Hi. I have an interesting question about bass reflex enclosures. What do you get when you build a bass reflex enclosure that is tuned to one frequency, but then have the port exit into another chamber tuned to a different frequency by another port which exits to the outside?
Is there a set of equations that can determine the resultant tuning frequency for a series-tuned enclosure like this? How would one determine what frequency the second chamber is tuned to if the driver is only in the first chamber? And, would there be any guidelines for determining how large that second chamber should be?
If so, I think I have an idea, possibly even a patentable one, for getting really big bass out of a really slim, small speaker enclosure.
Also, if there are any books or AES papers that can help me on my search for this series-tuned enclosure idea that I have, I'd like to know.
This prompts my question: In this web page, there is a sample set of design calculations for a series-tuned double-chamber reflex subwoofer enclosure. At the bottom of the page, it tells what the resultant tuning frequency of each chamber is, but not how that number was determined. That is part of what I am really looking for.
I have a feeling that those calculations which use an equivalent electrical circuit to describe acoustic system behavior might be of use here.
Is there a set of equations that can determine the resultant tuning frequency for a series-tuned enclosure like this? How would one determine what frequency the second chamber is tuned to if the driver is only in the first chamber? And, would there be any guidelines for determining how large that second chamber should be?
If so, I think I have an idea, possibly even a patentable one, for getting really big bass out of a really slim, small speaker enclosure.
Also, if there are any books or AES papers that can help me on my search for this series-tuned enclosure idea that I have, I'd like to know.
This prompts my question: In this web page, there is a sample set of design calculations for a series-tuned double-chamber reflex subwoofer enclosure. At the bottom of the page, it tells what the resultant tuning frequency of each chamber is, but not how that number was determined. That is part of what I am really looking for.
I have a feeling that those calculations which use an equivalent electrical circuit to describe acoustic system behavior might be of use here.
It's called "double bass reflex" and it can do really good things, like reduce excursion at your second tuning frequency, thus increasing power handling.
From what I hear, they are very tricky to get right, and they're not completely general. IIRC, if your normal BR box is tuned to 40hz, the double bass reflex box will have a tuning at 40hz and a tuning at 80hz and will be just enough larger to accomodate the extra wood and ports. The excursion vs frequency plot will have two low points instead of one, and the impedance curve will have 3 humps instead of 2.
Also try Google...
From what I hear, they are very tricky to get right, and they're not completely general. IIRC, if your normal BR box is tuned to 40hz, the double bass reflex box will have a tuning at 40hz and a tuning at 80hz and will be just enough larger to accomodate the extra wood and ports. The excursion vs frequency plot will have two low points instead of one, and the impedance curve will have 3 humps instead of 2.
Also try Google...
Not hard at all;
http://www.diysubwoofers.org/prt/dual_chamber.htm
I used UniBox and designed a ported box with 2 ports.
You can also use WinIsd and just model it for two ports.
The baffle with the third port is then placed 1/3 the inside length of the box. Say your box is 30" high, the baffle would be placed at the 10" mark. One chamber would be 20" and the second chamber 10". 3 ports all the same length. It's true, you won't get over excursion below resonance. I tried really hard one time and all I got was a lot of wind come out the ports.
http://www.diysubwoofers.org/prt/dual_chamber.htm
I used UniBox and designed a ported box with 2 ports.
You can also use WinIsd and just model it for two ports.
The baffle with the third port is then placed 1/3 the inside length of the box. Say your box is 30" high, the baffle would be placed at the 10" mark. One chamber would be 20" and the second chamber 10". 3 ports all the same length. It's true, you won't get over excursion below resonance. I tried really hard one time and all I got was a lot of wind come out the ports.
I have also been facinated with this. The company that has used this alignment with the most success has been Audiophysic. The Tempo (old version), Spark (current and previous version) as well as the Yara all use this type of design.
http://www.audiophysic.com/produkte/spark/index_e.html
About a month ago, I built a test cabinet to try out the double reflex design. I modeled a box using win isd for my eton 7-372 monitors which turned out to be .7 cu ft with a 4" x 8.75" port tuned to if I remember correctly 50 Hz. I then added another identical cabinet to the bottom of this. I used the same size and tuning. The results were very positive. I did some measurments and the actual tuning was closer to 70 Hz for the first cabinet and then 40 for the second cabinet. Upon comparison, the series tuned cabinet produced much deeper bass, with good tonality, then the smaller single reflex box. I also learned, from both listening and measuring, that there is a dip in the response just below the tuning frequency of the first chamber around 40-55 Hz.
http://www.audiophysic.com/produkte/spark/index_e.html
As time allows, I plan on building quite a few more test cabs. Hopefully using a little more science....
Also, I've heard that LSPcad can effectively model series tuned enclosures.
Nate
http://www.audiophysic.com/produkte/spark/index_e.html
About a month ago, I built a test cabinet to try out the double reflex design. I modeled a box using win isd for my eton 7-372 monitors which turned out to be .7 cu ft with a 4" x 8.75" port tuned to if I remember correctly 50 Hz. I then added another identical cabinet to the bottom of this. I used the same size and tuning. The results were very positive. I did some measurments and the actual tuning was closer to 70 Hz for the first cabinet and then 40 for the second cabinet. Upon comparison, the series tuned cabinet produced much deeper bass, with good tonality, then the smaller single reflex box. I also learned, from both listening and measuring, that there is a dip in the response just below the tuning frequency of the first chamber around 40-55 Hz.
http://www.audiophysic.com/produkte/spark/index_e.html
As time allows, I plan on building quite a few more test cabs. Hopefully using a little more science....
Also, I've heard that LSPcad can effectively model series tuned enclosures.
Nate
Double chamber reflexes, with ports in the large and small chamber and one between the two will always have a notch in the response near the stopband - not really Hi-Fi, but since the ear is nearly impervious to relatively narrow, sharp dips (like TL's have) you probably won't notice.
A series-tuned reflex can be modeled with LspCAD - I doubt there is any real advantage to it or loudspeaker manufacturers would run to it.
TL's and TQWT's and all other designs that have been touted as giving you bass for nothing have all proved to be wishful thinking by the designer and/or to have serious downsides. There may be sonic effects from the inherent cabinet bracing but otherwise TL's have no more/better bass than a ported box. TL fans said as recently as 5-7 years ago that TL enclosures had a 6dB/octave rolloff. I remember an article in Speaker Builder where a guy using cookbook formulas (presumably from one of the Weems' books) thought he could get an extra octave out of some TL boxes he was putting in his truck - imagine his chagrine when they performed no better than the predictions for (smaller) sealed or vented boxes
Our knowledge of TL's has improved recently due to some very professional amateur efforts and we now know many of the original claims for them are false....
If you want to develop equations, get some college (typically senior or graduate level) textbooks on electroacoustics and circuit analysis.
Acoustics by LL Beranek
Acoustical Engineering by Harry Olson
Theory and Design of Loudspeaker Enclosures by JE Benson
These should give you a good start.
A series-tuned reflex can be modeled with LspCAD - I doubt there is any real advantage to it or loudspeaker manufacturers would run to it.
TL's and TQWT's and all other designs that have been touted as giving you bass for nothing have all proved to be wishful thinking by the designer and/or to have serious downsides. There may be sonic effects from the inherent cabinet bracing but otherwise TL's have no more/better bass than a ported box. TL fans said as recently as 5-7 years ago that TL enclosures had a 6dB/octave rolloff. I remember an article in Speaker Builder where a guy using cookbook formulas (presumably from one of the Weems' books) thought he could get an extra octave out of some TL boxes he was putting in his truck - imagine his chagrine when they performed no better than the predictions for (smaller) sealed or vented boxes
Our knowledge of TL's has improved recently due to some very professional amateur efforts and we now know many of the original claims for them are false....If you want to develop equations, get some college (typically senior or graduate level) textbooks on electroacoustics and circuit analysis.
Acoustics by LL Beranek
Acoustical Engineering by Harry Olson
Theory and Design of Loudspeaker Enclosures by JE Benson
These should give you a good start.
Take a look at Martin King's website.
http://www.quarter-wave.com/
Note: to use his worksheets, you will need a fairly recent version of MathCad, but there are some tables also.
http://www.quarter-wave.com/
Note: to use his worksheets, you will need a fairly recent version of MathCad, but there are some tables also.
This type can be modelled, using Adire drivers, on LSPCAD for Adire, which is freeware.
Notice that all three boxes about to be shown have different port configurations.
Note: If you can find an Adire driver with Lower Qts than the one you woant to model, , you can just add a series resistor to the Adire driver to raise the Adire's Qts to equal any non-Adire driver you have in mind. Tips on how to do that in coming post.
Notice that all three boxes about to be shown have different port configurations.
Note: If you can find an Adire driver with Lower Qts than the one you woant to model, , you can just add a series resistor to the Adire driver to raise the Adire's Qts to equal any non-Adire driver you have in mind. Tips on how to do that in coming post.
Attachments
And here is the Weems/Augspurger, (Weems gives full credit to Augspurger), type.
Here are two posts on it, #34 and 35, with instructions from Weems' Great Sound Speaker Manual.
http://www.diyaudio.com/forums/showthread.php?postid=93570#post93570
the thumbnail also shows how to add a series resistor to raise the Qts-which you can then model on LspCAD to convert the Qts of an Adire woofer to one identical to a non-Adire woofer you might have in mind.
And here are several pages devoted to it by diyAudio member Claudio Negro:
http://paginas.terra.com.br/educacao/claudionegro/english/
Here are two posts on it, #34 and 35, with instructions from Weems' Great Sound Speaker Manual.
http://www.diyaudio.com/forums/showthread.php?postid=93570#post93570
the thumbnail also shows how to add a series resistor to raise the Qts-which you can then model on LspCAD to convert the Qts of an Adire woofer to one identical to a non-Adire woofer you might have in mind.
And here are several pages devoted to it by diyAudio member Claudio Negro:
http://paginas.terra.com.br/educacao/claudionegro/english/
Attachments
Martin, I guess there is more than one way to take what I have written - especially given the context. By calling your efforts "amateur" I intended to give you credit for doing this on what apparently is your own time for your own interest and challenge for little or no monetary gain - it is not intended to belittle your work at all.
It is interesting that once the Bose patent on 6th order bandpass expired they jumped on pipes again as their "new" technology. We all know the marketing power of the word "new".
Cheers.
It is interesting that once the Bose patent on 6th order bandpass expired they jumped on pipes again as their "new" technology. We all know the marketing power of the word "new".
Cheers.
Ron,
No offense taken, actually I thought it was a very accurate description and quite amusing. A few years ago I had a nemisis on a different forum who consistently referred to me as an "amateur", he ment it as a put down and as a way of discrediting the work I was presenting. I think he considered himself a "professional", I think he was really just a professional wanna be.
Personally I consider being an amateur an advantage, I do what I want and find interesting without taking orders from a boss (other then my wife). I can say anything that is politically incorrect when it comes to established speaker theory, contradict the pro's when I see an opening, and present anything I write on my own venue. How much better could it get, I spend 8 hours a day being a professional engineer and find doing the amateur thing is much more fun and interesting.
Thanks for the quote,
My question has to do with the first picture Kelticwizard showed from LspCad for Adire. Is there an equation out there for determining the optimum size of that bottom box? I'm searching Claudio Negro's page for something like this, because that is all I need to finish off my potentially patentable design. It's like a very high order vented box design that integrates three chambers.
MJK worksheets
I've been fiddling with Martin's worksheets from time to time for a long time now.
Due to my lack of woodworking skills, most of the times the designs never left the paper (or screen...).
But I have learned a lot from them.
Now I'm buiding (real, this time) a ML TL or ported box - your choice as I've found a decent craftsman to do the box.
Going back on topic, I've never built a double tuned box, but I would take in consideration that not only freq response, but also time delay is involved in good bass.
Wouldn't the bass loose it's tightness?
(Note: I like better low Qts boxes)
I've been fiddling with Martin's worksheets from time to time for a long time now.
Due to my lack of woodworking skills, most of the times the designs never left the paper (or screen...).
But I have learned a lot from them.
Now I'm buiding (real, this time) a ML TL or ported box - your choice
as I've found a decent craftsman to do the box.Going back on topic, I've never built a double tuned box, but I would take in consideration that not only freq response, but also time delay is involved in good bass.
Wouldn't the bass loose it's tightness?
(Note: I like better low Qts boxes)
Thanks for the positive feedback. I learn a lot from the comments I receive via e-mail or on the forums, this gets funneled back into the next revision of my documentation, designs, and software.
Personal status .... I think that you are exactly correct. People in the industy and academia, or who have published articles in some form of printed medium, have almost always treated me as a second class citizen when we have informally interacted. The usual criticism is the lack of peer review and formal peer acceptance of what I have presented. I have seen this same posturing with other unrelated technical societies that I have been exposed to, the technical societies and academics in general seem to attract these types of individual. Needless to say I have no real interest in publishing beyond my website. The industry and academia can take what I have presented or ignore it, the choice is up to them.
Yes, I support several commercial audio efforts. With the exception of Bob Brines, all commercial products using the MathCad software or my design ideas are people taking what I have provided for free and exploiting it. I have not recieved anything in return, with the exception of Bob Brines. This is why the newer versions of the MathCad worksheets, and some of my new designs, are not being made completely available on my site to the DIYer.
BAM,
There is a MathCad worksheet that analyzes the first configuration shown above on my webite's models page. If I remember correctly, the worksheet is set up to analyze a Fostex provided design. Somebody sent me measured results for this configuration and they matched the worksheet predictions very closely. All you need to do is iterate the variables input to the worksheet to get a response that meets your goals.
There is a MathCad worksheet that analyzes the first configuration shown above on my webite's models page. If I remember correctly, the worksheet is set up to analyze a Fostex provided design. Somebody sent me measured results for this configuration and they matched the worksheet predictions very closely. All you need to do is iterate the variables input to the worksheet to get a response that meets your goals.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.