"tubes" instead of chambers?

[audiotang]

a member by the name of jens posted this question in a long thread regarding port designs and i don't think it ever was answered, but I'd like to know what eveyone thinks, as i have the same question, seems cabinet construction would be greatly symplified if hardware store piping and tubes could be used in lieu of standard construction to create folded horns... and also conceptially it seems that a rounded pipe or hose of the desired length would be "faster" since there would be no right angles and diffraction issues as occur in building a horn loaded cabinet. Here is his post:
____________________________________-
To me as a beginner in the DIY speaker area, it looks like a TL, which leads me to a question re. the the picture in the first posting. (sorry if it seems slightly off-topic).

It isn't the first time I've seen some debate about these edges, but why couldn't one make an "inner lining" with the kind of flexible metal tube used for kitchen ventilations etc., and fill the remaining volume with expanding foam?

This way there would be no edges and reflections. Kinda like the B&W Nautilus (?) snail shape....

Jens


__________________
The Hi-Fi set is ok when you don't notice it's there...

[BillFitzmaurice]

Folded horns have a taper, and piping/tubes do not, so that pretty much tells that tale. TLs can be made using a constant cross-section but they work much better when tapered.

www.billfitzmaurice.com

[kneadle]

Originally posted by audiotang
a and also conceptially it seems that a rounded pipe or hose of the desired length would be "faster" since there would be no right angles and diffraction issues as occur in building a horn loaded cabinet. Here is his post:

This way there would be no edges and reflections. Kinda like the B&W Nautilus (?) snail shape....

I built one on this principle. Here it is: Spiral Shaped Transmission Line Enclosure Project Complete

Bill, do you have any data to support your assertion that TLs work better when they're tapered?

Dave

[AGGEMAM]

If you use a series of tubes with a decreasing crosssection to make it so as to make it tabered I see no problem. You can achieve that by either buying a lot of different tubes or by buying some different tubes and then cutting them up lengthwise and glueing them back together now with a different crosssection.

It's probably going to be a lot of work but I think the end result will be almost perfect, if you remember to make the start of the "horn" take into account of the magnet behind the speaker.

[BillFitzmaurice]

Bill, do you have any data to support your assertion that TLs work better when they're tapered?

There is qute an extensive amount of data available on this, the best source being the 2003 series of articles on TL design in audioXpress by George Augsperger. Previous art by Bailey and Voght also note this.

Even without seeing data this is one area where intuition is correct, as a good TL has no resonant nodes other than the one at the line frequency, and a tapered line by dint of being tapered naturally creates many small nodes over the frequency band rather than a few large nodes as occurs in a non-tapered line, making them far easier to suppress via stuffing.

[MJK]

Bill,

Even without seeing data this is one area where intuition is correct, as a good TL has no resonant nodes other than the one at the line frequency, and a tapered line by dint of being tapered naturally creates many small nodes over the frequency band rather than a few large nodes as occurs in a non-tapered line, making them far easier to suppress via stuffing.

I am sorry, but you completely lost me with this paragraph. What do you mean by a node?

A tapered, straight, or expanding TL will all have quarter wavelength standing waves. Depending on the TL design, the frequency of the standing waves will be slightly different for each type of taper requiring different length pipes for the same tuning frequency. You can see this clearly in the attachments to my alignment tables where many sample simulations are provided.

[BillFitzmaurice]

You have to consider more than one source of resonant nodes. Yes, the length of the pipe provides the major source, but the distance across the pipe is another one that is too often overlooked. It is less problematic in round cross-sections versus quadrilateral or polygonal to be sure, but it is still something to be considered. A constant distance across the line will encourage highly peaked nodes both additive and subtractive at a few frequencies; a varying distance across the line will encourage a larger number of nodes of smaller amplitude. This phenomena is well documented in horns, and it is present in TLs as well. It is more of a crtical issue in horns where internal damping to suppress the nodes is not an option.

www.billfitzmaurice.com

[kneadle]

My spiral design actually takes that into account. If you think about a spiral, which is a multi-plane curve, there is never a uniform cross-section that extends beyond a point at any given time.

Actually, in my (practical) case, the mounting piece provides a uniform cross-section for two inches, and the "port" provides a uniform cross-section for 2.5 inches. While my line is technically constant, without a taper, I wonder what kind of cancellation/attenuation can be supported in an enclosure of my design.

Dave

[MJK]

Bill,

You have to consider more than one source of resonant nodes.

I think the word you want to use is modes. In mechanical engineering, a vibrating structure has modes. Each mode has a frequency and a mode shape associated with it. For example a cantilever beam will have 1/4, 3/4, 5/4, ...... mode shapes at discrete frequencies just like a TL.

You are correct that standing waves across an enclosure will occur. This is true of closed, ported, TL, and horn enclosures. Since closed and TL designs tend to use some kind of stuffing they will probably be less likely to exhibit significant response anomolies due to standing waves between the sides or the front and back of the cabinet.

The cross-section shape of the enclosure will determine the frequency and mode shape, no shape is immune to this problem. If the walls are parallel or skewed the standing waves will still occur at some frequency. Even a round sphere will have internal standing waves.

[BillFitzmaurice]

I think the word you want to use is modes.

Perhaps, though nodes has come into the common usage to describe the effect, mosly when discussing room resonances, but the effect is the same, only the scale differs. In any event you can see the effect on both SPL and impedance charts of any undamped enclosure. True, no shape is immune to the problem, though rounded shapes and non-parallel walls help. The effect is most pronounced in horns where damping is never used and one has to rely on geometry to keep the reponse anomalies under control.