Yume, a Transmission Line WWMT speaker of moderate cost.

[JasonP]

Well, ok, so I've gotten out of speaker design. I had other things to do, and broke down and just bought a pair of Klipsch Reference. I do however have a home built sub powering my home theater, a ML TQWT 12" design that packs a serious punch.

Things have changed though, and now I'm thinking about a design that will last me a lifetime. Recent tests have made me realize I can't hear below 40 hz anymore - no big shock, I'm in my late thirties. So I'm going to design around that, just 40hz + and probably even subsonic filter the amp too (why waste the power?).

This all aside, I've been playing with Passive Crossover Designer 7 (man, I wish I had that years ago) and looking at different options. I think I finally settled on a design, and I'm kind of looking for general feedback from more experienced people before I embark on mission: Yume.

---------
YUME
---------

The design so far is looking like so: 2x Dayton RS180-8 woofers in the Tritrix Knockdown CNC cabinet with a custom front face, with a SB Acoustics SB15NRXC30-4 5" midbass, and the Seas Prestige 27TDFC for the tweeter. I'm thinking I'll invert the Tritrix box, ok I hacked up a terrible sketch using cut and paste which is attached :P

Anyway, the goals are: low distortion, efficiency (voltage) 90 or better, and 4 ohms nominal - my T-amp put out more clean power into 4 (70W) than 8 (40W). I think these drivers and overall design will fit the bill.

The Tritrix cabinet is a cost choice, since I'll have to do is cut the front piece into a frame and add a fancy pretty hardwood sheet in which to mount the drivers.

I'll post more here as I work on it.

[GranteedEV]

my T-amp put out more clean power into 4 (70W) than 8 (40W).

this is not true. 70w into 4 ohms, is 16.7 V while 40w into 8 ohms, is 17.9V and amps are generally more stable with less current running through them.

[JasonP]

Hrms, don't the distortion figures matter here too though? In this case 0.005% at 70W into 4 ohms or 0.007% at 40W into 8 ohms? Am I just confused here? *boggles*

[Wolf]

You'll still require a chamber on the mid.

My advice is to mount the tweeter near the top, and place the mid in it's own box on top, OR use the triangular 'curve' sector to your advantage and make that your chamber with the tweeter below it.

Later,
Wolf

[JasonP]

Wolf, oh right, I forgot that on the hacked diagram thanks! I was planning to chamber it, I'm just not sure about the size and such yet :)

[Paul K.]

Before you start making sawdust, I hope you plan on modeling the performance of the tapered TL using Martin King's software or something equally good.
Paul


Wolf, oh right, I forgot that on the hacked diagram thanks! I was planning to chamber it, I'm just not sure about the size and such yet :)

[JasonP]

Paul, thats a great idea and should have been obvious.

My initial plan was to:
1) Get the drivers.
2) Build the cabinets.
3) Assemble and measure.
4) Build a mock crossover from those measurements.
5) Test.
6) Repeat 4 and 5 until satisfied.

However, like you say, checking out the existing TL design with these woofers would make sense, duh. Thanks!

[fastbike1]

I wouldn't invert it, if you still want TL performance.


Paul, thats a great idea and should have been obvious.

My initial plan was to:
1) Get the drivers.
2) Build the cabinets.
3) Assemble and measure.
4) Build a mock crossover from those measurements.
5) Test.
6) Repeat 4 and 5 until satisfied.

However, like you say, checking out the existing TL design with these woofers would make sense, duh. Thanks!

[JasonP]

The change of position of the back port will impact it that much relative to the floor?

I inverted it for a couple of reasons: positioning the many drivers on the front panel workably, and coupling the two woofers to the floor, which got rid of a serious reflection problem in all the room simulations.

[Paul K.]

Since Jason is going to place the pair of RS180s towards the bottom of the baffle, he has shown correctly how the angled divider should be located in order for it to properly function as a tapered TL. The line will have a decreasing area, starting with the largest area immediately below the pair of RS180s, and ending with the smallest area out the back.
Paul


I wouldn't invert it, if you still want TL performance.

[Paul K.]

I had a couple of thoughts because I don't know how much modification you plan on needing/doing to the TriTrix cabinet. Will the existing line length and taper provide an optimum 1/4-wavelength resonant frequency for the RS180s, and will the volume in the line be appropriate for the F3 you hope to achieve?
Paul


The change of position of the back port will impact it that much relative to the floor?

I inverted it for a couple of reasons: positioning the many drivers on the front panel workably, and coupling the two woofers to the floor, which got rid of a serious reflection problem in all the room simulations.

[Pete Schumacher ®]

I'm in my 50s, and still have no trouble hearing into the low 20s.

But, if you're having trouble hearing those freqs, I'd suggest tactile transducers to impart the "feel" of the low frequencies at your seating position. You won't need much power at all for that last octave if your mains run from 40Hz up.

[johnnyrichards]

I'm in my 50s, and still have no trouble hearing into the low 20s.

But, if you're having trouble hearing those freqs, I'd suggest tactile transducers to impart the "feel" of the low frequencies at your seating position. You won't need much power at all for that last octave if your mains run from 40Hz up.

I was gonna say, I am in my mid-30's, and can still hear to at least 25 Hz.

That someone can only hear to 40 Hz in their late 30's is a surprise.

Jason, have you worked in manufacturing or something similar for a long time?

[JasonP]

I had a couple of thoughts because I don't know how much modification you plan on needing/doing to the TriTrix cabinet. Will the existing line length and taper provide an optimum 1/4-wavelength resonant frequency for the RS180s, and will the volume in the line be appropriate for the F3 you hope to achieve?
Paul

Yea, I really need to pick up a set of Martin King's Mathcad sheets and explore the ramifications of using the cabinet. Though, oddly enough, the line length is almost a match to King's Focal test tubes. Those have nice bass performance even though the driver's F3 is much lower than the t-line's resonance. *shrug* I'm no t-line expert or anything, so if anyone with more experience wants to shed some light on the matter I'd love to hear it. The volume may indeed be an issue, since these are larger surface area drivers than the Tritrix was designed for. Yea, gonna use those Mathcad sheets...

[JasonP]

I was gonna say, I am in my mid-30's, and can still hear to at least 25 Hz.

That someone can only hear to 40 Hz in their late 30's is a surprise.

Jason, have you worked in manufacturing or something similar for a long time?

Not at all, I just blame my high-output low distortion Klipsch headphones, electronic music, and volume creep. Also, poor judgement. Note, I haven't done any serious listening tests, I just messed with my EQ on playback and cut everything under 40 and I really didn't notice it. It could have been my sample tracks too I suppose... Hmmm, well I have subs floating around if I need to add that last octave after all.

[Pete Schumacher ®]

Not at all, I just blame my high-output low distortion Klipsch headphones, electronic music, and volume creep. Also, poor judgement. Note, I haven't done any serious listening tests, I just messed with my EQ on playback and cut everything under 40 and I really didn't notice it. It could have been my sample tracks too I suppose... Hmmm, well I have subs floating around if I need to add that last octave after all.

Most music recordings won't have much content below 40Hz, and most speakers don't have output below that point as well. So applying a 40Hz high pass to music resulting in little change to the sound isn't surprising.

[JasonP]

Ok so actually, it may have been purely a test issue. I was using my Alessandro MS1 headphones (modified Grado 125is I understand) and if their response is anything like the 125i measured here: http://www.headphone.com/headphones/grado-sr-125i.php , their output drops -6 to -16 db over the last octave. That could explain the EQ 20-40 Hz test results, since I may already have just been accustomed to them barely being there. Testing with my speakers and subwoofer, I can detect the EQ change though not all that well. (I put on some Deep Forest - Boheme that has some solid electronic deep octave output). hah.

- Thanks Pete

[Pete Schumacher ®]

Ok so actually, it may have been purely a test issue. I was using my Alessandro MS1 headphones (modified Grado 125is I understand) and if their response is anything like the 125i measured here: http://www.headphone.com/headphones/grado-sr-125i.php , their output drops -6 to -16 db over the last octave. That could explain the EQ 20-40 Hz test results, since I may already have just been accustomed to them barely being there. Testing with my speakers and subwoofer, I can detect the EQ change though not all that well. (I put on some Deep Forest - Boheme that has some solid electronic deep octave output). hah.

- Thanks Pete

If you can get the Gladiator soundtrack, you'll be treated to tremendous dynamic range to go along with the very full octave ranges.

[JasonP]

So I picked up the Mathcad sheets from M. King yesterday, and just fudging in some numbers, the tritrix doesn't look like a good fit. There is a hump in the response right at the upper bass, centered at around 200 hz and 4 db tall. I'm thinking that might have something to do with the volume of the line. I'll be playing around with it tonight and see if I can't fudge in more fitting values, and see what a more optimal TL design would look like.

[Paul K.]

Based on my modeling TLs for the RS180 and building one having a pair of RS180s in separate lines, the TuLines, the RS180 works much better in an ML-TL than a tapered TL. That doesn't mean you can't use a tapered TL, just that an ML-TL gives significantly better results and is easier to build to boot. In the TuLines each line for an RS180 was 55" long, and had a cross-section of 7.5"W x 4.5"D. The modeled F3 was ~35 Hz.
Paul


So I picked up the Mathcad sheets from M. King yesterday, and just fudging in some numbers, the tritrix doesn't look like a good fit. There is a hump in the response right at the upper bass, centered at around 200 hz and 4 db tall. I'm thinking that might have something to do with the volume of the line. I'll be playing around with it tonight and see if I can't fudge in more fitting values, and see what a more optimal TL design would look like.

[JasonP]

Right, I imagine that has something to do with the lower qts of the driver, to my knowledge .45 or higher drivers work better in pure TL. I tossed a couple DA175-8 in the same tapered TL, and those worked much better. I'd switch to them, but then I'd have to lower the target sensitivity of the design. That's not a price I am willing to pay.

I'll take a look at ML TL then, thanks!