Baffle simplicity yields sonic dipole benefit?
Over the years I have read Siegfried’s excellent material several times eg www.linkwitzlab.com/rooms.htm.
After several years of planning and saving etc, a room (6.7 * 5.3 * 2.6 m, 22 * 17 * 9 feet) mostly for music, is now available.
I am now planning to use some great drivers that I already have to build a higher output (HE drivers off tube amps) semi active dipole system.
With highish ceilings and a fair degree of human & animal floor traffic, I’d prefer to suspend the baffle off the ceiling. With generous space, I want to:
1. Minimize loss of efficiency
2. Minimize EQ
3. have a simple & cost effective baffle structure.
Material aside, the simplest way to build a large baffle, with WAF, especially off a high ceiling: is a *single transparent baffle.
Transparent dipole baffles http://www.diyaudio.com/forums/showt...ht=#post343896 where Nuuk commented:
When I designed my open baffles, I had some reservations about using perspex or other plastics. Apart from the possibility of colouring through aging/exposure to ultra-violet light, there is a real possibility of scratching them.
And there is the cost of course! In the end I went for glass”
If a single large baffle (for discussion in principle, say 10 feet wide * 2 feet high, circa 3 * 0.6 m) were used for dipoles: how would this perform compared to two separate baffles of similar total area?
Part of the answer to your question depends on the specifics of your proposed installation and the frequencies that the driver(s) in the baffle will be covering.
Part of what is likely to happen with a single large baffle (for two channels) is there will be two waves traveling into a collision across the front and back of the baffle. What happens when they hit each other will be the main question and will be related to the frequency.
If the two baffles are separate, there may be an effective discontinuity depending on the distance between the two baffles.
The second issue in the single baffle will be the sound that travels through the material and sums or cancels... usually the speed of sound through materials is higher than the speed through air. This may or may not be important.
Also the 2 foot dimension will play a big role compared to the 10 foot dimension with respect to sound within the mid bass...
Take a look at the IEC baffle response for some clues here...
I am looking at this right now, as a builder is very soon to build two small storage lofts in this room, and I want to make sure these do not impinge on enough space for the dipoles.
The ceiling height now is 12.5 feet (3.8 m), and the lofts as currently planned would reduce the ceiling height to 9 feet or 2.6 m.
To experiment initially I will use an adjustable electronic crossover. I will use the Phoenix woofers that i have.
The dipole “add on” to the Phoenix woofers would be likely two drivers:
- midbass (say 80 – 400 Hz) with high Vd 98 dB Lambda TD15M’s
- say 400 – 2000 Hz with 98 dB PHLs.
(Driven by a 12 watt 845 SE amp).
So my question concerns from about 1800– 80 Hz, ie wavelengths of about 0.6 - 14 feet (0.2 – 4.3 m)
Does that make it any easier to predict?
I think filtering can overcome any problem here, though it's hard to predict the results as interaction with sound and anti-sound is very much dependent on reflected waves in the room as well.
Therefore I think placement may be even more critical than the differnces between the two baffles. I think some lower frequencies might be a bit increased in your example, but that-again-will probably be easy to solve with the filter.
My dipole experience is limited to midbass and high frequency electrostatic speakers though, where changes in placing of only centimeters can make a large difference and such intense filtering is not necessary...
If the dipole is for bass, why not use an IB set up? or quasi IB?
That assures you of excellent LF extension and avoids almost all of the other problems associated with bass and dipoles...
My feeling is that EQ will almost certainly not solve many of the peak and valley issues likely to occur with a narrow and long baffle... in the LF...
Above 80Hz, ur probably going to be ok, but pay attention to the transitions from the baffle to free space, as the freqs go up so does the diffraction... and you can use the typical "baffle step" calculators on the web to see where the dimensions will have an effect...
maybe the room acoustics can help getting more bass, but eq-ing (filtering) will be necessary anyway...
Here's how you can keep it simple and affordable:
The link appears to point to basic "crossover filters."
You'll need to take into acount the actual acoustic response in the design of any real-world crossovers.
To make the 80-1800 range with little EQ you don't need a room-width sized baffle. My shallow U-frame baffles are 48" high, 24" wide and 7" deep, and true rolloff starts at ca. 100-120 Hz (don't forget that from 200 to 100 Hz you gain ca. 6 dB from floor boundary effect). You could make two tall perspex panels with a thick wooden frame around them for stability (flexing) in the 30" x 60" range and you'd be covered to be almost dipole EQ-free. Driver low Q EQ might still be needed though. Also beware that you don't plan to position the speakers too close to the rear wall - IME 4 feet is the minimum to get an airy sound.
> an IB set up? or quasi IB?
as in, in a roof space? Alas it’s a cathedral ceiling – no adjacent enclosed space . .
> shallow U-frame baffles
Do you have a drawing or pic?
> from 200 to 100 Hz you gain c. 6 dB from floor boundary effect)
~ again, to leave the floor uncluttered, mine will be suspended from a cathedral ceiling, so no floor boundary gain
BTW, boundary gain, and with his final update at the end of the article, driver Qts is done very scientifically by Martin King here: www.quarter-wave.com/Project07/Project07.html
I wish it were in Excel :bawling:
> You could make two tall perspex panels with a thick wooden frame around them for stability (flexing) in the 30" x 60" range and you'd be covered to be almost dipole EQ-free.
That’s basically my aim. Just need to get the maths right!
> Driver low Q EQ might still be needed though.
As proven Martin in www.quarter-wave.com/Project07/Project07.html, low Q drivers need EQ.
FYI I want to use these two drivers I have, - see attach.
the floor boundary gain should come in roughly when the distance to the floor is less than 1/2 wavelengths, regardless of whether they are standing on the floor or suspended. If you suspend the bass drivers high in the air, the effect will be shifted towards lower frequencies.
The low Q EQ is relatively easy to calculate and achieve - either using Linkwitz transform, or for very low Q drivers, using a shelving lowpass (as for the dipole EQ). JohnK had a post very recently on one of the forum pages how to calculatue just that, if you do a search.
For the baffle dimensions needed (excluding Q effects) there is
with the relevant formulas under "design models". For a simulation program there is
which should give you a good approximation of what's needed. Final response of course should be measured on a test baffle.
And finally, a pic of my baffles:
|All times are GMT. The time now is 08:15 AM.|
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2013 DragonByte Technologies Ltd.
Copyright ©1999-2013 diyAudio