How critical are box dimensions 2.6/1.6/1 ???

[avaserfi]

I think you linked the wrong post because there are only two posts and the are about Harry Potter??

I tried the magic route, didn't work to well. I edited the link it goes to the right place this time!

 

[Guiria]

What bandwidth is your response good through +/-1db???

WinISD only goes up to 1 kHz so my +/-1db is from 66Hz-up given a ported .25 ft^3 enclosure. It would be nice to see how the high end models but I don't think WinISD has that capability. Maybe I'm just a newb at it.

The speakers will be in a bedroom that is 12x13.5x8. I will rarely be cranking these things since my daughter's bedroom is exactly over mine.

 

[WmAx]

How does one achieve a true non-resonance type enclosure then?

As far as achieving vanishingly low resonance that is not likely to be audble, one can take several routes to partially or fully address the issue:

(1) Shift the resonance band of the enclosure above that of the midrange band driver.

(2) Increase stiffness/mass to a level that reduces resonance amplitude to a sufficiently low level.

(3) Dampen the resonance by converting to an alternate energy form(heat).

(4) Forcing the resonance into a high Q band of [ideally] single primary resonant frequency that is not likely to be excited substantially by most source material.

(5) De-couple midrange/midbass driver(s) from the cabinet with a suspension.

(6) Combination of the above.

The B&W Nautilus model seems to have a near non resonant cabinet based upon it's design I would think?

If you mean the top units: 801D, 802D and 800D; yes. These cabinets are virtually inert for so far as the purpose of audibility is concerned.

Where did you get a non-resonant reference speaker?

My virtually non-resonant reference is an extraordinary linear headphone(with some modification): a Beyerdynamic DT880 (pre-2006 model). With modifications, it has but two minor resonances (in the treble band) at a very low level, and a midband cleaner than any speaker system of which I know about. Combine this with zero room influence(no reverb contributes to making resonances harder to detect; improving this aspect yet further), and it's a superb resonant-free reference for midrange.

-Chris

 

[annunaki]

As far as achieving vanishingly low resonance that is not likely to be audble, one can take several routes to partially or fully address the issue:

(1) Shift the resonance band of the enclosure above that of the midrange band driver.

(2) Increase stiffness/mass to a level that reduces resonance amplitude to a sufficiently low level.

(3) Dampen the resonance by converting to an alternate energy form(heat).

(4) Forcing the resonance into a high Q band of [ideally] single primary resonant frequency that is not likely to be excited substantially by most source material.

(5) De-couple midrange/midbass driver(s) from the cabinet with a suspension.

(6) Combination of the above.



If you mean the top units: 801D, 802D and 800D; yes. These cabinets are virtually inert for so far as the purpose of audibility is concerned.



My virtually non-resonant reference is an extraordinary linear headphone(with some modification): a Beyerdynamic DT880 (pre-2006 model). With modifications, it has but two minor resonances (in the treble band) at a very low level, and a midband cleaner than any speaker system of which I know about. Combine this with zero room influence(no reverb contributes to making resonances harder to detect; improving this aspect yet further), and it's a superb resonant-free reference for midrange.

-Chris

What does a set of those headphones run?

As for the B&W's I meant the Nautilus Model. The snail shell looking speaker, the top of the line $40,000 a pair Nautilus.

Which of the above options are the best to use in combination? I am thinking that a combination of 5, 3, and 1 would work fairly well in designing a full range speaker?

 

[WmAx]

What does a set of those headphones run?

If you find a pair of this older model, I guess you would end up paying about $150-$200, depending on condition. I can not recommend the current version. It has, according to measurements, an elevated treble band, compared to the older unit.

Which of the above options are the best to use in combination? I am thinking that a combination of 5, 3, and 1 would work fairly well in designing a full range speaker?

It depends on the specifics of a particular application, as well as the degree to which you execute the method(s).

Take a B&W 802D as an example. It is a combination of methods. For the woofer module, it uses extensive matrix bracing every few inches internally to push the panel resonances above the bandwidth of the wooofer. For the midrange module, they use an extremely stiff, high density composite that has high damping properties. The mid-range driver is also mounted on an extensive suspension system to decouple it from the module. The entire mid-range module in turn is decoupled by way of suspension from the woofer module, to prevent the midrange module from exciting the woofer module panels.

-Chris

 

[annunaki]

Chris,

I am quite familiar in how the B&W system works and is put together.

Is there a certain formula for finding the enclosure resonance after bracing has been installed? Assuming one does seperate enclosures for each driver to ensure low/no resonance, how does one test to ensure they have succeded?

What materials work best for each type of driver? MDF/HDF for the woofer/bass portion? High grade marine plywood for the midrange? Vice versa? Does a tweeter need it's own enclosure?

 

[annunaki]

WinISD only goes up to 1 kHz so my +/-1db is from 66Hz-up given a ported .25 ft^3 enclosure. It would be nice to see how the high end models but I don't think WinISD has that capability. Maybe I'm just a newb at it.

The speakers will be in a bedroom that is 12x13.5x8. I will rarely be cranking these things since my daughter's bedroom is exactly over mine.

You will need a better program that you can use for simulation for those drivers. The response above 1kHz will be critical. I do not think that WinIsd will go to a full range signal. Do not begin your purchase or build until you can model them full range.

 

[Guiria]

You will need a better program that you can use for simulation for those drivers. The response above 1kHz will be critical. I do not think that WinIsd will go to a full range signal. Do not begin your purchase or build until you can model them full range.

So what do yo use or wish you were using for speaker building software? WinISDPro will allow you to model the graph full range, however, it measures it pretty much ruler flat which I find hard to believe from a 26 dollar driver.


This is the link that shows the SPL vs. Frequency graph for the driver
http://www.partsexpress.com/pdf/264-844.pdf

How it performs in the given box, however, is a good question.

 

[annunaki]

WinIsd Pro Alpha is a good program. You may want to look into Leap or some others. The really good ones you usually have to pay for

 

[WmAx]

Is there a certain formula for finding the enclosure resonance after bracing has been installed?

Yes, given sufficient physical parameters for the materials to be analyzed, it is possible to estimate. But it's not efficient to do manually for a cabinet system; it is complex/labor intensive math that is best suited to software analysis. You can read some about panel resonant mode analysis at this link:

http://www.audioholics.com/education/loudspeaker-basics/lowering-mechanical-noise-floor-in-speakers

Assuming one does seperate enclosures for each driver to ensure low/no resonance, how does one test to ensure they have succeded?

To measure the panel resonant frequencies one needs to attach an accelerometer to various points on the cabinet and derive the frequency response of the cabinet surface(s) by using a driver mounted in the cabinet to provide the stimulus. The resonant points will be indicated by large response peaks in the response.

I am not aware of an existing way for a DIYer to practically measure final acoustic output response of a cabinet vs. direct driver output. You need a far field, multi-axis impulse response analysis in order to derive waterfall type response charting. With a very wide bandwidth anechoic chamber, or the speaker mounted 50ft in the air on a pole, it's possible. I have an idea on how to test in a normal room, but I have not yet tested this, nor am I ready to talk about this procedure in public before I test it.

What materials work best for each type of driver? MDF/HDF for the woofer/bass portion? High grade marine plywood for the midrange? Vice versa?

Best material(s) depends on the specific application as well as other construction techniques used in the process.

If your woofer is being cut at 100Hz, standard MDF moderately braced, or a high grade hardwood plywood, somewhat less braced, will have primary resonant modes well above 100Hz. If your woofer is being cut at 500Hz, far more consideration is required in shifting the primary resonance(s) well above this band; this is where materials like HDF or very high grade hardwood ply comes into play as being easier to use for the application as compared to MDF.

Does a tweeter need it's own enclosure?

I am not aware of any benefit unless the tweeter's back is not air tight.

-Chris

 

[Clarence Hyder]

resonance and fooling a driver

Are you sure about this? From my research in the subject acoustic foam or what I am planning on using (OC705 which is far more effective) is useful if the proper thickness is used (2" should suffice for most applications in eliminating modal resonance and stop internal energy reflection). There is one caveat however, using too much can cause bass roll off this is caused by the extreme dampening effect of OC705 which can inhibit the cabinets ability to resonate the air volume properly.

I am pretty sure this in no way will trick a driver. The best advice would be compromise and don't use the specific box dimensions. Use the T/S parameters to build the best possible box for the drivers.

Also, if you want to properly combat resonance issues the best and only way a DIYer can reasonably deal with this situation is proper bracing on all axises within the given cabinet.

You can definatly attenuate resonance issues with cabinet dampening materials especially Acouta stuff, simalar to fiberglass but with much better energy absorbing properties. Also you absolutely can build a box and with material such as acousta stuff or fiberglass make the driver think its in a larger box. As much as 30 to 40 per cent.

If your brace is too close to the back of the woofer a "loading" issue can occur where the air flow around the woofer is different on one side than the other causing non linear movement of the coil and cone. Distortion grows from these gardens/

Panel resonance cant be eliminated with braces as it is the entire cabinet that vibrates. Many test have been done on this subject and its found that bracing mainly just raises the frequency of the resonance not eliminate it. Damping more or less eliminates it.

The golden ratio as noted in the fist post is a formula that if followed wont allow any standing waves to couple and therefor make them grow in magnitude. Damping does the same thing. Generally 1/2 pound of acousta stuff per cubic foot is ideal less if a ported system more

 

[Clarence Hyder]

How does one achieve a true non-resonance type enclosure then? The B&W Nautilus model seems to have a near non resonant cabinet based upon it's design I would think? Where did you get a non-resonant reference speaker?

I am quite happy with the speakers, however, they are not as detailed as newer models. B&W's drivers have improved much since my speakers were in production.

The closest I know of non resonant cabinets is one that is a cabinet within a cabinet. build the initial cab from 1 half inch mdf and another of 1 half or 3 quarters over it with a layer of roofing felt btween them only attached at the corners. Test proved this method to reduce cabinet resonance by 97 percent. Bracing didnt eliminate it at all it just changed the frequency.

 

[WmAx]

You can definatly attenuate resonance issues with cabinet dampening materials especially Acouta stuff, simalar to fiberglass but with much better energy absorbing properties. Also you absolutely can build a box and with material such as acousta stuff or fiberglass make the driver think its in a larger box. As much as 30 to 40 per cent.

Low density fiberglass, wool, poly fill and other fibers at low density have been shown to increase a closed box's apparent internal volume by some small degree. But I believe the poster was talking about high density fiberglass or mineral board, in which case, an excess of this material can prevent the tuned box resonance from occuring; this material is extremely efficient at damping acoustic energy vs. thickness. Much more effective at damping as compared to standard low density fiberglass or Acousta-Stuff(Dacron fiber that has been machine crimped). As an exampe: place 4" of the high density (6-8lb/ft^3) fiberglass or mineral wool board in all walls(type A mounting) of a an enclosure, and virtually no acoustical energy could reflect(or resonate) from about 125Hz and over. However, the material does not suddenly stop working at a fixed frequency. There would still be considerable effect well under 100Hz in this case.

Panel resonance cant be eliminated with braces as it is the entire cabinet that vibrates. Many test have been done on this subject and its found that bracing mainly just raises the frequency of the resonance not eliminate it. Damping more or less eliminates it.

Bracing is a critical part, if done correctly. While bracing raises the resonant frequency, if properly distributed with enough bracing points, the resonance will also be reduced in amplitude, to a degree. Bracing alone is certainly seems to not be the answer to a midrange enclosure(as it is very difficult to push the panel resonances completely over the midrange bandwidth), though it is applicable as an only solution for woofer modules, where one can push the panel resonances substantially over the bandwidth of the driver(s).

The golden ratio as noted in the fist post is a formula that if followed wont allow any standing waves to couple and therefor make them grow in magnitude. Damping does the same thing. Generally 1/2 pound of acousta stuff per cubic foot is ideal less if a ported system more

There is no shape one can use that will completely prevent internal cavity resonances sufficiently as a single solution. But this is useful when combined with low effectiveness damping materials as are commonly used in speakers(polyfill, standard grade foam, etc.). While these materials effectively damp upper midrange and treble resonances, they do not have much effect on lower midrange/midband frequencies in normal application unless the entire space is filled, and in my experiences, it still will not be as effective at eliminating lower mid-bass range reflections/resonances as compared to the high density board type fiberglass or mineral wool in a normal enclosure.

-Chris

 

[WmAx]

The closest I know of non resonant cabinets is one that is a cabinet within a cabinet. build the initial cab from 1 half inch mdf and another of 1 half or 3 quarters over it with a layer of roofing felt btween them only attached at the corners. Test proved this method to reduce cabinet resonance by 97 percent. Bracing didnt eliminate it at all it just changed the frequency.

To properly build a cabinet-in-a-cabinet system, the two cabinets must be effectively decoupled mechanically(a single solid connection will dramatically reduce the attenuation effect of the isolated cabinet systems) and have air-tight connections to prevent an air breach between the two. It is no simple task to properly build a cabinet using this method. The inner cabinet system must still take precautions to prevent a high degree of resonance, as an unbraced/non-damped inner cabinet system would have a very high amplitude that may breach acoustic output through the driver. BTW, there are better materials to use as compared to roofing felt for this application. A high density, high ILD latex foam would be more reliable in this capacity. Or, alternatively, a very low durometer rated durable rubber used in block points.

-Chris

 

[Clarence Hyder]

ouch! is this a competition?

To properly build a cabinet-in-a-cabinet system, the two cabinets must be effectively decoupled mechanically(a single solid connection will dramatically reduce the attenuation effect of the isolated cabinet systems) and have air-tight connections to prevent an air breach between the two. It is no simple task to properly build a cabinet using this method. The inner cabinet system must still take precautions to prevent a high degree of resonance, as an unbraced/non-damped inner cabinet system would have a very high amplitude that may breach acoustic output through the driver. BTW, there are better materials to use as compared to roofing felt for this application. A high density, high ILD latex foam would be more reliable in this capacity. Or, alternatively, a very low durometer rated durable rubber used in block points.

-Chris

#30 Roofing felt was used in the test because it was recommended by Vance Dickason in his book. Mainly because its cost at .17 cents per foot is significantly cheaper than the 2.00 per sq. foot for loaded vynyl. and performed better in several areas and as well in allmost all of them.

And I didnt say it was easy just that this was the best way to construct an affordable anti resonant cabinet. And I know it needs to be isolated with minimal mechanical coupling. Airplanes of all types use constraining layers to dampen and attenuate the noise outside the plane. They use 5mil aluminum sheets and E.A.R's ADC-124 Lightweight damping foam , but you could buy a house for what that material cost. I try to live in the real world of DIY. I know some people have lots of money to spend, but 99 percent of the people I have attempted to help were always interested in a less expensive alternative if the end result was even relativly close. I guess I need to clarify every single statement I make in massive detail unless I want to get called on every detail of every post. Im not here to claim the throne of "King of the DIY Knowledge" just to help people that ask questions.

No one is doubting your loudspeaker prowess, I am simply trying to offer another intelligent opinion. But thanks for clarifying all those points for me. (Not exactly necessary) Not everyone has as deep an understanding of the subject as we may have. So I tend to leave out alot that I feel will only confuse novice builders. I admit I did miss the fact that he was talking about a dense material on the walls of the cabinet. And I do agree with you on that. And everything else for that matter. I just didnt type in all the details for each comment I made. Thanks lets do this again soon. Jeff Hyder