ed corner as my likely placement since it is definitely a high pressure area for this mode. I'm curious to see how it affects the superchunk's high frequency absorption.

Panel traps being meant to be sealed, I'd worry about the elasticity of a resonant back panel, as it will likely affect the tuning of the front panel.

Good point which is reiterated below. I'm going with a stiff back panel.

Originally Posted by DanDan

Panel and other resonant absorbers get discussed quite frequently.
What seems to be consistently missing is some way of evaluating the potency of them vs other LF designs.

I hear ya. I haven't been able to find any data that gives a clue on how much surface area I need to affect the mode. All of my mineral wool panels (4") and superchunks haven't touched anything below 50Hz, which is why I decided to experiment with a panel absorber.

You might find my experiment here interesting Acoustics Forum • View topic - Helmholtz formula, is this incorrect?

This looks like you were able to get a 6dB reduction? If I can get that I will be happy as that is half the mode and a lot less EQ I'll have to do.

Originally Posted by jayfrigo

Welcome. Don't worry, you're in the right place. As for the location and back panel, unless the box is firmly fixed in place (strongly coupled to the mounting surface), or the box is extremely heavy with the back on the floor (unlikely given the quoted size, plus a back-to-the-floor orientation may not be best), it likely will still be able to vibrate and flex enough to relieve pressure in back. Therefore the effectiveness will probably be somewhat diminished by keeping a thin back, based on what I can surmise from your description.

Thank you, this is very helpful. I have decided to have a stiff back panel so as not to compromise the performance.

I assume you plan to build several? A single device of that size is unlikely to change the room's response very much. The more area you can cover, the greater the impact will be.

The plan was to build one and see how well it works, whether I can hit the frequency I'm after, etc. The first will be 8 square feet. If I can get some results, I'll build more.

Thanks everyone for the insight!

[ToBeFrank]

Originally Posted by Ethan Winer

In the OP he wrote "I am planning to build a 17.25" tall panel resonator." So my comment was that it's too small to work well.

I was thinking of the panel itself as being 2D (l x w) so tall was meant to indicate the depth of the whole device. I understand the confusion and it was my bad for not being more clear.

[DanDan]

Hi 2BF, that Helmholtz post was long and about a couple of issues. However you have read some part of it very optimistically. i.e. the 6dB. Nothing like that happened. I certainly achieved resonance in the device and a tiny reduction in SPL at resonance very close to the device. However no change in the Waterfall or Frequency Response or EDT plots. There is no way that I could get anything like the performance seen in the book. I cannot explain this, my device was very similar, but tunable. This result puts a question mark over the experiment in the book IMHO.
Some posters have suggested that one would need a wall of these to get any significant result.
I am of the opinion that you would need a similarly large amount of panel traps.
Both of these trap designs use damped resonance. If the resonance is left undamped they can actually add to the problem, returning a resonant boom to the room. With too much damping they will have no effect at all. A tricky balance.

Regarding your mode frequencies. Are these calculated or measured?
These two are rarely the same. It may be worth playing sines in your room, finding the modes. This will 100% tell you which mode is where and where best to place traps. There is a wonderful little LabMeter at rustykat.com
DD

[ToBeFrank]

Originally Posted by DanDan

Hi 2BF, that Helmholtz post was long and about a couple of issues. However you have read some part of it very optimistically. i.e. the 6dB. Nothing like that happened. I certainly achieved resonance in the device and a tiny reduction in SPL at resonance very close to the device. However no change in the Waterfall or Frequency Response or EDT plots. There is no way that I could get anything like the performance seen in the book. I cannot explain this, my device was very similar, but tunable. This result puts a question mark over the experiment in the book IMHO.
Some posters have suggested that one would need a wall of these to get any significant result.
I am of the opinion that you would need a similarly large amount of panel traps.
Both of these trap designs use damped resonance. If the resonance is left undamped they can actually add to the problem, returning a resonant boom to the room. With too much damping they will have no effect at all. A tricky balance.

You guys are probably right that I won't get much change, if any. But hey, I'm a gearhead type who likes to build and measure stuff so don't deprive me of that.

Regarding your mode frequencies. Are these calculated or measured?
These two are rarely the same. It may be worth playing sines in your room, finding the modes. This will 100% tell you which mode is where and where best to place traps. There is a wonderful little LabMeter at rustykat.com
DD

Both. I've calculated the modes, and I use REW and a calibrated ECM8000 to measure the room. The mode I'm going after is definitely related to either length or width or both. The corner where it is loudest is where I'm going to experiment with placement of the panel absorber.

[DanDan]

The mode I'm going after is definitely related to either length or width or both.

If you play a sine at the mode's frequency you should be able to tell exactly which dimenion is involved. Simply walk each axis, you will pass through alternate deep null(s) and peaks when you are in the right path.

Note that that frequency readouts in software are not necessarily accurate.


DD

[ToBeFrank]

I thought I'd give an update. The formula appears to be a pretty good indicator of the tuned frequency as the box I built does indeed attenuate the 43Hz mode by 2dB. The box is attenuating maximally at just below 43Hz and has effect +-5Hz from there. Interestingly, the box also has a large effect above 60Hz. I'm guessing this is because of the 2 layers of R30 fiberglass insulation batts (19" total depth) that are in the box.

[DanDan]

Thanks for getting back to us. That seems like really good result. If there is some way of introducing some dry sand into the box you should be able to 'tune' it upwards exactly. Might be best to wait a few months though, it may dry a bit.
2dB or so in Frequency response suggests to me that there might be a very much bigger effect in the time domain. Have you looked at the Waterfall and EDT graphs? Could you post the Waterfalls?

DD

[SAC]

As dd suggests, additional FR/waterfalls would be of benefit, as one of the insidious aspects of resonant absorbers is that they tend to interact with the environment and cause the frequency and location of the modes to shift just a bit, which is part of the iterative (meaning : pain in the posterior small incremental adjustment) process of the undertaking.