Double Driver Bass reflex design

[endrek]

Hi,
if i would like to build a double driver bassreflex from this box (single driver bassreflex):

http://www.faneinternational.com/downloads/Cabinets18inch200Lres.pdf


how would i have to modify the design? How will the T/S parameters change, how can i do the calculations of the formulas, etc?? any help please??




pd: i know simulators do the calculations for you, but i'd like to know wich formulas to use, and how to use them. Because i have to write about how i would do it

thanks to all!!

 

[zobsky]

DCR
Project


Basically, for a given single reflex design of volume V and tuning frequency f
1. divide the box into two chambers in the ratio 2:1 i.e. 2/3V and 1/3V
2. Instead of using a single port tuned to f, ..calculate the equivalent port size and length using two ports. Prepare 3 ports of these dimensions
3. Port each chamber to the outside using one port per chamber
4. Connect the internal chambers with the third port

You may have to fudge the volumes a bit to account for the extra port and dividing baffle volume but thats the principle

 

[zobsky]

DCR
Project


Basically, for a given single reflex design of volume V and tuning frequency f
1. divide the box into two chambers in the ratio 2:1 i.e. 2/3V and 1/3V
2. Instead of using a single port tuned to f, ..calculate the equivalent port size and length using two ports. Prepare 3 ports of these dimensions
3. Port each chamber to the outside using one port per chamber
4. Connect the internal chambers with the third port

You may have to fudge the volumes a bit to account for the extra port and dividing baffle volume but thats the principle

I misread your initial post. If your intent is to using 2 drivers in the mentioned enclosure, this is probably not a good idea

 

[Arty]

basicly the box will be 2 times larger.
Take Vas and double it.

Tuning freqvency should remain the same. But naturaly the port crossectional area needs to be doubled. Allso, You need to re-do the math to find a starting point on how long should your port be.
That is, if it is not an isobaric build You want.
/and You do not want that.

 

[weltersys]

Hi,
if i would like to build a double driver bassreflex from this box (single driver bassreflex):

http://www.faneinternational.com/downloads/Cabinets18inch200Lres.pdf

how would i have to modify the design? How will the T/S parameters change, how can i do the calculations of the formulas, etc?? any help please??

pd: i know simulators do the calculations for you, but i'd like to know wich formulas to use, and how to use them. Because i have to write about how i would do it

thanks to all!!

No calculations are needed, simply build double the box, each with the same ports as the original design, sharing a center wall.

You can eliminate the center wall, but distortion will go up slightly.
The center wall can be made of thinner material to save a little weight.

If you have to write about it, I suggest you use a spell checker

 

[endrek]

i have to write it with my own language Spanish, and i write it a little bit better than English

But anyways i'll use a spell check just to please you


Well here i have my design, could you check it? And tell me if i'm completely wrong? nearly nearly? or good enough to do a test?

I got one drivers T/S parameters (i know that when i buy the driver Fane Colossus 18XB, i'll calculate T/S parameters instead using the spec from manufacturer ) Cause i want to drivers 2*Vab and 2*Sd right? I'm not sure how to predict, or mesure how Fs will change once they are together (electrically in paralel) any ideas please??

Now i can calculate the Vab,Fb,Fl,Lv, Sv, etc.. for a C4 Alignment for a QL = 3, that gives me a volume Vb(L) =442,14539 . Wich i can verify that my suposition of QL was correct because Vb>80L (atleast this is how my book says to verify your QL)
Important parameter is Lv=0,10997 or 0,112368 or 0,0755 (with constant=23500 proposed by Small) ..... I think i'll get try the first one and start trimming while tunning as they've told me at this forum

Then i divide the Sv by 6 (Sv/6=Sv') because i want to make 6 cilindrical ports. That gives me Sv' and Dv'=0,103m I'll keep the Lv the same right? Or i divide it by 6 too? :S

Well, with a volume of Vb(L)= 442,1453 L and using the golden ratios of 0,62x1x1,62 i get a box dimensions of W=1,23m , H=0,47 and D=0,76 , wich are really bad dimensions, because my driver has a diameter of 0,48m and i kind of draw it in paper and the 2 drivers with the 6 ports don't really fit in 1,23 of width. So i changed a bit the dimensions, keeping the same volume W=1,29m , H=0,54 and D=0,63 .

I think its really really big, and before i start building it, and spent money with the woods, i would like to know some opinions, because this is my first design/construcion of a speaker. And maybe i'm just going completely wrong
here i leave a table i copied from my excel calculations:

T/S param
1 driver Colossus 18XB
2 drivers Colossus 18XB
Fs
33 Hz
33 Hz
Qts
0,337
0,337
Vas
236
472
Xmax
7,5e-3
7,5e-3
a(m)
0,2056605
0,26833131
d (2a ) (m)
0,41151
0,53666266
Sd (m^2)
0,1131
0,2262


C4 alignment
α
1,2616
h
1,1034
q
1,2798
Vab (L)
374,128091
Fb (Hz)
36,4122
Fl (Hz)
42,2334
Vs (L)
34,00864933
Vb (L)
442,14539
Vdmax (m^3)
1,70E-03
dv (m)
0,194066525è0,25
Sv(m^2)
((0,25/2)^2)*Pi=0,05
Lv (op1)
0,10997276
Lv (op 2)
0,11233682
















Design with 6 ports

Sv’ (6ports)=Sv/6èDv’ (m)
0,103m


Dimensions

W
1,29m
H
0,54m
D
0,63m


Pd: any advice for construction? wich cheap wood is good enough for my test-speaker and won't influence that is cheap on the results?? Wich material should i use for the ports?? Rigid PVC right? any thickness better than other?

And another dumbie question i just though, to trim the ports while tunning to the Fb, do i glue them all time and test? Or how do i stick them, test, and cut ?? Any good method??

Sorry for this long long post and if there are too many questions, and if there are not really smart I'm new, and my english either makes me look dumber

Thanks to all and for all. Really apreaciate your help and this cool forum
You are all invited to our parties if the speakers project succeed

cheerz from Spain

 

[weltersys]

Well here i have my design, could you check it? And tell me if i'm completely wrong? nearly nearly? or good enough to do a test?

Pd: any advice for construction? wich cheap wood is good enough for my test-speaker and won't influence that is cheap on the results?? Wich material should i use for the ports?? Rigid PVC right? any thickness better than other?

And another dumbie question i just though, to trim the ports while tunning to the Fb, do i glue them all time and test? Or how do i stick them, test, and cut ?? Any good method??

Sorry for this long long post and if there are too many questions, and if there are not really smart I'm new, and my english either makes me look dumber

Thanks to all and for all. Really apreaciate your help and this cool forum
You are all invited to our parties if the speakers project succeed

cheerz from Spain

endrek,

Sorry, don’t have time to closely check your work.
If you doubled the cabinet volume and doubled the ports leaving them the same size as the Fane 200L, your dual cabinet should be the equivalent of two of the Fane 200L cabinets.

Fs will go down slightly in a dual compared to a single, but that does not change Fb.

If you changed port lengths and depths to “equivalent “ values, they may or may not result in an equivalent tuning (Fb). Details like bracing strip position and distance of the port to walls will affect tuning a bit.

I almost always use ports made of plywood, in various shapes, triangular, or square are preferred, though in some cabinets a wide rectangular port fits best, in which case dividing the rectangle in to more square like portions gives better results.

These ports usually need some adjustment in length to match predicted Fb using circular ducts, so I make the ports longer initially, then cut them down until the cabinet tuning is correct.

Rounding the ends reduces port noise.
PVC or cardboard ducts work OK, but don’t provide much area on the inside to round off.
Putting a larger circle ( or whatever port shape used) around the inside of a port improves the high volume response (less port compression) but also makes the port appear “longer” than predicted.

I would suggest building your box from good plywood, there is no reason the “test box” can’t be used as the “real” box. The more plys the better, 13 ply 19 mm Baltic Birch would be the best material if you can afford it.

Different wood will sound slightly different, MDF requires more bracing than plywood to be as stiff, but is heavier, and slightly more inert than lightweight plywood.

Use plenty of glue and screws, brace all the large panels.

To check Fb start by painting a white dot on the cone.
Using a sine wave frequency generator into an amplifier connected to the speaker, choose a frequency about 10 Hz below Fb. Turn up the level until you see cone movement less than Xmax (less than 15mm peak to peak for your speakers).

Now raise the frequency until the cone movement is at minimum.
Cone movement will be at minimum at Fb. If the Fb is too low, cut off the ports a bit, and retest. If Fb was too high to start with, you will need longer ports.

Fewer large ports make this process easier than using multiple small ports.

 

[Arty]

to verify tuning freqvency You can allso mesure impedance.
No paint needed.

 

[weltersys]

to verify tuning freqvency You can allso mesure impedance.
No paint needed.

Arty,

That is true, but requires either an impedance test tool, or a VOM, resistor, clip leads, verifying the VOM reads properly at different frequencies (and recalibrating at different frequencies when it does not), and a lot more patience than I have.

I would like to own a good impedance tester, but have not been able to justify the expense.

I'll continue putting a tiny dot on the cone with a silver sharpie, as I like to verify the excursion over the sub range, and listen for any cabinet leaks, port noise, rattles, resonances and distortion.

Art

 

[endrek]

really thankfull for all

have a lot of work to do

big up free education for all
really appreciate it

 

[DSP_Geek]

I recommend an internal partition between the two drivers because the maximum internal dimension will be smaller. Consider a 210 litre box per driver, with internal dimensions (ignoring the port for now) of 70 cm x 60 cm x 50 cm. Now double the height of the box: the longest internal measurement is now 1.4 metres, which means the first half-wave resonance (the nodes would be at the top and bottom boundaries) will occur about 120 Hz. On the other hand, splitting the box in two internally means that resonance now occurs at 240 Hz, which is more tractable for damping without messing up your Qb too badly for port tuning.

In any event, one can treat a system with two woofers as the moral equivalent of two systems with one woofer each. Joining them together means you have twice the speakers, twice the box volume, and twice the duct area but the same duct length.

Also consider this: a single box with two 18 inch drivers will be HEAVY. Be really sure you want to drag this thing around.

 

[Arty]

Arty,

That is true, but requires either an impedance test tool, or a VOM, resistor, clip leads, verifying the VOM reads properly at different frequencies (and recalibrating at different frequencies when it does not), and a lot more patience than I have.

I would like to own a good impedance tester, but have not been able to justify the expense.

I'll continue putting a tiny dot on the cone with a silver sharpie, as I like to verify the excursion over the sub range, and listen for any cabinet leaks, port noise, rattles, resonances and distortion.

Art

if you have acces to ANY amplifier and a simple dmm, you can do it with ease.
R is still U / I.
Most probably You know the range where You are searching for an impedance peak, a simple computer based sine wave generator can help You.
A resistor can help allso, say a 5 watt one in series with the speaker.
Anything around 4 ohm will do.
You just mesure the voltage depending on the freqvency, and You will be able to recognise the tuning freqvency. Not at all hard to do.

Anyways i just wrote this since i do not realy think an impedance test tool is required to check tuning freqvency. The exact value of the impdeance is not important. Just the freqvency where the peak(s) is(are).