[mtnbob]

First, hello to everyone, been reading for a while, finally registered.
So here's my first question. I'm building a pair of U Frames with 15" woofers. I have 32" from the front of the baffle to the wall behind. As I understand, (been to John K's site) basically the deeper the U Frame, the better, up to a point - but given my limitation, what's the best trade-off of depth vs distance from the wall? I'm thinking about 16-18" for the depth, but does that leave enough distance behind?

[tubamark]

You really need at least 3' from the rear of the driver to the wall behind it.  Bass gets even cleaner at 5-7 feet.
Even a properly (heavily) stuffed U-frame as John K prescribes still behaves as a dipole over a portion of it's range (not as a pure cardioid).   Placing it closer than that will cause the rear output reflection to occur too soon, resulting in not-clean, disappointing bass.

If you have a rather wide front wall, you can place U-frames sideways and against the front wall--firing towards each other (not toward the listener).  listener sits within the equilateral triangle.  Frames would have to be 3' from sidewalls in this case, which is why your room needs be wide (or alternately, your listening triangle very small).
This alignment also maximizes output (same goes for placing against sidewall).  Bass quality won't be as tight/clean as when spaced from all walls, but will still be cleaner than any monopole.

'Hope this makes sense.  In addition to John K's site, I would recommend www.Linkwitzlab.com for everything that you need to know.

--Tubamark

[mtnbob]

Thanks for the info. I just couldn't seem to find a simple answer (like yours) to the question. I've been to Linkwitz many times,but usually get a headache trying to remember all that math I haven't used in 40 years. I could probably put them along the front wall, but would have to move the TV out another 8-10" to accommodate. Maybe I will also look at ripoles, since they're thinner although less efficient.

[johnk...]

A correctly damped U behaves more an more like a true cardioid as the frequency drops towards zero Hz. A correct analysis of a damp U shows that the damping results in a low pass filter on the rear radiation introducing a constant delay at low frequency. When correctly tuned the delay means that the rear radiation will exit the  u with the same delay as the propagation delay experienced by the front radiation resulting is cancellation. A u acts as a dipole at (very) low frequency only when it is undamped. 

With regard to u length, my experience has shown that the length should be such that the resonance of the undamped U -duct is at least 1/2 octave above the desired crossover frequency.

Whit regard to placement, if we consider only the wall behind the speaker and assume that the axis of the speaker is perpendicular to that wall we can make some simple observations. The wall acts like an acoustic mirror. Thus, if the speaker is X feet from the wall we can mimic the wall by imagining a second speaker X feet behind the position of the wall (2X feet behind the real speaker) with the back side of this imaginary speaker facing the listener.  At low frequency, when the wave length is much longer than the distance 2X we have the following results. For a dipole as X decreases the output form the speaker will decrease. As X goes to zero we would have the effect of two dipole speaker sitting next to each other, one facing forward and one facing backward and the low frequency output from sum of both dipoles will be zero. For a monopole woofer the radiation is omnidirectional at low frequency so it doesn't matter whether the speaker faces forward or backwards. Thus as X goes to zero the summed output will be that of 2 woofer side by side, a 6dB increase compared to an isolated single speaker. For a cardioid, the summed response as X goes to zero is warps into that of a single monopole. When we look at only the on axis low frequency response as a function of distance we find that the dipole SPL goes down, the monopole goes up and the cardioid stays the same as X decreases. This is of course an over simplification since real rooms have other walls. And with any speaker radiating to the rear there must be sufficient room for the rear source to radiated freely without being "loaded" by the wall. But this at least gives some indication why, in theory, cardioid woofers are the lease sensitive to room placement.

[JohninCR]

....This is of course an over simplification since real rooms have other walls. And with any speaker radiating to the rear there must be sufficient room for the rear source to radiated freely without being "loaded" by the wall. But this at least gives some indication why, in theory, cardioid woofers are the lease sensitive to room placement.

John, thanks for that analysis.  At least to my ears, I can definitely confirm that "theory" is backed up with real world experience.  For those of us chasing open alignments 100hz and down, but can't place their speakers 2m away from the front wall, I'm completely sold on U's.  In direct comparisons of U's vs dipoles below 100hz and placement less than 1m out, the more cardioid response sounds more like live especially for bass drums and bass guitars.  In my room, dipoles need to be out 2m from the front wall for that same sound.  Conversely, the U's don't sound as good to me with side wall placement in my deep narrow room. 

Also don't forget, with U's the room limit of dipole bass extension is far less of a factor.  This can be a major setback for dipoles in a small room with solid wall construction.

I don't have any H baffles, but a good experiment would be direct comparison of a fairly deep H with only the rear cavity damped like a U.  You'd probably need to cross lower, but you should gain some bass extension and freedom of placement.

[mtnbob]

Well, this is what I did - before I saw what the Johns wrote (BTW, thanks to both of you, I have learned mucho from reading your various posts)

I had an old speaker box sitting around, so I stuck one of my subs in it and listened. Really surprising. Ran some test tones and it was strong down to 40Hz, 30 was obviously down some, but still good, 20 was just barely there. Music ranged fro Dvorak and Strauss to Mark Knopfler, Crooked Still and Brubeck. I like to check bass out with tympani and acoustic bass. Sounded pretty good, a little bit of boominess up near the crossover (100 Hz)

But now, let me tell about the boxes! Just about everything is "wrong" with them. The internal dimensions are 21"x15"x10" deep. So the shape is wrong and they're too shallow. Being so shallow, I could only put a bit of stuffing in - only a couple of inches. Throw in a 15" driver firing through a cutout for a 12" and the 3" hole in the baffle where the tweeter was. AND they're only 20" from the front wall. I'll bet a lot of people would have told me it just wouldn't work, and frankly, I'm pretty surprised myself.

Now that it's stopped snowing (24" last 3 days) I can do something besides shovel snow for a while. I'm going to throw together a 16" cube and stuff it good, put the other sub in it and listen to both of them. Should be interesting.

One more thing, I don't have any test equipment - much too limited budget, retired teacher with no COL in 5 years. Got time, not much else.

[mtnbob]

I see this thread hasn't exactly been a hotbed of activity (guess the Johns were pretty much the last word), but I see people are still looking, so I'll report in. I got around to building a u-frame. It's a 16" cube. The stuffing is 2" of acoustical fiberglass, 2" cotton batting and 6" of polyester wrapped in 3 directions - left-right, up-down and diagonally. Kinda looks like a pillow.
I put it on the other side of the TV and listened. I thought I would probably hear some difference in the bass from side to side, but I couldn't detect any real difference?? I put in Fifth Element and explosions sound real enough and loud enough for my taste. BTW, I changed the crossover back to 80 Hz.
So I guess the moral of the story is the oft heard - " build it and listen" and don't agonize over the math and doing everything the "right" way.

[mcgsxr]

I appreciate the update, but lack the math or experience to comment, so call me one of the avid watchers of the thread!

I will likely be construcing similar dimension U's, how far are these from the rear, and side walls - just looking to get an idea of the placement in your room.

And what is the total room size?

Thanks,

[mtnbob]

mcgsxr (and others) I realize it's a really odd couple at the moment, but it is a experiment after all 

The old speaker cabinet is only 3 ft from the side wall. The 16" cube is by an opening into the dining area, so it's 15 ft from a side wall. They are both only about 16" from the front wall. Not ideal, I realize, but I'm sitting just 8 ft away from them, so ideal placement would put them almost in my lap.

The room is hard to describe, I'll give it a shot. The basic size is 22x15, but the end the speakers are on opens to that dining area, so it is like 26 ft wide there. The ceiling starts at 10 ft and goes to 22 ft. And then there's the loft, on the same side as the dining area, so at 10 ft high, the room is 22x26. The longest diagonal is 35 ft. I'll bet that's hard to visualize from my crappy description, sorry. Anyway, it's a big room, volume wise.

[mtnbob]

One other thing I forgot, when it comes to speakers, like everything else. I have to live by the teacher's first rule - "Do the best you can with what you've got"