Folding 1/4 Wave Tapered pipes

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi, I'm new to this forum but a long time fan of diy speakers. I failed algebra in highschool so I find alot of the math daunting. I have recently purchased a Jensen TS1212 Sub which cost next to nothing and happened to have the right characteristics for an existing ported enclosure I have been using. It performs reasonably well considering the investment, however, I have been tinkering with the idea to build a 1/4 wave TL enclosure.

My reading so far has lead me to choose a Tapered Pipe. I downloaded the Free version of MathCad and plugged in my driver specs and pipe length and taper that I deduced from the available space I have. I have arrived at a 140" pipe with a closed end cross section of 2X driver area (ds) and an open end CS of .5 ds. This gives a fairly nice SPL curve in the BW I intend to use. (20-80hz). Impedence curves look good to and port output remains in phase with driver output down to <15hz.

My problem is I haven't yet figured out how to plug my folded horn sections into the program so I'm not certain how the 7 folds will effect performance. My space dictates how many folds I put in.

My question is, will excessive folding and small cross sections and taper ratio result in worse performance than a shorter pipe with larger cross section and higher taper ratio.

Any imput is welcome, however, keep the math simple if possible.
:xeye:
 
Astro - have you considered a non-tapered pipe? You can get away with alot less length than you might think, and still get true 1/4 wave action percolating in your enclosure. I built a pair of these subs earlier this year using Martin's worksheets and some Dayton 18" pro drivers and they absolutely bring the house down...literally, if I'm not careful. :cool: It has an 8" dia port firing out the top which loads the ceiling and with room gain, there's nothing music-wise that they won't play with authority - and with all that smoothness and the quick transients that TL's are known for.

An externally hosted image should be here but it was not working when we last tested it.


PS - The cab next to the sub is using the same concept with a Fostex FX200 and an FT17H supertweeter. Usable bass down to 30Hz.
 
Tapered pipe Subwoofer

Thanks for the feedback guys.

Wow AJ, that's a pretty imposing looking SW. Don't think it would pass my WAF parameters though (Wife acceptance factor) :bawling:

I actually have a very specific space limitations. The sub will be placed in a corner behind a TV on a stand and my main speakers, Angstrom Reflexions. It was a major coup to get these full range speakers into our relatively small living room. 10X15X 9. I've already gotten the go ahead for a cabinet measuring 16"WX26"DeepX30" tall (inside measurements). I worked on several configurations for the pipe and arrived at one with 7 folds that allows me to place the driver at a approx a .6 position closest to the closed end, which was the best position according to the MathCad predictions. Any shortening of the pipe regardless of stuffing density, port configurations or pipe shape resulted in lowend roll off, well before fs.

I'm pretty certain this is the way to go, unless there will be serious deleterious effects from folding the pipe so severely. I've read most of Martin King's articles and haven't come accross anything that would suggest folding the pipe would have serious problems with a subwoofer, because most of the harmonics would be well above the Xover point of 80Hz. The most important point he stresses is to avoid any drastic changes in cross section along the length of the pipe, which I will avoid doing.

I'd really like to know if anyone has a good sense of the type of problems I'll likely encounter that can't be mitigated with stuffing.
 
Simple methods for calculating bend geometry in a tapered tube (Zipped PDF)

I figure that smoother bends => smoother pipe => smoother sound.

So, perfectionist that I am, I decided that I wanted really smooth curves in the folded TL sub project I've been meaning to build. And thus, back in August I started work on the attached "Simple Method" and had been hoping to add the 3D methods before offering it up. As you might expect, only the 2D version is attached and my sub project is still un-started. Still, 2D tapers are where most people apply their time so I offer this up now for all the post-holiday projects about to be started...

As always, constructive critique is very welcome, especially if there is an error to point out.

Enjoy!

:)ensen.

PS: Many thanks to MJK and Planet10 for being kind enough to take a peek at my first draft (way back when).
 

Attachments

  • simple methods for calculating bend geometry in a tapered tube.zip
    54.6 KB · Views: 370
Smooth bends

Hi PurplePeople (is PP ok ;) ),

Thank you very much for the file. I'll review my geometry with it to see if I've got it right or not.

From what I've read so far about TLs, your phylosophy about smooth bends is true for full range designs, but there appears to be some controversy about whether curved bends, or corner reflectors make measurable differences in a subwoofer application. Check out this page from Bob's Speaker Stuff:
http://geocities.com/rbrines1/ According to his measurements, there is no need to invest much effort into "smoothing the bends", as it were! :confused:

It seems counter intuitive though doesn't it. You would think that corner reflectors or rounded corners would preserve a more constant taper and minimize standing waves in the corners. However, I suppose that in a SW's bandwidth, the standing waves would be too short to be exited. :cool:

As for feed back on your paper, it appears very well written, and the equations are probably simple to anyone with a grasp of basic algebra, however, I am so easily overwhelmed by math equations that I tend to shut down as soon as I see equations without descriptive text for every permutation. To me it's like a foreign movie without subtitles. :confused: :xeye: ;) If you are targetting the novice as well as affictionatoes, you may want to add a legend defining all of your variables ie., X = Cross section etc... I found myself scrolling back and fourth to your diagram to follow the equations. :(

Hope you found this constructive PP :angel:

Thanks again :D
 

GM

Member
Joined 2003
>My question is, will excessive folding and small cross sections and taper ratio result in worse performance than a shorter pipe with larger cross section and higher taper ratio.

====

Well, with no specs I can't sim it, but FWIW, a steeper taper will allow a much shorter pipe, though net Vb won't change. Also, as the reverse taper increases, the driver moves toward the closed end.

Anyway, folding a pipe intended for sub duty relieves you of worrying about damping any pipe BW beyond the XO BW, so if using an 80Hz/4th order XO, then -24dB = 160Hz, though some folks consider -12dB as adequate, or 160*0.707 = ~113Hz. To keep standing wave reflections from coming back through the driver or setting up standing waves between bends, the distance from the driver to the first bend and between each succeeding bend needs to be < 1/3WL of 160Hz, or ~13560"/pi/160 = ~27" (113Hz = ~38") until the pathlength is 1WL long. After this point is reached, the bends can be much farther apart since 1WL of 160Hz = ~pi*WL of ~51Hz (~36Hz), where the process repeats itself.

Where most folks screw up is in designing the bends themselves. If the taper doesn't continue around the bend, or at least remain constant, out of BW standing waves can reflect back through the driver as if it is in a closed pipe, i.e. at 27" there would be standing waves beginning at ~6780"/27" = ~251Hz (~178Hz), then ~502Hz (357Hz), etc..

====

>According to the theory on the above site folding is not a good thing

====

I haven't read every word he's published, but if true, then either there's some misunderstanding or flat out error. For sure it must be done properly for the desired BW, but folding per se is neither good nor bad, just another 'tool' in the speaker designer's 'cookbook'.

====


>I figure that smoother bends => smoother pipe => smoother sound.

====

Right, where there's a need for the pipe to cover a wide BW, such as a plane wave tube or horn, but for TLs, ML-TLs, you want to attenuate the pipe's out of BW upper harmonics, so reflectors aren't a good plan performance wise. Indeed, if they're not sized/positioned to continue the bend correctly for the desired HF BW then they can do more harm than good, and if not needed then more stuffing is required to attenuate the extra BW they pass. Instead, put damping material in adjacent corners to keep mid/HF standing waves from developing.

GM
 
Design details

Hi guys,

Sorry Timn8ter, the driver is a Jensen KS212 (I had posted from work and memory usually doesn't serve me well ;) ). The FS is 26.13 Hz Vas = 4.231 Ft3, Qts = .379, Qms = 7.37, Qes = .399, Sd = 74.41"2, Lvc@1000hz = 1.009 mH, Re = 3.55 Ohm

Gm, the XO I'm using is, I believe 24db/octave, but I can't know for certain because I purchased it through a local surplus store without specs. Please forgive my ignorance GM, but I'm not certain I follow you on the standing wave problems. Are you suggesting that to suppress standing waves of 113Hz, the mimimum pipe length between driver and first bend has to be 38" ? How then to build any folded configuration with a cabinet size of less than 38"? :confused: Below is a diagram I just did in paint. It's by no means to scale. I tried to scan my sketch in but my scanner has a love/hate relationship with my computer and today they aren't talking! ;) :mad: Let me know if I'm way out in left field or not.
 

Attachments

  • enclosure.jpg
    enclosure.jpg
    14.2 KB · Views: 864
Hi Sreten,

I've read the article and agree with the rationale. As noted in previous posts above, this may be more critical in full range TL design than in SW designs. Perhaps I'm wrong, but I haven't come accross any literature that indicates otherwise. A single fold SW would necessitate minimum cabinet lengths of approximatel 6 ft. Not an option for most real world applications! ;)
 
Astro, thanks for the feedback. I will make sure the 3D version has the full list of variables. My intent was to keep it small and make the final calculation easy to do. If a builder has decided on a driver and dimensions for a straight line, then filling in the variables is easy at each bend. The overall volume of the line will be greater, but the equation ensures that the TL length will always be the same.

GM: My original thinking on this comes from playing the sax in my teens. Even then, I knew intuitively that any change in the shape of the waveguide creates a tonal change. By giving the waveform a smooth transition around corners, we lessen the reflections that create the acoustic mess. I think a lot of muddy LF comes from enclosures that bounce around too much energy, creating diffractive effects within.

I agree that using stuffing to limit upper harmonics is good for subs and the theory that the low frequencies have a wavelength too long to make a difference is probably correct for fairly simple waveforms. But even bass notes have tonal qualities and accurate presentation of that tone becomes important. Even a basic 30Hz wave should have harmonics at 60 and 120Hz and these should be presented by the sub (up to the XO frequency). More complex waveforms will have even more that really should not be attenuated more than needed. The stuffing should never attenuate harmonics, just clean up the reflections.

:)ensen
 
>referring to to the observations on page 9 regarding the straight
and one fold TQWT.

====

The way I interpret it is he's proving my point WRT the detrimental effect on performance a simple diagonal divider board has in a rectangular cab due to not maintaining the flare rate through the bend(s). ;)

====

>My original thinking on this comes from playing the sax in my teens. Even then, I knew intuitively that any change in the shape of the waveguide creates a tonal change. By giving the waveform a smooth transition around corners, we lessen the reflections that create the acoustic mess. I think a lot of muddy LF comes from enclosures that bounce around too much energy, creating diffractive effects within.

====

Depends on the BW Vs pipe cross sectional area and length.

====

>I agree that using stuffing to limit upper harmonics is good for subs and the theory that the low frequencies have a wavelength too long to make a difference is probably correct for fairly simple waveforms. But even bass notes have tonal qualities and accurate presentation of that tone becomes important. Even a basic 30Hz wave should have harmonics at 60 and 120Hz and these should be presented by the sub (up to the XO frequency). More complex waveforms will have even more that really should not be attenuated more than needed. The stuffing should never attenuate harmonics, just clean up the reflections.

====

Not sure what you mean here. Excepting pipe organs, low frequency music is composed mostly of suppressed fundamentals, so it's their upper harmonic structure that defines their tonal qualities. Since we don't want the increasingly out of phase output of the vent to comb filter with, or modulate, the driver's output or comb filter with the mains, doesn't it seem reasonable to you that damping the out of passband pipe harmonics is required to maintain the proper tonal balance within the passband?

GM
 
But I would want a sub to reproduce big pipes. As stated, I haven't yet built my sub, so I have no empirical evidence yet. However, my thinking has always centred around the TL opening allowing the exit of phase corrected backwaves, such that a forward facing "port" would be need to be 180deg from the backwave of the driver. So whatever FR is coming from the front of the driver needs to be duplicated at the port. In an 2-way design, the LF driver may be handling BW up to 3kHz. That means the port should also. Ideally, the stuff should not attenuate the harmonics of a wave since the driver also puts those harmonics out the front. But the interior of the enclosure should prevent internal reflections as these will interact and create acoustic mush which will eventually show up at the port.

What I haven't been able to surmise yet is whether putting the driver right at or near a fold will cause problems with the air being "pumped" in two directions. Like Astro-Muse, every cabinet I've designed so far places the driver at a bend and it worries me. I've tried to move it but doing so forces the design away from MJK's alignment table parameters.

:)ensen.
 
So whatever FR is coming from the front of the driver needs to be duplicated at the port.

I'm sorry but you appear to have a fundamental misunderstanding
of transmission lines and the way the 1/4 wavelength works, and
the consequences of the 1/3 wavelength, 1/2 wavelength and 1/1
wavelength which are alternatively in and out of phase with the driver.

Replication of the front output of the driver is simply not possible.

:) sreten.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.