Opening a can of worms but...how do you calculate the slope inside a TL enclosure

Sorry about this thread but...calling all TL experts!!!

Trying to find some good material to which I can understand Transmition Line enclosures. Got lots of very helpfull info on planet10's site...but still a little confused in how to calculate the "slope" inside a TL. What I meen is the "Line" or "Channel" that is made to produce the inside of the encloser.

I know that this is a large question but very willing to learn. If any of you can give me some laymence (need spell checker) termed formulas, or some good sites that explain...that would be most helpfull.

Thanks,

Rino Odorico
 
Basically, just wondering were to put the baffels to get the proper taper etc.

No driver in mind...yet. What it looks like, I will "play" with different drivers over time. I understand the length of baffle is dependant upon the Fs of the driver (1/4 wave).

I also understand, keep away from any 90 degree angles etc. Just, were do you start in placement of the "Line"...or does it matter as long as it works out to the proper volume displacement?
 
Hi Rino

Yup. a big can of worms...

QWTLs are great speakers, if you get them right, but most of what you get on the web is hearsay, random experimentation or just plain wrong.

The few people who have tried to construct real models of TL design, esp. Augsperger and Martin J. King disagree on the effect of line taper on box response.

I have had good results with both tapered and non- tapered lines, but this was before I started measuring my boxes properly, previously I was just using my ears.

No driver in mind...yet. What it looks like, I will "play" with different drivers over time. I understand the length of baffle is dependant upon the Fs of the driver (1/4 wave).

You can't expect any driver to work in a generic TL enclosure, just like any other style of box it has to be designed to work effectively with the driver.

Rather than trying different drivers in the same box, I would get a good quality driver, and build several different enclosures to find one that sounds good to you, or you could download Martin's software and try it all out in a virtual world ( not an option open to me as I run a Mac, and no one in England seems to have second hand copies of Mathcad!).

So, pick a driver first, then we can help :) :)
 
Hi Bill

According to a study carried out by Tyrland ( I think that's the correct spelling!) , tapering the line means that you can get away with a slightly shorter line and decreases the resonances at unwanted higher frequencies, but these are minor in effect compared to just folding the line rather than just having a linear pipe.
 
pinkmouse said:
The few people who have tried to construct real models of TL design, esp. Augsperger and Martin J. King disagree on the effect of line taper on box response.

At 1st Martin didn't think taper had any effect. But as he refined the model based on actual results, and compared results with lines that Rick Schultz (exolinear) was doing using Augspurger's software, he has come to agree with Augspurger (not surprising since Augspurgers electrical analog and Martin's mechanical analog generate essentially the same lines).

A strong taper 3 or 4:1 improves the low pass function of the line allowing the use of less stuffing for the same amount of ripple thereby improving efficiency. This is at least partly due to the shifting of the harmonics of the fundemental up in frequency where the stuffing is more effective. The primary fundemental stays in the same place -- if you are designing for maximum bass reinforcement the line should remain at 1/4 Fs.

Bill: no matter how long you make the line, if it is closed at one end and open at the other end it is a quarter-wave resonator.

Back to the original question: getting a smooth taper is simply a matter of geometry. no folds is trivial. With one fold, you set the cross-section of the box to the area of the closed end + the area of the open end. Start your divider such that it divides the total area into these sub-areas. Then at the other end the partition is exactly in the middle and away from the end by the same distance it is from the walls. More folds is a little tricker, but you use the same technique. Any more explanation will require some pictures.

It should also be noted that the taper does not have to be linear. One fellow has gotten good results with an exponential taper and one has to admire the shape of exolinear's Augspurger modeled RS 40-1354 EXL.

dave
 
Pictures please :>)

Sorry Dave,

But could you send some pictures, or drawings?

I'm trying to picture this in my head so I can understand it fully. It's just like pinkmouse had said, there are a lot of myths revolving around the TL design. And unfortunatly, I fell right into some of these myths...

Once I get on a new project I like to learn everything about it...so sorry about the anal retentavness (need spell checker)

Thanks to all that responded

Rino Odorico
 
Rino,
I have been going crazy trying to figure out a cheap and easy way to build TLs. No matter what kind of design you end up going with there seems to be at least one common thing that starts the whole confusing aspect of this.
It is easy to figure out the length of the pipe needed for QW. But since hollow pipes tend to sound like hollow pipes you stuff them to stop the resonance.
When you stuff them you are effectivly adding length to the pipe by slowing down the sound from the driver. So you can not use a pipe whose length is based on a QW even though it is a QW you want to produce.
Isn't this fun?
 
Thatch_Ear said:
It is easy to figure out the length of the pipe needed for QW. But since hollow pipes tend to sound like hollow pipes you stuff them to stop the resonance.
When you stuff them you are effectivly adding length to the pipe by slowing down the sound from the driver. So you can not use a pipe whose length is based on a QW even though it is a QW you want to produce.

And here we have one of the myths -- at least in some respects. To gain the advantage of extra extension & efficiency in the bass we do not want the stuffing to have any effect at the fundemental. Hence the speed of sound at that frequency is not affected by the stuffing and the line still needs to be a quarter wavelength of the Fs (ignoring end correction).

If you are building a TL for its benefits as a way to remove any reflections from coming back thru the cone (as i like to say "sucking the output from the back of the speaker down the line") then the length becomes less crtitical and you stuff for flattest impedance curve. The output from the terminus is much lower and you do not achieve maximum bass extension or efficiency (but you get less ripple and lower group delay).

So it is relatively easy. Make the line a 1/4 WL of Fs for max efficiency & extension or shorter for an aperiodic line and tune stuffing by impedance.

dave
 
There are ways to work around standing waves and you can work with the QW of a drivers FS and get that Hz and hopefully it does sound like you want.
Most tend to use an easier to build line and use acoustical stuffing.
If you use stuffing it will change the length that the driver "sees"and it is no longer the QW of the Fs of the driver.
You can take advantage of that aspect and go for a lower Hz or a longer QW if you will.
Most commonly used methods of TL building require use of acoustical stuffing and so most commonly built TLs will function at a different QW than the Fs of the driver even if the lenghth of the line was based on the QW of the FS of the driver used.
Or look at it in this aspect; if the Fs of the driver is the Hz level you want to acheive a transmissin line using typical contruction methods and thus requiring use of acoustical stuffing, the TL can not be the actual length of a QW of the Hz of the driver's Fs.
 
David,

No. What i am saying is that even in a stuffed line, the stuffing should not affect the speed of sound at the fundemental (assume the design is going for efficiency & bass extension). We are aided in this by 2 factors:

1/ as the wavelength increases the wave sees less damping relative to its wavelength (ie a significant amount of the cycle has to be traveling thru the stuffing for it to have an effect)

2/ stuffing is, in general, less effective as the frequenct goes down. Take a look at the graphs at the bottom of this page to get an idea of the magnitude.

The whole idea of shortening the line & stuffing to compensate is an approach that is counter to what most people are trying to achieve with a TL. even John Cockcroft, one of the biggest proponents of the short-line, said as much to me in my last telephone conversation with him.

dave

of course you could use dynamic damping, as in the picture below and all bets are off

[IMGDEAD]http://p10hifi.net/tlinespeakers/TLS/projects/victor/images/SuperDempher-s.jpg[/IMGDEAD]
 
Thatch_Ear said:
I hate to say this but that pup does not look to happy, what kind of music you listening to?

a picture sent to me by a fellow in Siberia.

So just enough packing to stop standing waves will not effect the lower, longer waves like it does higher shorter ones.

perfect would be a step function...

dave