The making of: The Two Towers (a 25 driver Full Range line array)

That's quite the combination! Given the relative sizes of the 2 breeds that would average out to a normal-sized cat.

That's pretty accurate... she still fits in the most unlikely of places...
attachment.php


You're going to have a lot of hair stuck on the intake vents of your computer!

And everywhere else in the house! :)

The computer is fine and protected(*), it's not in an area where the cat houses... but you're right about the rest of the house! :eek:

(*) = Only the keyboard/mouse/monitor are in the living room, the PC is in our corridor. Less noise to worry about :).
 

Attachments

  • cat-small.jpg
    cat-small.jpg
    185 KB · Views: 662
Right now there is a fun discussion going on over on the other side of the Loudspeaker forum: Dynamic is Preferred over Electrostatic

While the title could be misleading, the resulting discussion is interesting. Rather than throwing out my opinion right there, I'm more comfortable to sprout it on my own thread.

I've said before, what I look for in a replay system is for it to do the "soundtrack of my life" justice. Meaning it has to be able to play a lot of different sorts of music and still do well.

So how do we get there? My opinion would be: take your time.

I've been playing with ambience again and had my reverb settings longer than ever, enjoying many songs this way. Last Friday I had a visitor over to listen. Most songs we played sounded pretty good, until I threw in "September in Montreal". While that song certainly does not belong in the soundtrack of my life, it has been used a lot to judge my system in the past. One thing it did make obvious: my ambience settings were a bit too much.
There's a slap echo in the song, that bounces off the (virtual) right wall from a drum being played on the left. My ambience enhanced that blow out of proportion at the back. However, this morning I meant to fix that.

I rearranged early and late levels, and more importantly changed the slope of those levels making lower sounds resonate longer than the higher ones.
No more exaggerated slap echo, but awesome depth on the songs I tried this session. So, is the recording to blame? Or was it my (virtual) room response. Yes, Koldby, I know how you feel about that song ;). Anne is doing something out of the Exorcist while sitting behind that piano. :)

All I'm saying is that the room is as important to what we hear as the speaker system is. They have to play well together to stand any chance on random material. I'll keep working on my dream, can't wait to add the subs!

Do you hear more flaws on a good system? Lets first make sure that it was a flaw on the recording, not something in our system or room before we judge. Do I hear differences in recordings? You bet I do! But I find more and more recordings that work, the longer I keep working on and in my room.

Just saying... spend a little more time on it before ruling out certain songs.
Are there badly mixed songs or recordings out there? No doubt. But good and well recorded songs are more than a handful!
 
Administrator
Joined 2004
Paid Member
I'm in agreement with you there. Excellent systems in excellent rooms make many more recordings enjoyable. It doesn't make bad recordings good, it just makes their faults more understandable and less objectionable. With a superb room and system and room combo you can hear past the recording, to the music.
 
Using everything on hand as tooling:

attachment.php

(it is good to save everything :))

Cutting the sides for a brace:

attachment.php

(not showing the actual brace)

Test fit:

attachment.php


More testing:

attachment.php

(yes I know the logo is upside down ;))

Good progress, but not nearly done yet! Lots of material on its way to complete things, hopefully more to come soon!

P.S. Yes, the 30W is an excellent fit into the cut-out. I did try that to be sure. The baffle will be 48mm thick in total with the woofer in-between the panels. I think it's going to work out nicely.

P.P.S. I know I said no pictures, but I decided to just take my time instead of racing to the finish. I was in a hurry to be able to get paint on before the winter sets in. I guess a layer of colored epoxy will have to do though.
 

Attachments

  • Tableguide.jpg
    Tableguide.jpg
    228.7 KB · Views: 521
  • Bracecutout.jpg
    Bracecutout.jpg
    208.8 KB · Views: 511
  • subsides.jpg
    subsides.jpg
    236.4 KB · Views: 511
  • 2subs.jpg
    2subs.jpg
    206.4 KB · Views: 587
Last edited:
Ronald, I'm curious about something. I noticed your brace panel appears to be bisecting the side and top walls. I'm sure this will raise the "excitement frequency" of each panel high enough so the frequency range covered by the subwoofer won't excite the panels and cause a vibrational node but my practice has been to apply bracing at off-center points to avoid a double excitation at a higher frequency. Is this something you (or others) normally practice as well or does it simply depend on the design? Your Two Towers have more than adequate bracing between drivers even though they are evenly spaced, and your brace holes were alternately offset as well, although I believe that was more for damping purposes than anything else. Continued thanks for all of your knowledge sharing in this thread and on DIYAudio in general. The subs are coming together nicely!
 
Looking really nice!
Did you ever decide on the power amp for the subs?

Thanks, sanford1

Yes I did. I decided to get two of the Hypex NC500MP OEM modules. Member koldby mentioned the 250 watt amp as a possibility, I wanted to have slightly more reserve. It felt wrong to have 250 watt for each sub, with each array having 350 watt available. Each sub has double the volume displacement of a single line of 25 TC9's.

Ronald, I'm curious about something. I noticed your brace panel appears to be bisecting the side and top walls. I'm sure this will raise the "excitement frequency" of each panel high enough so the frequency range covered by the subwoofer won't excite the panels and cause a vibrational node but my practice has been to apply bracing at off-center points to avoid a double excitation at a higher frequency. Is this something you (or others) normally practice as well or does it simply depend on the design? Your Two Towers have more than adequate bracing between drivers even though they are evenly spaced, and your brace holes were alternately offset as well, although I believe that was more for damping purposes than anything else. Continued thanks for all of your knowledge sharing in this thread and on DIYAudio in general. The subs are coming together nicely!

Hi Craig,

Nothing wrong with your reasoning here. After looking at sub enclosures from both Troels Gravesen and Siegfried Linkwitz (may he rest in peace) I figured Linkwitz had no worries about non-braced lengths of panels that were longer than mine. So I looked up the corresponding wave length of my top/side panels. It's about 942 Hz. Gravesen went wild with bracing but also had passive radiators in his design (and the smaller 10" sub). I wanted to maximize my internal volume, so less internal bracing would help, but I did decide to do one (because otherwise both the length and the width of a side or top panel would be identical, so that would mean all dimensions would be the same).

But I did stress a little less about having to place the brace out of the center (which was my original plan), due to the non-braced panel being this high up in frequency already.

Further more, there will be 8x M8 rods all the way through the enclosure, holding the baffle and end plate in place. They could, in theory, be used to tune the enclosure by tightening them. In contrast of both the above example sub enclosures, I do intend to add internal damping materials, both real wool felt on all walls as well as fiberglass insulation. As my external dimensions were as good as fixed from the start, I want to use this damping material to 'fool' the driver into 'thinking' the volume is larger than it actually is.

So, yes, because we're only talking about sub frequencies here, I stepped away from the need to place the brace off-center. In a mid enclosure I would have done the off-center thing though. This placement was an easier way for me to make the cuts with a 8mm router bit. Just 2 cuts just by flipping the panel.
 
...That would prove very handy. :nod:

Think its really not so difficult when knowing sound travel 343 meter in seconds and frq swing is also in seconds, then set up 343 / 942Hz is aprox a panel lenght of 0,364 meter the other way around having the lenght is same 343 / 0,364 is aprox 942Hz, further on divide with 2 to get half wavelenghts, divide with 4 to get quarter wavelenghts etc, also for such cases its handy in most computer operating systems one can have as many calculators opened at one time as computers screen area can handle.
 
Do you have an online reference for the wavelength chart/calculator you used? That would prove very handy. :nod:

BYRTT is right, it is as easy as he shows there. However if you're lazy you can just type in the numbers here:
Wavelength frequency convert lambda Hz sound conversion acoustics acoustic audio radio measure speed of sound and radio typical waves wave length light vacuum equation formula for frequency speed of light color electromagnetic spectrum - sengpielaudi

Another valuable page from that site is time delay per distance:
Time difference per sound path distance ms per meter or length millimeters time of arrival milliseconds calculation calculate delay line noise sound wave in air calculator variance ITD Haas effect duration - sengpielaudio Sengpiel Berlin

To make it complete, here is the calculator for both wavelength and cycle time of a sound wave:
Wavelength acoustic sound wave air sound waves frequency calculation temperature wave no air pressure speed of sound - sengpielaudio Sengpiel Berlin

That last one can come in handy for spotting reflections in your IR and knowing how many cycles are completed before that reflection.
It makes reading the REW plots a little easier, as you get an idea of what happens over time and why.
 
Last edited:


Well, apparently the Internet doesn't want to indulge my laziness. The only answer I get is "NaN" no matter what frequency I enter or which browser I use. That being said, just having the formulas and equations for sound is enough and the wavelength line graphs are a great quick reference.