Cain & Cain Abby Clone

[EUVL]

Just wonder if any one (Mr. Cain?) has tried a tapered pipe with circular cross section, in which case the area would change quadratically (rather than linearly) with pipe length.

Any advantages / disadvantages ?


Thanks in advance,
Patrick

[planet10]

Originally posted by EUVL
Just wonder if any one (Mr. Cain?) has tried a tapered pipe with circular cross section, in which case the area would change quadratically (rather than linearly) with pipe length.

Why would the taper not be linear?

Andy got the Conehead using translam before he set the project aside (forever he says but you never know)

dave

[planet10]

Rick Shultz built a TL with a non-linear taper with good results...

http://www.t-linespeakers.org/projec...ear/curve.html



Another fellow developed a TL with a reverse exponential taper -- he was going to try to get a patent, but finally just put it into the DIY community.

dave

[cheesehead]

Originally posted by planet10

Why would the taper not be linear?

The taper (increase in diameter with length) would be linear, but the cross-sectional area A=(pi)*r^2 would be quadratic with length. (Seems weird to me too, but the math don't lie.)

Should be able to sim it in MJK's sections worksheet, approximating a quadratic profile by using many different linear sections?

cheesehead

[GM]

Greets!

AFAIK the worksheets (WS) are only using the cross sectional areas (CSA) at each end and the length to define a conical expansion of whatever its 'M' factor (taper ratio) turns out to be, so doesn't matter what CSA shape you input, ergo if you build a very high aspect ratio cab WRT width/depth, the sim won't be as accurate above the cab's compression BW.

Of course if you want to define a specific 'M' factor with 'x' BW, such as exponential (M = 1), then SL and length become the variables and the 'sections' WS will need to be used.

GM

[azira]

Originally posted by cheesehead


The taper (increase in diameter with length) would be linear, but the cross-sectional area A=(pi)*r^2 would be quadratic with length. (Seems weird to me too, but the math don't lie.)

Should be able to sim it in MJK's sections worksheet, approximating a quadratic profile by using many different linear sections?

cheesehead

I think you're mistaken about a rectangular cross-section being different than a circular cross-section.

circular:

A = pi*r^2 -- quadratic.

rectangular:

A = x * y -- linear?
Since the ratio of x:y stays constant because of the linear taper,
x = k*y where k = x/y
so:

A = (k*y)*y = k*y^2 -- quadratic as well.

Since MJKs worksheets just consider CSA w/rt to taper, it doesn't know whether you're using a circular or rectangular shape. However, the taper ratio is different.
--
Danny

[cheesehead]

I think we're saying the same thing, just not reading the same thing...

In most cases, when you're changing the cross-sectional area of a speaker, only one dimension is changing. Say, rectangular baffle and triagonal side. Let h be the distance from the top of the "pipe", with a cross-sectional area (CSA) of zero at the top, then

x = constant (the baffle is the same width throughout)
y = c * h (the width of the side is linear with the distance from the top)

CSA = x * y = c * x * h linear with distance from the top

For a cylindrical pipe where the radius is linear with the distance from the top,

r = c * h

Then,

CSA = pi * (c * h)^2 = pi * c^2 *h^2 quadratic with distance from the top

You could get the same thing by building a square pyramid...

In any case, I think you could use MJK's sections program, and enter a number of sections with linear increase in CSA but increasing "taper" to approximate a quadratic increase in CSA.

[purplepeople]

The transmission line geometry must be a smooth linear transition, along the length, from the closed end area S0 to the open end area SL as shown in Figure 1.

This is from page 18 of MJK's TL Alignment Tables.

I read this to mean that a cross-sectional area at any point X along the line length is predictable by equation. Any deviation from the original rate of change would make this impossible. Never mind that in practice, a stepped transition doesn't make much difference.)

:)ensen.

[FrankDIY]

Hello,

I finally finished my Clabby. See the attached picture. They sound great! I may have a chance to compare them with the real Abby at the Hi-Fi Montreal Festival in april. Usually there is a room with the Abby and I always spend a lot of time in it.

I decided to ignore a bit all the calculation I did with MJK sheets and also take into accound the dimensions of the real Abby. It is a compromise between both design.

Just for fun, I tried to use a BSC and it completely destroyed the mid-range. Basically I was using a very similar version of the BSC used by MJK for his FE164 project.

I am now using them with a pair of ASL wave-8. Honessly I prefer this match over my Pass Zen V4. I will probably build the first version of the Zen (apprently better for full-range speaker?) or maybe the Zen Lite is better?

Anyway, I will have to do some experiementation.

Nothing special in these speakers: FE167E, silver-plated copper wire with teflon insulation, high quality binding post (but not the cardas), 5/8'' MDF, African Mahogany veneer finished with danish oil.

BTW, please start a different subject if this is not to discuss about the Abby, thx.

[Nanook]

They look good Francois.
I too am looking at building something like this. My nephew has the TL/Voight bug. His Dad gave him 2 pairs of DCM TF 250's and he loves em and wanted to "hotrod" them. I suggested that a completely new speaker might be just as easy (He still wants to mod them in addition to building some others).

I have always like TLs and have heard a few. When I stumbled across the Voight pipes on the Internet I was intriqued.

Email me if you would allow a couple in depth questions.

nanook