which taper ratio for TL?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
basically, i have nearly finished my design of latest speaker. It is going to be a tapered TL using a 5 inch driver, Fs 40Hz, Qts 0.43, and Vas 11 litres.

I have chosen to taper the TLine in order to shorten it and negate the need for any bends.

I have 2 options:

1: Sl= 4.5Sd tapering down 20:1 ratio, So=0.25Sd, line length about 900mm

2: Sl= 4.5Sd tapering down 10:1 ratio, So=0.4Sd, line length about 1050mm

Presently I am considering design #1, as i figure that if i need to tune the pipe slightly higher for any reason, then i can increase the vent area a little. Indeed i may be able to increase it up to the same So as design #2 if necessary. As i cant see much in the way of size difference, it occured to me..."Why dont i just build the 10:1 taper instead".

Now im confused ..hehe:warped:

my question is this:

which design would be preferable and why?
 
Your 10:1 tapered line will have a 1/4-wavelength resonant frequency around 50 Hz, but your driver's combination of Fs and Qts will likely fair better if the line's resonant frequency is at or a little below the driver's Fs of 40 Hz. Now the 20:1 tapered line will have a bit lower resonant frequency, but it still might not be low enough for the flattest, smoothest overall response, and you have to be careful about potential terminus air velocity at larger and larger tapers with smallish beginning line areas (you may be okay starting at 4.5 Sd, however). Of course a benefit of larger taper ratios is that they will smooth out the overall response.
Paul

basically, i have nearly finished my design of latest speaker. It is going to be a tapered TL using a 5 inch driver, Fs 40Hz, Qts 0.43, and Vas 11 litres.

I have chosen to taper the TLine in order to shorten it and negate the need for any bends.

I have 2 options:

1: Sl= 4.5Sd tapering down 20:1 ratio, So=0.25Sd, line length about 900mm

2: Sl= 4.5Sd tapering down 10:1 ratio, So=0.4Sd, line length about 1050mm

Presently I am considering design #1, as i figure that if i need to tune the pipe slightly higher for any reason, then i can increase the vent area a little. Indeed i may be able to increase it up to the same So as design #2 if necessary. As i cant see much in the way of size difference, it occured to me..."Why dont i just build the 10:1 taper instead".

Now im confused ..hehe:warped:

my question is this:

which design would be preferable and why?
 
Last edited:
Your 10:1 tapered line will have a 1/4-wavelength resonant frequency around 50 Hz, but your driver's combination of Fs and Qts will likely fair better if the line's resonant frequency is at or a little below the driver's Fs of 40 Hz. Now the 20:1 tapered line will have a bit lower resonant frequency, but it still might not be low enough for the flattest, smoothest overall response, and you have to be careful about potential terminus air velocity at larger and larger tapers with smallish beginning line areas (you may be okay starting at 4.5 Sd, however). Of course a benefit of larger taper ratios is that they will smooth out the overall response.
Paul

yes a smoothening of passband ripples is the intention, hence the initial idea of a 20:1 ratio taper...both lengths quoted here are approx. as i dont have the drawings to hand, but the intended Fb is about 45-50Hz, as i figure stuffing will lower that a small amount, plus with a Qts of .43, i would like to aim for a finished line QWfrequency of 45Hz.

These figures are approx and especially as i have just measured Fs on both drivers and the average is actually about 46Hz with 40hrs of run in. However, in a ported system the port area works out to be very small, in fact a 1.5" tube, and i would be happier with a minimum of 2" so an area of .25Sd or about 3 in³ would meet that criteria. I was more worried about over attenuating the vent output. though that being said, wouldnt a smaller vent are also decrease the acoustic LP frequency of the vent; thus keeping more mids from being radiated from the vent?
 
Last edited:
I'm not sure I follow your logic entirely but in a vented system, as well as in an ML-TL that uses a port, making the port's diameter smaller will lower the tuning frequency which will, indeed, reduce the frequency of signals that are allowed to exit the port. If you want, I'd be happy to model the expected performance of your TL, just give me the proposed line dimensions and all of these driver's T/S values: Fs, Re, Le, Bl, Vas, Sd, Qes and Qms.
Paul

yes a smoothening of passband ripples is the intention, hence the initial idea of a 20:1 ratio taper...both lengths quoted here are approx. as i dont have the drawings to hand, but the intended Fb is about 45-50Hz, as i figure stuffing will lower that a small amount, plus with a Qts of .43, i would like to aim for a finished line QWfrequency of 45Hz.

These figures are approx and especially as i have just measured Fs on both drivers and the average is actually about 46Hz with 40hrs of run in. However, in a ported system the port area works out to be very small, in fact a 1.5" tube, and i would be happier with a minimum of 2" so an area of .25Sd or about 3 in³ would meet that criteria. I was more worried about over attenuating the vent output. though that being said, wouldnt a smaller vent are also decrease the acoustic LP frequency of the vent; thus keeping more mids from being radiated from the vent?
 
Here's my modeling results...

Mondo, I got the automated email from this forum containing the T/S values for your driver but for some reason, your post didn't end up being posted in this thread. Anyway, I did some modeling. I don't know exactly what kind of driver you're using but for the sake of modeling I assumed it was a ~5-inch fullrange type. So, I ended up with a line tapered at 15:1 and 42" long, starting at 4.13 Sd and ending at 0.276 Sd. Dimensionally the starting area was 7.5" square and the ending area was 7.5"W x 0.5"D. The terminus would be 7.5"W x 0.5"H. I placed the driver's center 8" from the beginning of the line. I used a stuffing density of 0.75 lb/cu.ft., stuffing with a total of ~7.5 ounces in the first 28" of the line. I've attached the system response graph (red line). I used Martin's ML-TQWT worksheet which allows modeling a tapered line, too. As you will see on the response graph, F3 is ~39 Hz and the ripples in the response are acceptably low. If the stuffing density and/or length were to be increased, the ripples would be diminished but at the expense of a higher F3. The terminus air velocity with this 2.83-volt modeling input was under 1% of the speed of sound.
Paul

Edit: The peak of the terminus air velocity occurs at ~32 Hz. Without stuffing, the system's tuning frequency is 40 Hz, and with stuffing it's ~38 Hz.
 

Attachments

  • MondoTTL1.gif
    MondoTTL1.gif
    6.6 KB · Views: 269
Last edited:
aha you noticed my deletion.......!

many thanks...basically im working nights and at work. The figures i gave were approximate and after checking them in the excel sheet i used previously i couldnt get them to tie up quite the same. So i deleted the post, with the intention of posting the figures i have on my home spreadsheet calculations i did from my measurements. I think for the most part the numbers are accurate enough though, but BL maybe 0.1 or so higher or lower, whatever difference that would make!!

FR graph does indeed look really quite acceptable, i was expecting somewhat higher F3 given that the driver is fairly low Qts. And this is tuned to 40Hz unstuffed...given that FR graph, coupled with room modes etc, maybe tuning slightly lower WOULD give subjectively better results...

Thanks very much again..hopefully i remembered all my numbers ok!!! If theres any glaring errors then ill be sure to repost them, correctly. As a guideline design i think maybe the 10:1 or your comprimise between my 2 ideas, the 15:1 taper may both be pretty good starting places.

P.S. the driver is Visatons AL130 aluminium cone mid bass driver. Never played with rigid cones before, but it impresses and has a nice smooth midband. Obvious main breakup mode at about 7 kHz but the crossover im using rolls that out of audibility way before you get that high. Cone resonance seems very well damped, considering the metal cone. Would recommend highly to anyone.compared to my past project; audax HDA 130mm 2 way in SBB4, their resolution and tonal balance is awesome. Strangley that what i thought when i built the audax's, going from paper to HDA cones.
 
Last edited:
I can always model again with different T/S values just to be sure, and I can see whether tuning the line lower will give you something you'd prefer. Also, I didn't model with any d.c. resistance assumed to be in series with it, like that from a series crossover inductor, and we really ought to incorporate that, too. BTW the Qts isn't particularly low and if we add in this series resistance, it will be higher yet. What tweeter are you planning on using with this Visaton driver? Based on my modeling the tweeter would be located above the Visaton.
Paul

many thanks...basically im working nights and at work. The figures i gave were approximate and after checking them in the excel sheet i used previously i couldnt get them to tie up quite the same. So i deleted the post, with the intention of posting the figures i have on my home spreadsheet calculations i did from my measurements. I think for the most part the numbers are accurate enough though, but BL maybe 0.1 or so higher or lower, whatever difference that would make!!

FR graph does indeed look really quite acceptable, i was expecting somewhat higher F3 given that the driver is fairly low Qts. And this is tuned to 40Hz unstuffed...given that FR graph, coupled with room modes etc, maybe tuning slightly lower WOULD give subjectively better results...

Thanks very much again..hopefully i remembered all my numbers ok!!! If theres any glaring errors then ill be sure to repost them, correctly. As a guideline design i think maybe the 10:1 or your comprimise between my 2 ideas, the 15:1 taper may both be pretty good starting places.

P.S. the driver is Visatons AL130 aluminium cone mid bass driver. Never played with rigid cones before, but it impresses and has a nice smooth midband. Obvious main breakup mode at about 7 kHz but the crossover im using rolls that out of audibility way before you get that high. Cone resonance seems very well damped, considering the metal cone. Would recommend highly to anyone.compared to my past project; audax HDA 130mm 2 way in SBB4, their resolution and tonal balance is awesome. Strangley that what i thought when i built the audax's, going from paper to HDA cones.
 
Thank you for the offer. Maybe you could model the response with a line tunes to 35Hz or 30 Hz instead of 40, to see what difference it would make as im probably just as interested in F10 as F3, so here are my reviewed figures:

Qms 4.75
Qes 0.45
Qts 0.4118
Re 5.6
DCR L1=0.25 Ohm
So assume 0.4 Ohm with wiring

Vas 10.82 litres
Sd 79cm²
Bl 5.78

with your 15:1 ratio, well 0.08 is the nearest i can do on excel sheet:
Sl=49in²
So=3.7in²
length=48in

Oh and im using the visaton g20sc tweeter, due to its very flat response graph on their site, the waterfall looks pretty clean too. The only thinh that lets it down for me at least is the 'average' Fs, lower would be nice, but we cant always have it all!

Again thanks alot.
 
Last edited:
Okay, I'll post my results.
Paul

Thank you for the offer. Maybe you could model the response with a line tunes to 35Hz or 30 Hz instead of 40, to see what difference it would make as im probably just as interested in F10 as F3, so here are my reviewed figures:

Qms 4.75
Qes 0.45
Qts 0.4118
Re 5.6
DCR L1=0.25 Ohm
So assume 0.4 Ohm with wiring

Vas 10.82 litres
Sd 79cm²
Bl 5.78

with your 15:1 ratio, well 0.08 is the nearest i can do on excel sheet:
Sl=49in²
So=3.7in²
length=48in

Oh and im using the visaton g20sc tweeter, due to its very flat response graph on their site, the waterfall looks pretty clean too. The only thinh that lets it down for me at least is the 'average' Fs, lower would be nice, but we cant always have it all!

Again thanks alot.
 
More modeling results

I used your latest T/S values, including adding 0.4 ohms of series resistance, and simply entered them into the very same tapered TL I originally modeled. The resulting system response is shown in the first attached graph (BTW, you didn't give me new values for Fs or Le so I used the original ones of 46 Hz and 0.9 mH). The system tuning with stuffing was ~38 Hz.

Next I lengthened the line to 48", keeping everything else the same, and the system response is shown in the second attached graph. The system tuning frequency with stuffing was ~32 Hz.

Last, I shortened the line to 36", keeping everything else the same, and the system response is shown in the third attached graph. The system tuning frequency with stuffing was ~41 Hz.

I prefer the overall shape of the higher-tuned system but I would make the enclosed volume larger in order to get a lower F3. Tuning the system lower, as shown in the second graph, does make F10 a couple of Hz lower.
 

Attachments

  • MondoTTL2.gif
    MondoTTL2.gif
    6.6 KB · Views: 151
  • MondoTTL3.gif
    MondoTTL3.gif
    6.6 KB · Views: 148
  • MondoTTL4.gif
    MondoTTL4.gif
    6.6 KB · Views: 150
Last edited:
thanks again...wow what a speedy modeller you are ! HAHA

The shortest design does look more 'maximally flat', and is what id tune to acheive in a reflex design. I like the original best if im to be totally honest. Mainly as the length isnt too long, WAF and all that, and i like the slightly less than flat to F3. The low tuned design is also quite appealing, albeit the length is a bit much and may put my 'no bends' idea on the backburner...

I think i may try a combo of the low tuned version, and my original taper of 20:1 in order to reduce line length a little further. Then if i find the bass a little light i can simply increase the vent area to bring the resonance back up to 40Hz or even to the 45Hz mark, where i assume the shortest of your suggestions was tuned.

Many thanks. Just drawing up cutting plan and panel designs out for the CNC man to cut(probably NY now:deer:). Will post some pics as and when i get the build going. Once again, many thanks for your ears, your time, and another set of eyes to look over what im planning--Well i COULD have been trying to make an awful design!!! As it turns out the sims you ran on my behalf indeed reinforce what i had hoped.

Merry Christmas!!
 
I didn't check to see what the tuning frequency was on the shortest line without stuffing but I really don't think it would be any higher than 43 Hz as adding stuffing usually causes only a 1-2 Hz decrease in F3.

I'm glad I could help out and Merry Christmas back at you.
Paul

thanks again...wow what a speedy modeller you are ! HAHA

The shortest design does look more 'maximally flat', and is what id tune to acheive in a reflex design. I like the original best if im to be totally honest. Mainly as the length isnt too long, WAF and all that, and i like the slightly less than flat to F3. The low tuned design is also quite appealing, albeit the length is a bit much and may put my 'no bends' idea on the backburner...

I think i may try a combo of the low tuned version, and my original taper of 20:1 in order to reduce line length a little further. Then if i find the bass a little light i can simply increase the vent area to bring the resonance back up to 40Hz or even to the 45Hz mark, where i assume the shortest of your suggestions was tuned.

Many thanks. Just drawing up cutting plan and panel designs out for the CNC man to cut(probably NY now:deer:). Will post some pics as and when i get the build going. Once again, many thanks for your ears, your time, and another set of eyes to look over what im planning--Well i COULD have been trying to make an awful design!!! As it turns out the sims you ran on my behalf indeed reinforce what i had hoped.

Merry Christmas!!
 
Thanks again for all your help

thanks for the interest and help offered..ive finished designing and have posted my mock up of the finished article, while i await the CNC man, cutting the MDF to my blueprints!!!

then i have just the biulding, and finishing to do, prior to the tuning.

again many thanks and feel free to comment

heres the link:

http://www.diyaudio.com/forums/mult...e-speaker-eventually-started.html#post2037972
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.