First, the way I simulated this. I used a freeware version of Bullock and White's Transmission Line DOS simulator. Playing around with this program some time ago, I found that the program had a provision for an enclosure behind the woofer that is NOT part of the Transmission Line. Experimenting around with this, I discovered some months ago that if you make this enclosure large and the Transmission Line the size of a vent in a bass reflex, it perfectly models a ported box. Well, almost-more about that momentarily. Program is downloadable at Bob White's home page here:
http://www.hal-pc.org/~bwhitejr/
Bullock and White also have a different freeware program, available at the same place, that models closed, vented and drone cone, (passive radiator), boxes all by itself. The program I am using is meant for Transmission Lines. but you can manipulate it to model vented and even closed boxes.
The advantage of using the Transmission Line program to model ported boxes is that the Transmission Line program has provision for type of stuffing and amount of it. Ported and Sealed programs generally do not. So we can model a ported box with stuffing in the port by calling the the volume of the enclosure the small enclosure at the top of the line, and calling the port the Transmission Line.
It would be useful to download WinISD or any other quick simulator to get the dimensions of the port that tunes the box to a certain length. The Transmission Line Simulator does not have provision to tune by frequency, because it was designed to model Transmission Lines, not ported boxes.
A glitch: when you model a box of a certain size with a port of a certain size, the Transmission Line program for some reason shows it as tuning the box half an octave higher than any program, with a response curve to match. Why? Who knows? All I can say is to use Win ISD to model the box you wish, set WinISD to give you a port length that tunes the box half an octave, (.7 of the desired Fb) and transfer that port length and area to the Transmission Line model. The impedance curves of the Transmission Line model confirm that you will have the correct tuning when you use this method!!
Smaller glitch: even when you use the above method to get the correct tuning, the Transmission Line model will give you a ported box that is [1]slightly[/i] optimistic. A box that Win ISD will show an F3 of 30 Hz, for instance, will show up here with an F3 of maybe 28 Hz or so. Nothing to get excited about, just bear it in mind.
Incidentally, the separate program by Bullock and Whtie meat for modelling closed, vented and drone cone boxes do not have either of these glitches. Go figure.
http://www.hal-pc.org/~bwhitejr/
Bullock and White also have a different freeware program, available at the same place, that models closed, vented and drone cone, (passive radiator), boxes all by itself. The program I am using is meant for Transmission Lines. but you can manipulate it to model vented and even closed boxes.
The advantage of using the Transmission Line program to model ported boxes is that the Transmission Line program has provision for type of stuffing and amount of it. Ported and Sealed programs generally do not. So we can model a ported box with stuffing in the port by calling the the volume of the enclosure the small enclosure at the top of the line, and calling the port the Transmission Line.
It would be useful to download WinISD or any other quick simulator to get the dimensions of the port that tunes the box to a certain length. The Transmission Line Simulator does not have provision to tune by frequency, because it was designed to model Transmission Lines, not ported boxes.
A glitch: when you model a box of a certain size with a port of a certain size, the Transmission Line program for some reason shows it as tuning the box half an octave higher than any program, with a response curve to match. Why? Who knows? All I can say is to use Win ISD to model the box you wish, set WinISD to give you a port length that tunes the box half an octave, (.7 of the desired Fb) and transfer that port length and area to the Transmission Line model. The impedance curves of the Transmission Line model confirm that you will have the correct tuning when you use this method!!
Smaller glitch: even when you use the above method to get the correct tuning, the Transmission Line model will give you a ported box that is [1]slightly[/i] optimistic. A box that Win ISD will show an F3 of 30 Hz, for instance, will show up here with an F3 of maybe 28 Hz or so. Nothing to get excited about, just bear it in mind.
Incidentally, the separate program by Bullock and Whtie meat for modelling closed, vented and drone cone boxes do not have either of these glitches. Go figure.
A word about stuffing. This Bullock and White Transmission Line freeware program allows you to use specify both fiber size and density in the program. Since most speaker buildlers do not have data sheets full of fiber sizes of various densities hanging around the house, Bullock and White happily give us a default stuffing: it is DuPont Hollofil.
Hollofil is a polyester stuffing material that is used extensively in jackets, sleeping bags, pillows and comforters. You can also take comfort in the fact that it is nonallergenic, so you don't have to worry about your speakers making you sneeze. DuPont assures us that Hollofil has has at least 10% more fluffiness than that awful, grotesque unbranded polyester stuffing material that no sleeping bag manufacturer or pillow magnate would let himself be caught dead using.
http://www.dupont.com/insulations/product/hollofil808.html
Fiber size aside, the Transmission Line freeware modelling program has a nice feature called Packing Density, which tells you how many kilograms of stuffing per square meter the Transmission Line, (the vent, in our case) is stuffed with. That way, you can calculate what percentage of a cubic meter your vent is, and multiply a kilogram by that fraction to find the weight of stuffing your vent requires.
Those unfamiliar with the Metric System, (like every man, woman and child between Canada and Mexico), should know that a kilogram is 2.2 pounds, or 35 ounces, and a cubic meter equals 61,000 cubic inches.
The default Packing Density is 8 kilograms per meter, which the authors figure is what most Transmission Line builders pack their Lines with. However, I found the simulations work best with only 1 kilogram per cubic meter. This might create problems with packing the vent with a thin enough density. Perhaps only parts of the vent can be packed, and spaces in between left empty, like the manufacturer that Phase Accurate discovered who used an otherwise clear vent with one section full of open cell foam. Perhaps our Transmission Line experts can help us out here.
Anyway, I shall compare an imaginary, (but based on a Peerless 10" CSX) woofer in a 2 cubic ft. (56 liter) box tuned to 27 Hz with both stuffed, (at 1 kilogram per cubic meter), and unstuffed port. For cone excursion, I shall compare both to a sealed box with in a large enclosure with an F3 the same as the stuffed port box.
Hollofil is a polyester stuffing material that is used extensively in jackets, sleeping bags, pillows and comforters. You can also take comfort in the fact that it is nonallergenic, so you don't have to worry about your speakers making you sneeze. DuPont assures us that Hollofil has has at least 10% more fluffiness than that awful, grotesque unbranded polyester stuffing material that no sleeping bag manufacturer or pillow magnate would let himself be caught dead using.
http://www.dupont.com/insulations/product/hollofil808.html
Fiber size aside, the Transmission Line freeware modelling program has a nice feature called Packing Density, which tells you how many kilograms of stuffing per square meter the Transmission Line, (the vent, in our case) is stuffed with. That way, you can calculate what percentage of a cubic meter your vent is, and multiply a kilogram by that fraction to find the weight of stuffing your vent requires.
Those unfamiliar with the Metric System, (like every man, woman and child between Canada and Mexico), should know that a kilogram is 2.2 pounds, or 35 ounces, and a cubic meter equals 61,000 cubic inches.
The default Packing Density is 8 kilograms per meter, which the authors figure is what most Transmission Line builders pack their Lines with. However, I found the simulations work best with only 1 kilogram per cubic meter. This might create problems with packing the vent with a thin enough density. Perhaps only parts of the vent can be packed, and spaces in between left empty, like the manufacturer that Phase Accurate discovered who used an otherwise clear vent with one section full of open cell foam. Perhaps our Transmission Line experts can help us out here.
Anyway, I shall compare an imaginary, (but based on a Peerless 10" CSX) woofer in a 2 cubic ft. (56 liter) box tuned to 27 Hz with both stuffed, (at 1 kilogram per cubic meter), and unstuffed port. For cone excursion, I shall compare both to a sealed box with in a large enclosure with an F3 the same as the stuffed port box.
As I understand Pjotr's goals, he wants an enclosure that has
A) A bass rolloff similar to a second order, (12 dB/octave);
B) A substantial degree of output from the port (to provide cone excursion relief for the woofer);
C) Box size, sensitivity and F3 cutoff roughly comparable to a bass reflex box.
The simulations show that, within reason and with certain bass reflex setups, a stuffed port box can provide ALL of these requirements!
Basically, the simulations show that by properly stuffing the port, you can move the bass rolloff curve UP one order, (say from fourth to third), by sacrificing about a quarter octave of response at F3. So a speaker with an unstuffed port that has an F3 of 30 Hz and a 24dB/octave rolloff in the first octave under F3 will have, when you lightly stuff the port with 1kilogram per cubic meter of Hollofill, an F3 of 36 Hz and an 18 dB/octave rolloff in the first octave under F3.
Clearly then, the smart thing to do is to choose a bass reflex speaker/box combination that has a third order, (18 dB/octave) rolloff to start with. Then, when you move it up an order, (for newbies, 6 dB/octave is an "order"), you have something getting close to a second order, (12 dB/octave) rolloff, at least for the first octave under F3.
Such a combo-third order rolloff with unstuffed port-requires a speaker with a Qts of around 0.24. A little lower than average, but these woofers are available.
I have also seen where speaker author David Weems raised the Qts of a speaker by putting a resistor in series with it. Page 3 of this thread shows a reprint of part of a book chapter where he did this.
http://www.diyaudio.com/forums/show...page=15&highlight=double chamber&pagenumber=3
He did not state that you can lower the Qts of a speaker by putting a resistor in parallel with it, but the implication would seem to be there. Of course, the SPL of the speaker will be cut the more you lower the Qts, so if you must use this method, start with as low a Qts woofer as possible.
A) A bass rolloff similar to a second order, (12 dB/octave);
B) A substantial degree of output from the port (to provide cone excursion relief for the woofer);
C) Box size, sensitivity and F3 cutoff roughly comparable to a bass reflex box.
The simulations show that, within reason and with certain bass reflex setups, a stuffed port box can provide ALL of these requirements!
Basically, the simulations show that by properly stuffing the port, you can move the bass rolloff curve UP one order, (say from fourth to third), by sacrificing about a quarter octave of response at F3. So a speaker with an unstuffed port that has an F3 of 30 Hz and a 24dB/octave rolloff in the first octave under F3 will have, when you lightly stuff the port with 1kilogram per cubic meter of Hollofill, an F3 of 36 Hz and an 18 dB/octave rolloff in the first octave under F3.
Clearly then, the smart thing to do is to choose a bass reflex speaker/box combination that has a third order, (18 dB/octave) rolloff to start with. Then, when you move it up an order, (for newbies, 6 dB/octave is an "order"), you have something getting close to a second order, (12 dB/octave) rolloff, at least for the first octave under F3.
Such a combo-third order rolloff with unstuffed port-requires a speaker with a Qts of around 0.24. A little lower than average, but these woofers are available.
I have also seen where speaker author David Weems raised the Qts of a speaker by putting a resistor in series with it. Page 3 of this thread shows a reprint of part of a book chapter where he did this.
http://www.diyaudio.com/forums/show...page=15&highlight=double chamber&pagenumber=3
He did not state that you can lower the Qts of a speaker by putting a resistor in parallel with it, but the implication would seem to be there. Of course, the SPL of the speaker will be cut the more you lower the Qts, so if you must use this method, start with as low a Qts woofer as possible.
The imaginary speaker I used has:
Fs = 15 Hz
Qts = 0.24
Vas = 224 liters
SPL = 86.8 dB @ 1Meter/1Watt
The box is 56 liters tuned to 30 Hz.
Here are the Thiele-Small listings and screen shot of the Bullock-White Transmission Line Box Modelling Program, version 1.3. Shown is how to model between stuffed and unstuffed ports for the same box.
Fs = 15 Hz
Qts = 0.24
Vas = 224 liters
SPL = 86.8 dB @ 1Meter/1Watt
The box is 56 liters tuned to 30 Hz.
Here are the Thiele-Small listings and screen shot of the Bullock-White Transmission Line Box Modelling Program, version 1.3. Shown is how to model between stuffed and unstuffed ports for the same box.
Attachments
Here is the frequency response charts for the identical box with both a stuffed line and an unstuffed line.
Unstuffed line in Light Blue.
Stuffed Line in Red.
Sealed Box with a different woofer and larger enclosure, with a Qtc of 0.7, and a response and F3 identical to the stuffed line in Purple. This is for comparisons with cone excursion.
Unstuffed line in Light Blue.
Stuffed Line in Red.
Sealed Box with a different woofer and larger enclosure, with a Qtc of 0.7, and a response and F3 identical to the stuffed line in Purple. This is for comparisons with cone excursion.
Attachments
Impedance chart for stuffed and unstuffed port. Note how the stuffed port moves the resonance frequency down slightly. Remember the speakers and enclosures are identical except for the stuffing in the ports.
The unstuffed port resonance means the box is tuned to around 27 Hz, not 30 Hz as I stated earlier.
Note how the stuffed port also has a lower impedance at the high resonance peak. In other words, it is making the speaker act a little more like a closed box.
The unstuffed port resonance means the box is tuned to around 27 Hz, not 30 Hz as I stated earlier.
Note how the stuffed port also has a lower impedance at the high resonance peak. In other words, it is making the speaker act a little more like a closed box.
Attachments
Here, on the impedance phase, we see that the stuffed port acts very much like the unstuffed port.
The purple line is the closed box with the frequency response identical to the speaker with the stuffed port. Very, very different.
Unlike the impedance plot, there is little evidence the stuffed port makes the speaker act more like a closed box at all in this measurement.
The purple line is the closed box with the frequency response identical to the speaker with the stuffed port. Very, very different.
Unlike the impedance plot, there is little evidence the stuffed port makes the speaker act more like a closed box at all in this measurement.
Attachments
Pinkmouse:
I got started on this because I was playing around some months ago with Bullock and White's Transmission Line freeware and discovered that I could transform it into something could model a bass reflex system as well. Never thought it would do me any good until this thread came along.
Now that I have started down this path, I really should download Martin King's program and compare, especially since he is a member here.
The way I modelled the closed box, even though the Bullock and White freeware have no provision for it, is to
A) Punch in Thiele-Small number for a woofer with the SPL of 86.8, a Qts of .5 and an Fs that is one half octave below the stuffed stuffed vent's F3 of 30 Hz.
B) Made the box equal to Vas
C) Since the software has no provision for a closed box, I just made the Sd of the vent real, real small, and stuffed it fully. This lowered the tuning frequency to smething like 5 Hz and essentially gave the speaker the characteristics of a closed box in the octaves of interest.
Anyway, back to the speaker modelling salt mines. Next up: Cone Excursion!!
I got started on this because I was playing around some months ago with Bullock and White's Transmission Line freeware and discovered that I could transform it into something could model a bass reflex system as well. Never thought it would do me any good until this thread came along.
Now that I have started down this path, I really should download Martin King's program and compare, especially since he is a member here.
The way I modelled the closed box, even though the Bullock and White freeware have no provision for it, is to
A) Punch in Thiele-Small number for a woofer with the SPL of 86.8, a Qts of .5 and an Fs that is one half octave below the stuffed stuffed vent's F3 of 30 Hz.
B) Made the box equal to Vas
C) Since the software has no provision for a closed box, I just made the Sd of the vent real, real small, and stuffed it fully. This lowered the tuning frequency to smething like 5 Hz and essentially gave the speaker the characteristics of a closed box in the octaves of interest.
Anyway, back to the speaker modelling salt mines. Next up: Cone Excursion!!
Here we see the notch in the cone excursion is lessened in the stuffed port box, but it still gives substantial relief to the woofer.
The stuffed port speaker only has to move about half as much as the closed box speaker with identical passband response at 30 Hz, which is F3. The stuffed port box only has to move about 75% as the closed box speaker one half octave above F3-which is 42 Hz.
Again:
The unstuffed port is in Light Blue
The stuffed port is in Red
The closed box with frequency response identical, (down to F3) to the stuffed port is in Purple.
So here we see some evidence that the stuffed port acts somwhat like a closed box, but still retains most of the characteristics of the bass reflex.
The stuffed port speaker only has to move about half as much as the closed box speaker with identical passband response at 30 Hz, which is F3. The stuffed port box only has to move about 75% as the closed box speaker one half octave above F3-which is 42 Hz.
Again:
The unstuffed port is in Light Blue
The stuffed port is in Red
The closed box with frequency response identical, (down to F3) to the stuffed port is in Purple.
So here we see some evidence that the stuffed port acts somwhat like a closed box, but still retains most of the characteristics of the bass reflex.
Attachments
The Bullock and White TL software is based on Bradbury's moving fiber model. The software does not work very well for simulating a TL. Even Bullock, in his paper based on this program, noted that the correlation between test data and his calculations was not very good.
I tried the program a few years ago and even built up a MathCad model using the same moving fiber equations. The MathCad model and Bullock's software were in good agreement. This was the original model that I used to try and correlate against my early test line measurements. It did not work.
My advise would be not to use this program. But the final choice is for the TL DIYer to make.
I tried the program a few years ago and even built up a MathCad model using the same moving fiber equations. The MathCad model and Bullock's software were in good agreement. This was the original model that I used to try and correlate against my early test line measurements. It did not work.
My advise would be not to use this program. But the final choice is for the TL DIYer to make.
Pjotr,
In the limit, the lossier the port is, the closer the box resembles a sealed box. You will get a small lower impedance peak, which if you then stuff the crap out of the box, you will lower the upper peak as well. ?Then you will have an impedance curve which looks like what TL fans like to see.
Will it sound like a TL? - methinks it will sound like a damped lossy sealed box.
TL's are an interesting design problem, but I have not found them to sound any better than a properly designed conventional enclosure, which would probably be 1/3 the size. They may sound different, perhaps due to the comb filtering that is often present.
Different <> better.
In the limit, the lossier the port is, the closer the box resembles a sealed box. You will get a small lower impedance peak, which if you then stuff the crap out of the box, you will lower the upper peak as well. ?Then you will have an impedance curve which looks like what TL fans like to see.
Will it sound like a TL? - methinks it will sound like a damped lossy sealed box.
TL's are an interesting design problem, but I have not found them to sound any better than a properly designed conventional enclosure, which would probably be 1/3 the size. They may sound different, perhaps due to the comb filtering that is often present.
Different <> better.
Pjotr:
I have not yet gotten started on Martin King's program.
However, as I understand it, you want an enclosure type which gives the cone excursion relief while producing a second order rolloff.
Such an enclosure would appear to exist. I have not built one yet, though. However, it was written up in two articles in the Journal Of The Audio Engineering Society.
It is the Augmented Passive Radiator by Thomas J. Clarke.
I have a page missing from the first article. I can Email you what is left of the first article and all of the second if you want.
I discussed this enclosure at length on page 5 of the following thread. It has charts, frequency response simulations, and diagrams of the unit.
http://www.diyaudio.com/forums/showthread.php?threadid=2960&perpage=15&pagenumber=5
The first post I wrote about it is in the middle of page 5. It begins with the words: "Here is the enclosure I was talking about."
I would think this enclosure would be of real interest to you. It would seem to supply either
A) A quarter octave lower F3 compared to a normal bass reflex with the same driver and the same 24 dB/octave rolloff rate; or
B) The same F3 as the conventional bass reflex only with a 12 dB/octave rolloff instead of a 24 dB/octave rolloff rate.
Hope this helps.
I have not yet gotten started on Martin King's program.
However, as I understand it, you want an enclosure type which gives the cone excursion relief while producing a second order rolloff.
Such an enclosure would appear to exist. I have not built one yet, though. However, it was written up in two articles in the Journal Of The Audio Engineering Society.
It is the Augmented Passive Radiator by Thomas J. Clarke.
I have a page missing from the first article. I can Email you what is left of the first article and all of the second if you want.
I discussed this enclosure at length on page 5 of the following thread. It has charts, frequency response simulations, and diagrams of the unit.
http://www.diyaudio.com/forums/showthread.php?threadid=2960&perpage=15&pagenumber=5
The first post I wrote about it is in the middle of page 5. It begins with the words: "Here is the enclosure I was talking about."
I would think this enclosure would be of real interest to you. It would seem to supply either
A) A quarter octave lower F3 compared to a normal bass reflex with the same driver and the same 24 dB/octave rolloff rate; or
B) The same F3 as the conventional bass reflex only with a 12 dB/octave rolloff instead of a 24 dB/octave rolloff rate.
Hope this helps.
Hi kelticwizard
If you'd like to lower a driver's Qts (electrically via it's Qes) then you don't have to wire a resistor in parallel with the driver but a negative resistor in series with it !
There was once an issue of Wirleless World whose emphasis was on bass reproduction.
There was an author presenting a special tuning he had used:
He took a driver put it in a box whose volume was too small to even get a decent response as closed box, i.e. Qtc > 1, fc < 100 Hz.
He then ported the enclosure with quite a low fb (I think it was in the 30ies).
The result was a response with a hump above 100 Hz followed by a second order response down to the tuning frequency where it slowly changed into a fourth order response.
He then simply took a Linkwitz transform circuit to make it flat ......
I thought that his proposal was a bit extreme and played around with "perfect box". I haven't built such a box so far (I am more into "real" closed boxes at the moment) but in the simulations the following tuning didn't look that bad: You take the volume that would give a decent closed-box tuning (i.e. a Qtc between 0.5 and 0.8, depending on application and taste) and then tune the port about an octave lower than the fc of the closed box would be.
Regards
Charles
If you'd like to lower a driver's Qts (electrically via it's Qes) then you don't have to wire a resistor in parallel with the driver but a negative resistor in series with it !
There was once an issue of Wirleless World whose emphasis was on bass reproduction.
There was an author presenting a special tuning he had used:
He took a driver put it in a box whose volume was too small to even get a decent response as closed box, i.e. Qtc > 1, fc < 100 Hz.
He then ported the enclosure with quite a low fb (I think it was in the 30ies).
The result was a response with a hump above 100 Hz followed by a second order response down to the tuning frequency where it slowly changed into a fourth order response.
He then simply took a Linkwitz transform circuit to make it flat ......
I thought that his proposal was a bit extreme and played around with "perfect box". I haven't built such a box so far (I am more into "real" closed boxes at the moment) but in the simulations the following tuning didn't look that bad: You take the volume that would give a decent closed-box tuning (i.e. a Qtc between 0.5 and 0.8, depending on application and taste) and then tune the port about an octave lower than the fc of the closed box would be.
Regards
Charles
distributed port
hello,
i realize this reply is a bit untimely, but here goes.
the original prototype of the jbl aquarius 4 speaker employed a distributed port. i have one of these cabinets.
subsequent revisions of this speaker used a port tuned to about 40 hz.
the speaker employed an upward firing 8" driver (le8t) in the top of the enclosure, similar to one of the allison speakers. the top 2/3 of the cabinet was stuffed to the magnet of the woofer with batting. the lower 1/3 chamber had no stuffing at all. the distributed port consisted of 4 1" holes drilled through the cabinet.
the distributed port had a more "mellow" sound, similar to a TL.
hello,
i realize this reply is a bit untimely, but here goes.
the original prototype of the jbl aquarius 4 speaker employed a distributed port. i have one of these cabinets.
subsequent revisions of this speaker used a port tuned to about 40 hz.
the speaker employed an upward firing 8" driver (le8t) in the top of the enclosure, similar to one of the allison speakers. the top 2/3 of the cabinet was stuffed to the magnet of the woofer with batting. the lower 1/3 chamber had no stuffing at all. the distributed port consisted of 4 1" holes drilled through the cabinet.
the distributed port had a more "mellow" sound, similar to a TL.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.