Hello All
After reading a lot about TQWTs I will start to build my first pair. The dimensions of Cyburgs Needle are too small for the driver I would like to use (Alpair 10), so I created a new design.
Cyburgs Needle was simulated, built, tested and enjoyed by so many other people, therefore I think it may be sensible not to start from scratch but carefully adapt the dimensions of the Needle to the parameters of the new driver. By this way I may get a decent result, even I cannot run a simulation on that cabinet.
A summarisation of information I found on the internet and how I will apply those rules and the math to my project:
(Unfortunately I could not compress the text further)
1. Qt
Qt of the driver used in a TQWT should be between 0.2 and 0.7, best results can be expected if Qt=0.4 .. 0.5.
I will use the Alpair 10. Qt of that driver is 0.33, so it should work well in a TQWT. The outer diameter of the Alp10 is 16 cm, this defines the (internal) width of the cabinet. The internal width will be 18 cm.
The baffle will be removable.
2. Line Length
The Needle is tuned to 50Hz, the length of the middle line is about 178cm, a variation of 1 Hz equates a change of the line length of 3.56 cm. The Alpair 10 allows a deeper tuning frequency, I aim for tuning frequency of 45Hz.
New line length L = 178[cm] + 5[Hz] * 3.56[cm/Hz] = 196 [cm]
That method may not be perfect, but I think the error is not that big. The resulting tuning frequency may be 1-3 Hz deeper than the planned 45 Hz, but the Alpair 10 can handle that.
3. So
According to a rule of thumb (also used for the Needle) So should be
0.75 .. 1 x Sd.
For So = Sd = 90 [cm^2]
length lo = 90[cm^2] / 18[cm, internal width] = 5[cm]
4. Sl/So
Depending of the author, a ratio of 2.25 .. 5 is proposed for Sl/So, the Needle uses approx. 4.
I will use a ratio of 3 as I do not want a cabinet that is overly deep (even if I lose some bass).
Mouth Area = 3 * 90 [cm^2]= 270[cm^2]
length at mouth ll = 3 * 5[cm] = 15[cm]
I will use 19mm plywood for the box and the divider, so the internal depth of the TQWT is
1.9+5+15 = 21.9 [cm]
5. Length of the Divider
-1.9 length of the middle line is shortened by an additional piece of wood
+ l_Divider
+ 21.9 rectangular shaped length of the middle line in the upper part of t. box
+ l_Divider
+ 1.9 reinforcement on top of the opening
+ 5 opening of the Helmholtz Resonator, adaption neccessary!
= 196 length of the middle line
l_Divider = 84.5[cm]
6. Calculation of the Geometry of the TQWT
I will use the Intercept Theorem (Strahlensatz) for the calculation of internal measurements. The included drawing should explain the calculation. First I calculate the length of the hypotenuse (m_complete) of the unfolded TQWT, as that simplifies further calculations:
7.5/2.5 = m_complete /(m_complete - 196)
m_complete= 294[cm]
The angle of the triangle is small (2.9°) so it is not neccesary to seperately calculate the cathetus (for the baffle).
15/294 = l(x) /x
l(x)=15 * x /294 where x is the point on the middle line measured from O.
l(x) is the distance between the divider and the baffle (or the other outside walls)
A(x)= 18 * 15*x /294 A(x) is the Area of the TQWT depending from the position on the middle line
I apply that formular to So (the measurement is already known)
l(x)=15*x /294
x=294 - 196= 98[cm]
#Important: x= length of the hypotenuse - line length
l(x) at So is 5 cm.
Horizontal distance between baffle and divider:
x=98[cm] + 84.5[cm]-1.9[cm]=180.6[cm]
l_1=15 * 180.6 /294= 9.2[cm]
Vertical distance between divider and the top of the cabinet:
x=98[cm] + 84.5[cm] -1.9[cm] + 10[cm] = 190.6 [cm] (simplified calculation)
l_2=15 * 190.6 /294= 9.7[cm]
Now I can calculate the total internal height of the box:
H = 9.7[cm] + 84.5[cm] + 1.9[cm] + 5[cm] = 101.1[cm]
The internal dimensions of the box are 101.1 x 18 x 21.9 cm
and the capacity is ca. 36.4 liter
7. Adaption of the Helmholtz Resonator
The opening must be adapted to the tuning frequency and the capacity of the cabinet.
There is also a rule of the thumb (the area of the port should be 60 to 80% of Sd), but I think the result is very likely to be faulty.
The dimensions of the port can easily be messed up, as small deviations have large effects. I decided to start with a opening that is too big (18 x 5 x 8.8 cm), but the overdimensionised port can easily be recuced. I added 2 Maple4 scripts that can be used to calculate the dimensions of the Helmholtz Resonator.
A port (1.8 x 18 x 8.8 cm) will result in a tuning frequency of 45 Hz, but I am not sure if there will be streaming noises caused by the small height of 1.8 cm of the opening.
If I recuce the width of the port of 2x1.9 cm, a tuning frequency of 45 Hz can be reached by following port dimemsions: 2.4 x 14.4 x 8.8
I think it is best to tune the port when listening to music.
8. Driver Position
The driver should be positioned at 0.33 to 0.45 of the line length, according to the internet.
First I will position the driver at the same hight as the upper end of the divider, the resulting Xi is 0.43
If the result is unsatisfactory I will create a test-baffle, cut it in segments of 2,3,5,7,11,13, ... cm, so I can test the effect of different positions of the driver just by changing the order of the segments.
9. List of Parts
2 Back: 104.9 x 18
2 Baffle 96.1 x 18
2 Divider 84.5 x 18
2 Bottom 23.8 x 18
2 Top1 25.7 x 21.8
2 Top2 21.9 x 18
2 Port1 5.0 x 18
2 Port2 8.8 x 18
4 Side 106.8 x 25.7
Thickness of all parts: 1.9 cm
The list does not contain the wood for reducing the port.
Comments and critisim of this project is welcome!
Peter
After reading a lot about TQWTs I will start to build my first pair. The dimensions of Cyburgs Needle are too small for the driver I would like to use (Alpair 10), so I created a new design.
Cyburgs Needle was simulated, built, tested and enjoyed by so many other people, therefore I think it may be sensible not to start from scratch but carefully adapt the dimensions of the Needle to the parameters of the new driver. By this way I may get a decent result, even I cannot run a simulation on that cabinet.
A summarisation of information I found on the internet and how I will apply those rules and the math to my project:
(Unfortunately I could not compress the text further)
1. Qt
Qt of the driver used in a TQWT should be between 0.2 and 0.7, best results can be expected if Qt=0.4 .. 0.5.
I will use the Alpair 10. Qt of that driver is 0.33, so it should work well in a TQWT. The outer diameter of the Alp10 is 16 cm, this defines the (internal) width of the cabinet. The internal width will be 18 cm.
The baffle will be removable.
2. Line Length
The Needle is tuned to 50Hz, the length of the middle line is about 178cm, a variation of 1 Hz equates a change of the line length of 3.56 cm. The Alpair 10 allows a deeper tuning frequency, I aim for tuning frequency of 45Hz.
New line length L = 178[cm] + 5[Hz] * 3.56[cm/Hz] = 196 [cm]
That method may not be perfect, but I think the error is not that big. The resulting tuning frequency may be 1-3 Hz deeper than the planned 45 Hz, but the Alpair 10 can handle that.
3. So
According to a rule of thumb (also used for the Needle) So should be
0.75 .. 1 x Sd.
For So = Sd = 90 [cm^2]
length lo = 90[cm^2] / 18[cm, internal width] = 5[cm]
4. Sl/So
Depending of the author, a ratio of 2.25 .. 5 is proposed for Sl/So, the Needle uses approx. 4.
I will use a ratio of 3 as I do not want a cabinet that is overly deep (even if I lose some bass).
Mouth Area = 3 * 90 [cm^2]= 270[cm^2]
length at mouth ll = 3 * 5[cm] = 15[cm]
I will use 19mm plywood for the box and the divider, so the internal depth of the TQWT is
1.9+5+15 = 21.9 [cm]
5. Length of the Divider
-1.9 length of the middle line is shortened by an additional piece of wood
+ l_Divider
+ 21.9 rectangular shaped length of the middle line in the upper part of t. box
+ l_Divider
+ 1.9 reinforcement on top of the opening
+ 5 opening of the Helmholtz Resonator, adaption neccessary!
= 196 length of the middle line
l_Divider = 84.5[cm]
6. Calculation of the Geometry of the TQWT
I will use the Intercept Theorem (Strahlensatz) for the calculation of internal measurements. The included drawing should explain the calculation. First I calculate the length of the hypotenuse (m_complete) of the unfolded TQWT, as that simplifies further calculations:
7.5/2.5 = m_complete /(m_complete - 196)
m_complete= 294[cm]
The angle of the triangle is small (2.9°) so it is not neccesary to seperately calculate the cathetus (for the baffle).
15/294 = l(x) /x
l(x)=15 * x /294 where x is the point on the middle line measured from O.
l(x) is the distance between the divider and the baffle (or the other outside walls)
A(x)= 18 * 15*x /294 A(x) is the Area of the TQWT depending from the position on the middle line
I apply that formular to So (the measurement is already known)
l(x)=15*x /294
x=294 - 196= 98[cm]
#Important: x= length of the hypotenuse - line length
l(x) at So is 5 cm.
Horizontal distance between baffle and divider:
x=98[cm] + 84.5[cm]-1.9[cm]=180.6[cm]
l_1=15 * 180.6 /294= 9.2[cm]
Vertical distance between divider and the top of the cabinet:
x=98[cm] + 84.5[cm] -1.9[cm] + 10[cm] = 190.6 [cm] (simplified calculation)
l_2=15 * 190.6 /294= 9.7[cm]
Now I can calculate the total internal height of the box:
H = 9.7[cm] + 84.5[cm] + 1.9[cm] + 5[cm] = 101.1[cm]
The internal dimensions of the box are 101.1 x 18 x 21.9 cm
and the capacity is ca. 36.4 liter
7. Adaption of the Helmholtz Resonator
The opening must be adapted to the tuning frequency and the capacity of the cabinet.
There is also a rule of the thumb (the area of the port should be 60 to 80% of Sd), but I think the result is very likely to be faulty.
The dimensions of the port can easily be messed up, as small deviations have large effects. I decided to start with a opening that is too big (18 x 5 x 8.8 cm), but the overdimensionised port can easily be recuced. I added 2 Maple4 scripts that can be used to calculate the dimensions of the Helmholtz Resonator.
A port (1.8 x 18 x 8.8 cm) will result in a tuning frequency of 45 Hz, but I am not sure if there will be streaming noises caused by the small height of 1.8 cm of the opening.
If I recuce the width of the port of 2x1.9 cm, a tuning frequency of 45 Hz can be reached by following port dimemsions: 2.4 x 14.4 x 8.8
I think it is best to tune the port when listening to music.
8. Driver Position
The driver should be positioned at 0.33 to 0.45 of the line length, according to the internet.
First I will position the driver at the same hight as the upper end of the divider, the resulting Xi is 0.43
If the result is unsatisfactory I will create a test-baffle, cut it in segments of 2,3,5,7,11,13, ... cm, so I can test the effect of different positions of the driver just by changing the order of the segments.
9. List of Parts
2 Back: 104.9 x 18
2 Baffle 96.1 x 18
2 Divider 84.5 x 18
2 Bottom 23.8 x 18
2 Top1 25.7 x 21.8
2 Top2 21.9 x 18
2 Port1 5.0 x 18
2 Port2 8.8 x 18
4 Side 106.8 x 25.7
Thickness of all parts: 1.9 cm
The list does not contain the wood for reducing the port.
Comments and critisim of this project is welcome!
Peter
Attachments
A rule of thumb that has proven to be pretty irrelevant, critical box dimensions are a function of Vas & Fs not Sd.
Positioning to kill the 1st undesirable pipe resonance is fairly critical. Needs to be withing cms if not closer.
You would do well do get Martin King's worksheets instead of guessing. Or contact Scott (scottmoose) or I about doing a beta test build of the Woden Design Samhain.
dave
Just found my way into this thread because I´d like to know what happens if you change the lenght of the TQWT line - like Peter said - to go for a deeper tuning frequency. I´m using the online TQWT calculator on this site mh-audio Tapered Quarter Wave Tube but they recommend that "the length of the TQWT should be agreed on about 1.0 to 1.2 x resonant frequency of the speaker".
I want to use a driver for my project with a resonant frequencyof 75 Hz. Calculation ends up with a pretty small enclosure at 50 Hz. But from the specs of my driver I know it can handle 40 Hz. So by changing the tuning frequency to 55 Hz, I´m pretty close to the intended result.🙂
But will it work to go that way?😕
I want to use a driver for my project with a resonant frequencyof 75 Hz. Calculation ends up with a pretty small enclosure at 50 Hz. But from the specs of my driver I know it can handle 40 Hz. So by changing the tuning frequency to 55 Hz, I´m pretty close to the intended result.🙂
But will it work to go that way?😕
Last edited:
I'd suggest at least looking at Martin's alignment tables. They do not thou do a ML-V (aka ML-TQWT), you'll need a copy of his worksheets for that. Mass loading adds some more poles to the terminus low pass significantly reducing ripple.
dave
dave
This is a calculator from Italy? I tried it, producing pretty strange results
Dave, results from the online calculator in the link I´ve posted base on the MK formulas. I saw a comparison between the calculator and the sheets, pretty much the same. It would be nice to have the alignment tables implemented in some sort of self-contained software in the future to make it easier to work with. I would gladly pay for something like that.😎
There is a construction picture in the calculator link on the right that looks like a ML-TQWT design.😕 Enclosure dimensions don´t change, only "internal setup"... maybe worth to experiment with.
I would still like to know if anyone extended the TQWT line to go for a deeper tuning frequency (in a way the driver is still able to handle) and how it worked.
- Status
- Not open for further replies.