dipole stereo bass?

[Kensai]

John,

Space savings, yes, but primarily just trying to learn.

I've got tons of questions, but forum searching is taking way too long and I'm not getting anywhere fast, so when I get the chance inside the framework of an active thread I try to sneak some in.

How are your 12"ers mounted (I mean, what sort of configuration are you using)?

I've also been looking at ways for altering a driver's Qts to make them more suitable for OB use. I've found a calculator for adding a series inductor to a normal woofer that gives a new, increased Qts depending on the DRC. I've also found some old discussion on using a dual voice coil woofer and using one coil as a brake to increase Qts (resistor or a pot wired across one coil while the other coil is driven). In that discussion, it sounds as if dual voice coil drivers, with only one coil driven, have double the Qts. This was never stated explicitly, and it was never stated whether that was just one coil driven and the other open or with the non driven coil wired closed.

Being able to raise the Qts of a woofer would theoretically open up a very large number of new driver possibilities for OB use, but the first question is, would artificially raising a driver's Qts using any of these techniques actually work to make a driver more suitable for OB? Second, I guess, would be, is it true that driving just one coil of a DVC driver actually double the listed Qts of the driver, and if so, should the second coil stay open or should it be wired closed?

Thing that kills me about this is that I can't imagine this sort of thing not being of interest to alot more people than it seems to be. Searching around various forums and googling about really only seems to bring up more questions. You just see people discussing things but they're not letting outsiders in on the basics of what they're talking about. They've already figured out all this stuff and are debating finer points which leave me trying to backtrack to the core truths of the discussion, but even then, I can't just take my conclusions on faith. It would be great if there was somewhere to go to just get concrete answers to simple questions like these.

If I had all my answers, I could actually figure out what I intend, build it then come back here and show it off for others to learn from (for good or ill ).

Kensai

[johninCR]

Kensai,

I discussed my cab earlier, but it's a double stack W-baffle with push/pull alignment and a 9" extension on the back making it a hybrid W/U-baffle.

You add a series resistor to raise Qts, not an inductor. An inductor is a 1st order low pass crossover. Yes, typically the Qes of a DVC driver is listed in the specs with both VC's connected. Leaving one open and unconnected doubles the Qes, which essentially doubles the Qts. Making some kind of connection across the terminals of the open VC will make it act as an electromagnetic brake and lower the Qes. The more resistance, the less braking.

Linkwitzlabs.com has almost all the answers and science regarding dipoles. There's not a lot of other information because OB's are just beginning to become more mainstream in the past couple of years, and most don't really understand their operation.

[Gary P]

Originally posted by johninCR
Nice work on those measurements and I'm sure the distortion reduction of push/pull would show even greater in a uniform setup, but your measurements already show it is significant. I've only done P/P with some cheapie 12's that I have a bunch of, and I swear the difference is audible with them. I speculate that P/P alignment becomes more important with cheaper drivers, since I'm sure their operation is less linear.

BTW, it sounds like you did a lot of measuring before starting your OB project. What type of design are you using?

Thanks,

Yes, I've made many measurements, hours and hours of measurements...

The system as it stands now is using 4 of the Dayton IB15 drivers, 2 each loaded into U baffle that have the drivers mounted PP in the W configuration. The U portion of the setup is damped with 4 layers of 3.5" recycled cotton building insulation.

Trying to come up with a damping method that could deal with the amount of air that these drivers move was an interesting challange. What ended up working well was the cotton building insulation. The bats are stapled to strips of 3/4" plywood on one edge. The edge with the strip of plywood is screwed to the top of the baffle leaving the bat hanging. The spacing of the bats of insulation is set so that there is ~1" air space between each bat. The fit of the bat in the baffle has to be just right... Too tight and the resonant frequency increases due to the resistance of the damping. Too loose and the resonant frequency decreases from the damping mass coupling to the drivers. With the bat lightly touching the sides and bottom the impedance at resonance is damped without moving the frequency up or down.

The resonant frequency of theses drivers in the W configuration is ~14Hz.

At full output at low frequencies the bats are moving up to 1/2"P-P.

How do they perform? I'll let the measurement speak...

Bass response measured 5-13-06

The current project that the OTS (Oregon triode Society) is a subwoofer. We have had several listening sessions auditioning different subs. The ones that impressed me the most were the open baffle subs. That was the catalyst that got this project moving. 2 weeks ago we had another listening session to compare the more common H frame OB sub with the damped U baffle.

The H frames were loaded with another set of 4 Dayton IB15 drivers in the PP arrangement. I generated a crossover curve on the Behringer 96-24 for the H baffle to get it as close to the damped U baffle as possible for the comparison. The match was very close with the exception of below 20Hz where the H baffle was rolling off.

Measurement comparing U to H

We started with the U baffles, played 3 songs, swapped in the H baffles, played the same 3 songs, then swapped back to the U baffles and played the same 3 songs over yet again. The comments after the H baffles had been switched in were mostly along the lines of "Wow, it only took a few notes to tell the difference..."

After the U baffles were swapped back in and the songs played the general consensus by unanimous decision was the U baffles out performed the H baffles by a good margin. Tighter, cleaner, more detailed were the most common comments.

The 3 songs played were
1. Les Vic, Ecstasy In Numbers Live, 2003 Diversity Records, Ltd. (CD)
2. Finlandia, Thomas Murray, The Transcribers Art, Gothic Records, (CD)
3. Fall 77, Oregon, Out Of The Woods, Electra/Asylum Records, (Vinyl)


The Ecstasy In Numbers cut has a kick drum that causes the whole room to shudder. The reverb of the kick drum gives the impression that you can feel the size of room the recording was made in

The Finlandia recording is recorded on pipe organ and has some of the deepest bass I’ve heard so far… Great fun.

The Oregon recording is just plain nice! All acoustical instruments. The bass that Glen Moore plays must be a 5 or 6 string as it goes lower than any string bass I’ve heard. Glen does some nice runs up and down the bass on this track that go from the middle 20’s to over 200Hz (guessing). Great test for sub integration.

Image of front of system
Image of rear of system

Another point of interest is the room interaction between the 2 types of subs. My listening room has a short hallway behind the left speaker that goes to 3 bedrooms and a bathroom. The frequency response of each speaker is influenced by the state of the doors.

The H frame woofers were very well behaved with all the doors closed. With the doors open the response was not as nice.

The Orange trace is the response with the doors closed, the blue trace is the response with the doors open.

Measurement comparing the H frame with doors open and closed

The U frame woofers were not influence nearly as much. I don't have a measurement of the current crossover setup with doors open so the comparison for the U baffle is from an earlier setup.

Once again, the Orange trace is the response with the doors closed, the blue trace is the response with the doors open.

Measurement comparing the U baffle with doors open and closed

Gary

[Kensai]

Thanks John.

I don't think I meant to say inductor there, but in the back of my mind I was wondering about that . . . inductors have DRC values, so if you were using a passive xover and your lowpass had high enough DRC (probably needing to stack alot of high gauge, small value air core inductors to get a high enough value), that should allow you to adjust the Qts upwards, too, right? What is the downside of a high DRC lowpass section (1st order, of course).

Also, what are the effects of running a DVC driver on only one coil (leaving the second open seems like a good idea to me right now)? Do we lose sensitivity? Is there any effect on sound quality?

Kensai

[johninCR]

Gary,

Wow, 2 IB 15's per side must really be something. I think you should know that there are ways around the use of cavity damping in your U-baffle. I use a straight pipe and angled specially shaped rear edge with my 15" coax OB with no damping and no resonance despite the rear wave running from up to over 3khz. Angled sides so the terminus is larger can also prevent 1/4 wave resonance, which I have also used along with different length sides. This has the added benefit of guiding the rear wave more to where you want, which may help with your hallway issue. A TL expert told me a 1" expansion for each 6" of depth is enough to eliminate resonance.

With the kind of force that the IB15's create, I'd feel compelled to do a W/U-baffle hybrid and I think I could fit it in about a 17" width, so I could cancel the mechanical forces and retain push/pull.


Kensai,
You just loose some max power handling, which is out the window anyway with OB because you run out of excursion long before you get near each VC's thermal limit. Also, efficiency is a meaningless number with OB woofers. Swept air volume determines output down low.

[Gary P]

Yep, the 4 IB15's do move a lot of air. It's been fun watching the looks on folks faces when they experience them for the first time.

The damping is something that I want. I started out with just the W frame running regular dipole. Then the extension got added to bring up the efficiency some. The extension did not improve things as much as the calculations led me to believe, but after reflecting on the issue it makes sense.

If I remember correctly, the calculations predicted about 10dB gain from the extension, more gain than the ~2.5dB measured. After some thinking I think I know what's going on. The calculation is for a baffle in free space. With the W sub one side (the floor) is already totally blocked so no gain there. The other 3 sides are open for improvement. In this case the path over the top is already much longer from the mains sitting on top of the subs. This gets us down to only the path around the sides is available to lengthen.

The frequency response was the same with the extension added to make the system into a U baffle, just higher output.

Measurement of efficiency change from 16" extension

After reading the information on John K...'s pages Music and Design I wanted to try adding damping to see what the difference in sound was configured as a cardioid as compared to OB.

I measured the in room frequency response and the Thiel Small parameters as each layer of damping was added.

Two things stand out.

1. The frequency response shifted from the same as the H frame with doors open to U baffle with doors closed.
Frequency response with varying damping

2. The phase shift at system resonance changed from +/- 60 degrees un-damped to +/- 20 degrees damped.
Phase change at resonance

As expected, the impedance curve also got much flatter at system resonance, dropping from 32 ohms at resonance to 7.5 ohms at resonance.

Having heard the difference between un-damped OB bass (both H frame and U frame) and damped OB bass from a U frame, I won't be building subs without damping in the future.

I'm basing my choice of compromises on the following points:

The bass sounds better to me (and everyone else who has heard them) with the damping. Cleaner, tighter, more detailed.

The bass goes deeper with the damped U baffle. Starts rolling off at the system resonance of ~14Hz and does not change much with room volume change (doors open or closed).

Gary

[ScottG]

Fantastic information Gary!

VERY much appreciated!

I'd bet JohnK would like to see this.. I'll post something on the mad board now to try to inform him..

[johninCR]

Gary,

For OB woofers I've always put some resistance in the back too. I couldn't with my single drive unit coax, so I figure out a way around it. Your work explains why I've preferred the woofer U's with damping, so thanks so much for sharing.

Also, I recently read that a dipole rolls off steeply below the longest room mode. After burning some grey matter on the subject, I decided that I agree except that I believe it's the on axis mode which is important. In addition, I believe placement anywhere near the front wall is detrimental to deep bass dipole extension.

I think the dispersion difference is also giving us deeper bass with the U alignments. With the wider rear wave dispersion, there is less of the rear wave for on axis reflection to net against the front radiation. Thankfully, I've yet to hear any stimulation of room modes other than on axis, so the added bass response seems to be without cost.

[AJinFLA]

I was going to type a long answer as to why this "test" is fatally flawed, but I see that John K already answered.

JPK) Hi Scott. I took a look at the thread. It is good to see others experiment. However, it is difficult to draw conclusions unless care is taken to carefully match the anechoic on axis response of the different woofers as I did with the CRAW design. And I don't think a general conclusion can be made about which woofer design sounds better, U, H, sealed or ported box, or even TL. They will each have different characteristics in a given room, and with variation such as open or closed doors or windows.

The fact that the subjective response was skewed to the sytem with greater tactile energy comes as no great suprise.
To me at least .

cheers,

AJ

[johninCR]

AJ,

Actually it's that kind of approach that is fatally flawed, and why the design of OB's isn't advancing as fast as it should. It's left to throwing an excess amount of drivers and fixing everything electronically. That's not how it's done with box speakers, so why should OB's be any different.

While anechoic measurements can be a useful tool, only in room measurements have any real meaning because the room is actually part of an OB speaker. If the guys with the right tools and experience would step in with anechoic measurements along with in-room measurements and explanations that reconcile the differences, it would be beneficial.

Let's get the ball rolling with a question. Does my earlier speculation hold true, that dipoles can't support a wave below the on axis room mode, but a U-baffle alignment can? Why please?


Where is this "mad board" ? It sounds like there's useful info there.