what do u think??
after reading about '' the acoustic wave cannon'' thread.. it made me bring up the mathcad files..
im using jbl2226h
1st question
The port diameter,does this need to be large like a ported box to prevent chuffing//
2nd question
Can i presume that 'area at large end of ML TQWT ' = the driver end?? because normaly he talks of Reducing area tlines..
also experimenting with Driver position ratio..
that image=
S0=1Sd
SL=7sd 880cm x7 = big hah
ah i fixed the peak,, i made SL=4x Sd so its flat unlike the image..
3rd question
can anyone give me some pointers? its good to explore posibilities not just build the regular ported box
what do u think??
4thquestion
is the Fs of the speaker,actualy lowered like it says in my old voigt pipe articles?>
http://homepages.paradise.net.nz/quadroph '' articles link''
*http://www.geocities.com/sc00byd0159/scans.html*
🙂
after reading about '' the acoustic wave cannon'' thread.. it made me bring up the mathcad files..
im using jbl2226h
1st question
The port diameter,does this need to be large like a ported box to prevent chuffing//
2nd question
Can i presume that 'area at large end of ML TQWT ' = the driver end?? because normaly he talks of Reducing area tlines..
also experimenting with Driver position ratio..
that image=
S0=1Sd
SL=7sd 880cm x7 = big hah
ah i fixed the peak,, i made SL=4x Sd so its flat unlike the image..
3rd question
can anyone give me some pointers? its good to explore posibilities not just build the regular ported box
what do u think??
4thquestion
is the Fs of the speaker,actualy lowered like it says in my old voigt pipe articles?>
http://homepages.paradise.net.nz/quadroph '' articles link''
*http://www.geocities.com/sc00byd0159/scans.html*
🙂
Attachments
mikee12345 said:
You say you've smoothed out the bottom a bit from the posted pics... can you post that simulation?
Looks like you can afford a bit more stuffing. this will reduce the rolloff and give you a little more roll-off in the bottom (which will help compensate for room gain)
dave
yep i reducd it to 3x Sd,makes the peak,flat otherwise the same.
also,3x Sd makes it more 'pipe ish' in my opinion,where as 7x Sd is basshorn size :-D
and i cant fit a basshorn,or only a 40hz one anyway-this is another possibility,but only if i could make a nautilus shell style one to cram it all into small space.
.is that all a tline is------a small volume/area pipe basshorn
i have LOTS of wool if i decide to build this :-D
im now on the 133mhz without mathcad lol,but i will be looking at these files again on the weekend, with the 1.6ghz
🙂
also,3x Sd makes it more 'pipe ish' in my opinion,where as 7x Sd is basshorn size :-D
and i cant fit a basshorn,or only a 40hz one anyway-this is another possibility,but only if i could make a nautilus shell style one to cram it all into small space.
.is that all a tline is------a small volume/area pipe basshorn
i have LOTS of wool if i decide to build this :-D
im now on the 133mhz without mathcad lol,but i will be looking at these files again on the weekend, with the 1.6ghz
🙂
Yes. As frequency decreases, more of the output is from the port. I like to use 0.5*Sd as a minimum port size.
If you are using one of the ML TQWT worksheets, be careful. MatCAD doesn't care about "small end", "large end". So is the driver end, Sl is the port end, regardless of your taper.
You don't want flat to cut-off. It will boom like H...! Make a guess at what your room lift is going to be and reduce the output at the bottom end by the amount of room lift.
🙂
Absolutey not. Fs is Fs, Qts is Qts. You can safely ignore everything that has been written about "Voigt" pipes prior to MJK's worksheets. All of the old advise ranges for dead wrong to simply misguided.
Bob
thanks heaps Bob!!!
also thats good,making it a more gradual roll off than 'flat'..
makes the end less area i think..il get back to the files on the weekend:-D also better sound,reminds me of open baffle roll off vs ported 4th order.lower the order better sound less phase change etc?
its good to get rid of old myths that hang around isnt it!
i wonder why they thought the Fs did lower in a low tuned pipe.
eg the Fs was 100hz or so,and it lowered to near Fb of pipe..(is what they say)strange😛
but i think that they are not understand air mass on the speaker,,and also the response graph shows a roll off below 100hz as if the driver is put in a too low tuned box 😛
what do you think of the old articles on the taper of the pipes? it seemed to agree with the mathcad files,i think.. i havent looked into it enough.
i know that using a smaller speaker would make it Much more easier
i have a 5x4m room ,and a 2.2x3.8metre room it could go in.anyways,ive finished the ported box plans..time to go get it cut on friday :-D
😎
also thats good,making it a more gradual roll off than 'flat'..
makes the end less area i think..il get back to the files on the weekend:-D also better sound,reminds me of open baffle roll off vs ported 4th order.lower the order better sound less phase change etc?
its good to get rid of old myths that hang around isnt it!i wonder why they thought the Fs did lower in a low tuned pipe.
eg the Fs was 100hz or so,and it lowered to near Fb of pipe..(is what they say)strange😛
but i think that they are not understand air mass on the speaker,,and also the response graph shows a roll off below 100hz as if the driver is put in a too low tuned box 😛
what do you think of the old articles on the taper of the pipes? it seemed to agree with the mathcad files,i think.. i havent looked into it enough.
i know that using a smaller speaker would make it Much more easier
i have a 5x4m room ,and a 2.2x3.8metre room it could go in.anyways,ive finished the ported box plans..time to go get it cut on friday :-D
😎
ok this time..
100inches
S0=3
SL=1
Density 0.05 (just to see)
it would only be used up to 70hz or lower anyway.
8inch port,0.5inches length
is a S0=1 SL=6 - a (conical) basshorn ??
the transient response looks jagged.
if i want a system that rolls off,then i would build a shiva in 50litres sealed 😛
but i know yeh,its good to keep it smooth
if i change it to be S0=4, Sl=1, the response just goes lower in freq but same amplitude which is good if i want extension.
100inches
S0=3
SL=1
Density 0.05 (just to see)
it would only be used up to 70hz or lower anyway.
8inch port,0.5inches length
is a S0=1 SL=6 - a (conical) basshorn ??
the transient response looks jagged.
if i want a system that rolls off,then i would build a shiva in 50litres sealed 😛
but i know yeh,its good to keep it smooth
if i change it to be S0=4, Sl=1, the response just goes lower in freq but same amplitude which is good if i want extension.
Attachments
Let me amend my comments a little. Juha Bachman wrote and AES article using computer sims that hinted how to do TL's right. George Augspurger did an AES article that is very enlightening. He then expanded that article and had a three-part article printed in Speaker Builder (2000 I believe). His computer program is not generally available, but I am told by those who know that given the same input GA's program produces nearly identical results to MJK's MathCAD worksheets.
GA's articles presume the pipe resonance (Fp) is higher than the driver Fs. The result is that the pipe has a much lower Q and the total ouput FR has a much more gradual roll-off. My approach, which mirrors MJK's, put Fp lower tha Fs. This makes for a high pipe Q and a rapid system roll-off. GA's approach is best used when the driver Fs is low, say in the 20's. MJK's approach is best when you are trying to get the lowest possible F3 out of a small driver. While MKJ's approach is certainly usable with a large driver, it does lead to really large cabinets. What you need to consider is whether the price/performance ratio of a TL is warranted in comparison to, say, a stereo pair of Shiva's in Adire's recommended 2ft3 sealed box.
Bob
GA's articles presume the pipe resonance (Fp) is higher than the driver Fs. The result is that the pipe has a much lower Q and the total ouput FR has a much more gradual roll-off. My approach, which mirrors MJK's, put Fp lower tha Fs. This makes for a high pipe Q and a rapid system roll-off. GA's approach is best used when the driver Fs is low, say in the 20's. MJK's approach is best when you are trying to get the lowest possible F3 out of a small driver. While MKJ's approach is certainly usable with a large driver, it does lead to really large cabinets. What you need to consider is whether the price/performance ratio of a TL is warranted in comparison to, say, a stereo pair of Shiva's in Adire's recommended 2ft3 sealed box.
Bob
here in NZ we dont get speakerbuilder,i think im missing out on alot!! 
i worked out the approximate dimensions of the TLine, 560litres lol!
as much as a labhorn
but the labhorn has disadvantages of course
-woodworking skill
-2x expensive drivers that are unavailable here
perhaps replace them with 2x XLS peerless units.
you can buy AES articles for 4$ US ea cant you???
i mite buy some ,sometime
as soon as i buy my wood for my 175litre Fb=33hz box,and my SPL meter, etc etc 😛
thanks bob :-D
i worked out the approximate dimensions of the TLine, 560litres lol!
as much as a labhorn
but the labhorn has disadvantages of course
-woodworking skill
-2x expensive drivers that are unavailable here
perhaps replace them with 2x XLS peerless units.
you can buy AES articles for 4$ US ea cant you???
i mite buy some ,sometime
as soon as i buy my wood for my 175litre Fb=33hz box,and my SPL meter, etc etc 😛
thanks bob :-D
Hey Mikee, What specs did you use? There's no way a driver with a Qts = 0.31 and a ~3.27% eff. is going to have an IB slope response like shown on your graphs. It's going to be off the plot before it hits 200Hz.....
FWIW, this driver doesn't have specs suitable for a low tuned ML-TL or TL for that matter, except with some serious digital EQ, so a standard ol' golden ratio 6.2ft^3 net tuned to between 18-20Hz with a pair of 18" PRs will net you ~5-6dB/20Hz over the same size sealed. Considering the cost of the PRs, that's a fairly expensive extra few dB.
Really, since it has good basshorn specs, a bit larger BP with a large, long damped vent tube might be your best bet for simple Vs SPL.
GM
FWIW, this driver doesn't have specs suitable for a low tuned ML-TL or TL for that matter, except with some serious digital EQ, so a standard ol' golden ratio 6.2ft^3 net tuned to between 18-20Hz with a pair of 18" PRs will net you ~5-6dB/20Hz over the same size sealed. Considering the cost of the PRs, that's a fairly expensive extra few dB.
Really, since it has good basshorn specs, a bit larger BP with a large, long damped vent tube might be your best bet for simple Vs SPL.
GM
i used Qes=0.301 and Qms=5.5 they are approximate values since i lost the pdf
yes i do expect it to reduce a few db,in winisd it prefers an enclosure tuned above resonance and small box,highly efficient.where as my 175litre 33hzFB usage is not so efficient .
http://homepages.paradise.net.nz/quadroph 'tline'
i actualy prefer the basshorn but it seems to unload badly.
mmm efficiency.
yes i do expect it to reduce a few db,in winisd it prefers an enclosure tuned above resonance and small box,highly efficient.where as my 175litre 33hzFB usage is not so efficient .
http://homepages.paradise.net.nz/quadroph 'tline'
i actualy prefer the basshorn but it seems to unload badly.
mmm efficiency.
OK, here's the averaged specs I have on the 2226H:
Fs = 40Hz
Vas = 175.58L
Qms = 5.0 (8ohms)
Qes = 0.33 (0.528)
Qts = 0.30956 (no need to input, the spreadsheet calculates it)
Le = 1.75mH
Pe = 600W
Re = 5ohms
Sd = 883cm^2
Xmax = 7.6mm
B*L = 19.441N/A
Cms = 1.5951e-04mm/N
Mms = 0.0993Kg
Plugging these into MLTQWT and Hornresp should give you a bit more accurate sim. Not as good as measuring yours, but should be in the ballpark.
Mass loaded pipe (MLTQWT, or ML-TL) designs was one of my stocks-in-trade so to speak, though they were usually called tower designs until MK's spreadsheets became popular, and quickly found that increasing the Re to 8ohms and using a straight pipe tuned to ~34Hz appeared to be about the best trade-off between BW/efficiency. It's a T/S max flat alignment (Vb = 10.36ft^3), so if too boomy in-room can be tuned lower to better blend with the room and/or lower system Q somewhat, though I don't recommend tuning below ~25Hz without heavy stuffing to damp the Fb peak 'ringing'.
I don't have a website to post stuff so to see the response, input the above driver values using those in parentheses, then pipe info:
L (in) 47.32
S0/SL mult. 2.7373 (this is the multiplier if Sd is the reference)
X0 (%) 0.5146 (this is how far down from the top the driver should be (L*X0), or a little below the midpoint in this case)
S0/SL (in^2) 374.63 (this is the cross sectional area of the pipe)
S0/SL (width) 24.62
S0/SL (depth) 15.21 (these two are based on the golden ratio, but somewhat flexible in a midbass/sub app)
X0 (in) 24.36
Vent radius (in) 4
Lv (in) = 4
stuffing density 0.25lb/ft^3
-----
Answering your website Qs:
>I presume that the port must be sufficient in size to prevent chuffing from high air speeds.
====
Correct, these are no different than a standard reflex in this respect.
====
>What is the Tline doing then/? Just simply resonating like an organ,not impedence matching,because it is not 103db efficiency like the basshorn…
====
Correct.
====
>I wasn’t sure which M number to do,so I copied the labhorn and did M=0.5
====
M, or flare rate, is determined by the driver's specs, desired BW and efficiency. To effectively use any horn response program to sim high efficiency horns, you'll need to find a Marshal Leach horn analysis spreadsheet or on-line calculator (not sure if one exists yet). Or get his JAES papers and do your own excel spreadsheet.
====
>Notice how the horn loads well at Fc,but unloads like a ported box in the 55hz region but with ripples,and below Fc excursion dramatically goes up,this is the part I am not fond of,as the input power is only 100watt rms,and my Xmax=7.5mm.
====
Horn's are 1/4WL resonators and this one's too short, with too small a mouth, hence the 'sinewave' excursion plot.
====
>But also,why is it loaded at 40hz,but the response continues flat to almost 30hz,where excursion is high.
====
Because the excursion is high. In-room, if a corner or something doesn't load it down it will be excursion/distortion limited, with a very ragged response.
HTH,
GM
Fs = 40Hz
Vas = 175.58L
Qms = 5.0 (8ohms)
Qes = 0.33 (0.528)
Qts = 0.30956 (no need to input, the spreadsheet calculates it)
Le = 1.75mH
Pe = 600W
Re = 5ohms
Sd = 883cm^2
Xmax = 7.6mm
B*L = 19.441N/A
Cms = 1.5951e-04mm/N
Mms = 0.0993Kg
Plugging these into MLTQWT and Hornresp should give you a bit more accurate sim. Not as good as measuring yours, but should be in the ballpark.
Mass loaded pipe (MLTQWT, or ML-TL) designs was one of my stocks-in-trade so to speak, though they were usually called tower designs until MK's spreadsheets became popular, and quickly found that increasing the Re to 8ohms and using a straight pipe tuned to ~34Hz appeared to be about the best trade-off between BW/efficiency. It's a T/S max flat alignment (Vb = 10.36ft^3), so if too boomy in-room can be tuned lower to better blend with the room and/or lower system Q somewhat, though I don't recommend tuning below ~25Hz without heavy stuffing to damp the Fb peak 'ringing'.
I don't have a website to post stuff so to see the response, input the above driver values using those in parentheses, then pipe info:
L (in) 47.32
S0/SL mult. 2.7373 (this is the multiplier if Sd is the reference)
X0 (%) 0.5146 (this is how far down from the top the driver should be (L*X0), or a little below the midpoint in this case)
S0/SL (in^2) 374.63 (this is the cross sectional area of the pipe)
S0/SL (width) 24.62
S0/SL (depth) 15.21 (these two are based on the golden ratio, but somewhat flexible in a midbass/sub app)
X0 (in) 24.36
Vent radius (in) 4
Lv (in) = 4
stuffing density 0.25lb/ft^3
-----
Answering your website Qs:
>I presume that the port must be sufficient in size to prevent chuffing from high air speeds.
====
Correct, these are no different than a standard reflex in this respect.
====
>What is the Tline doing then/? Just simply resonating like an organ,not impedence matching,because it is not 103db efficiency like the basshorn…
====
Correct.
====
>I wasn’t sure which M number to do,so I copied the labhorn and did M=0.5
====
M, or flare rate, is determined by the driver's specs, desired BW and efficiency. To effectively use any horn response program to sim high efficiency horns, you'll need to find a Marshal Leach horn analysis spreadsheet or on-line calculator (not sure if one exists yet). Or get his JAES papers and do your own excel spreadsheet.
====
>Notice how the horn loads well at Fc,but unloads like a ported box in the 55hz region but with ripples,and below Fc excursion dramatically goes up,this is the part I am not fond of,as the input power is only 100watt rms,and my Xmax=7.5mm.
====
Horn's are 1/4WL resonators and this one's too short, with too small a mouth, hence the 'sinewave' excursion plot.
====
>But also,why is it loaded at 40hz,but the response continues flat to almost 30hz,where excursion is high.
====
Because the excursion is high. In-room, if a corner or something doesn't load it down it will be excursion/distortion limited, with a very ragged response.
HTH,
GM
why (0.528)Qes???!
oh i did have it! it was just incredibly boring and i didnt know all the parameters..but i will try it,see if it can design me a good horn
oh i did have it! it was just incredibly boring and i didnt know all the parameters..but i will try it,see if it can design me a good horn
i found out why my excursion was so hig,on the pics on my website,cos i used 1000watts ,not 100watts
il check up on the cubic meters-cc etc stuff when i get back to this pc sometime in the next week
🙂
il check up on the cubic meters-cc etc stuff when i get back to this pc sometime in the next week
🙂
>why (0.528)Qes???!
====
To get a ~flat FR to a decently low Fb.
====
>theres the response with the ml util specs
====
That looks more like what I expected. That ML Util program, where did you get that?
TIA,
GM
====
To get a ~flat FR to a decently low Fb.
====
>theres the response with the ml util specs
====
That looks more like what I expected. That ML Util program, where did you get that?
TIA,
GM
http://melhuish.org > sound, software, ML UTIL
😱)
ive had it for ages but didnt realise its use lol
it doesnt give me a nice response ,, why is it so low in response? it reccomends a tiny throat! :-S
i realise you tell me the Qes number,is that to compare to the normal JBL Qes number/? to see the nice roll off
im on the 1.6ghz i can muk about with these things for abit :-D im using Qcad to draw up my box (175L Fb=33hz) so i can hurry up and get it cut and built.
🙂
😱)
ive had it for ages but didnt realise its use lol
it doesnt give me a nice response ,, why is it so low in response? it reccomends a tiny throat! :-S
i realise you tell me the Qes number,is that to compare to the normal JBL Qes number/? to see the nice roll off
im on the 1.6ghz i can muk about with these things for abit :-D im using Qcad to draw up my box (175L Fb=33hz) so i can hurry up and get it cut and built.
🙂
Ah, Michael Zhang's program. He lives near me, but I've never downloaded (or seen) his program since it requires another MS app to be able to run it.
====
it doesnt give me a nice response ,, why is it so low in response? it reccomends a tiny throat!
====
No, and it won't, for the reasons I gave. Small throat means a very long horn/large mouth to get best performance from it. Load it with a autoformer to raise its Qes and it will calc a larger throat/higher 'M', ergo shorter horn/smaller mouth required.
====
>i realise you tell me the Qes number,is that to compare to the normal JBL Qes number/? to see the nice roll off
====
Right, do two different sims, one with the specs and another with the revised to see the difference in performance, especially WRT transient response.
GM
====
it doesnt give me a nice response ,, why is it so low in response? it reccomends a tiny throat!
====
No, and it won't, for the reasons I gave. Small throat means a very long horn/large mouth to get best performance from it. Load it with a autoformer to raise its Qes and it will calc a larger throat/higher 'M', ergo shorter horn/smaller mouth required.
====
>i realise you tell me the Qes number,is that to compare to the normal JBL Qes number/? to see the nice roll off
====
Right, do two different sims, one with the specs and another with the revised to see the difference in performance, especially WRT transient response.
GM
yep i think it was a 20mb download
so your saying compare the 2 different qs in the ml util program?
originally i thought u meant in mathcad,which will do anyway.
🙂
so your saying compare the 2 different qs in the ml util program?
originally i thought u meant in mathcad,which will do anyway.
🙂
- Status
- Not open for further replies.