They actually have 4 woofers. What that accomplishes in practice, I can't say.
Legacy Audio Whisper loudspeaker | Stereophile.com
The main portion of the speaker consists of two baffles, mounted 2¼" apart. The rear baffle holds the second set of open-air woofers, which are mounted back-to-front; the front baffle holds the front woofers, four mid-woofers, and two tweeters.
The Celestion paper is attached to bolserst's post 17. He talks about some of his experiments and links to a discussion by Linkwitz (who didn't seem that enthused about the setup from a theoretical standpoint).
I've not tried this kind of open compound woofer dipole myself, so can't say what it does or doesn't do.
https://www.diyaudio.com/forums/multi-way/46111-legacy-whisper-bass-2.html#post3897099
I've not tried this kind of open compound woofer dipole myself, so can't say what it does or doesn't do.
https://www.diyaudio.com/forums/multi-way/46111-legacy-whisper-bass-2.html#post3897099
Surely Xmax is a defined maximum, (a design limitation), rather than that which may be reached by any given drive voltage.
My suggestion was meant to describe the possibility that the woofer may not reach fully the theoretical position for a given drive signal when alone, but being assisted the other, may reach a point nearer to the theoreticallly correct point, a better approximation.
Xmax is normally pretty easy to get to or exceed without large power inputs if you are running low frequencies for the driver. Depending on how it's specified and how the driver is designed, you can have very good linearity up to it, and decent a bit beyond it.
Xmech, being the maximum mechanical excursion of the suspension, is harder to run up against in typical use with a well designed driver that's not being abused. Xmech that's twice Xmax is pretty common.
Last edited:
Not completely sure but I think by having two drivers spaced apart you end up with a smaller baffle acting like a larger baffle. The dipole baffle cutoff point is a bit lower without having to use a larger baffle width.
I suspect this is its greatest "advantage". A smaller dipole wich would make a dipole more desirable in a world that want smaller speakers
Also by using high mass diaphragms ( compared to stats and planer magnetic s) the diaphragms motion is not as effected by the rear wave that bounces off rear wall an interacts with the diaphragm
I did an experiment once using very high mass planer magnetic diaphragms, probably around 3 to 4 times mass of typical planer diaphragm. These were limp , no tension diaphragms. they were the "fastest " sounding bass transient I ever heard. Like a live drum. The sensitivity was so low it was usless as a marketable product. They were also extreemly well damped. There was more rubber than aluminum in there just to give u an idea of the construction. It was like a sheet of lead hung by its own weight ha
I suspect this is its greatest "advantage". A smaller dipole wich would make a dipole more desirable in a world that want smaller speakers
Also by using high mass diaphragms ( compared to stats and planer magnetic s) the diaphragms motion is not as effected by the rear wave that bounces off rear wall an interacts with the diaphragm
I did an experiment once using very high mass planer magnetic diaphragms, probably around 3 to 4 times mass of typical planer diaphragm. These were limp , no tension diaphragms. they were the "fastest " sounding bass transient I ever heard. Like a live drum. The sensitivity was so low it was usless as a marketable product. They were also extreemly well damped. There was more rubber than aluminum in there just to give u an idea of the construction. It was like a sheet of lead hung by its own weight ha
Last edited:
Impressions and vague wordings are swell but measurement is nicer.
Much in that post (and a whole lot of imaginative* speculation throughout this thread) that grates against my intuitions. So it would be nice to learn either (a) some measurements or (b) some indication what amount of credibility is due lowness by virtue of training or experience or otherwise?
B.
* "imaginative" is the type of word that a very polite Canadian would use instead of......
It is my practice to write "My guess is..." or ".... just guessing" or something similar when I am "speculating". I should prolly add that to all my posts that have no measurements attached.
Real healthy for writer to know when they are BS'ing and even more important for readers.
B.
Here's a question I'm hoping someone can answer. In another thread acoustic lenses were brought up for ESLs. It's pertinent to this topic so I'll ask it here. If this principle was applied, could it be effective in preventing the cancellation effect by wrapping the lens around the the sides if a hyperbolic recess was located at the cancellation point. So for example on a single panel like an Acoustat Model 1, the lens would literally wrap around the panel 360deg with recesses front, back, and sides. This would help horizontal dispersion but would it serve to cancel the cancellation at the sides?
It is a commonly held illusion that in dipole settings, the rear wave sneaks around and substantially annihilates the front wave.
Just 10 seconds of thought about wave-lengths and wave math will clarify how silly that notion is. For example, above a certain frequency band (varying with size and shape of the baffle), half the time the front and back waves will reinforce one another and even that will vary as you go around the periphery of the baffle. Now you need only 5 seconds of thought.
Next time somebody is building a boxy cab, they should test it with the back off and on while located just where a dipole would be.
B.
Last edited:
You may need to draw a quick sketch of what you are proposing.
As long as there is a relatively short acoustic path between the front and the back, you are going to have cancellation at low frequencies no matter the baffle or lens shape. It's that "above a certain frequency" bentoronto mentioned that tends to be the problem for full-range electrostatics or dynamic dipoles.
Equation in top image here gives the basics for an ideal point source dipole:
Electro-acoustic models
If you cram it into Excel successfully, you can plot the basic destructive/constructive interference pattern for different widths. Image 2 shows the result with 10 inch spacing for the point sources. D is roughly half of baffle width (or driver diameter for an unbaffled unit).
Things are messier than this with real-world devices, but at the lower frequencies it gives you a feel for what's happening with cancellation.
As long as there is a relatively short acoustic path between the front and the back, you are going to have cancellation at low frequencies no matter the baffle or lens shape. It's that "above a certain frequency" bentoronto mentioned that tends to be the problem for full-range electrostatics or dynamic dipoles.
Equation in top image here gives the basics for an ideal point source dipole:
Electro-acoustic models
If you cram it into Excel successfully, you can plot the basic destructive/constructive interference pattern for different widths. Image 2 shows the result with 10 inch spacing for the point sources. D is roughly half of baffle width (or driver diameter for an unbaffled unit).
Things are messier than this with real-world devices, but at the lower frequencies it gives you a feel for what's happening with cancellation.
I have the speaker on the left, the ML Montis. It’s of course a hybrid design. As a guy who designs conventional loudspeakers, I think the Montis is very dynamic.
It’s impedance profile can be very complex, but with it’s powered woofer, even my 3 watt SET can blow me out of the room. For example, when I’m listening in the dark, a tympani being struck, which is mostly in the woofer range, but appears to emanate from the ESL panel, it’s a thunderous diapason. It’s a psychoacoustic trick, but the design makes the loudspeaker very dynamic.
Now, the presentation is generally along the plane of the panel, not at the listener. But, it’s a very interesting presentation that really should be heard.
Cheers,
Greg
I think in that case, you'd wind up with D (in the Linkwitz equation) going from the red line length in the image below, to the green path length.
Rough numbers, if you started with an overall panel width of 20 inches, D would be 10 inches (the red line), so low frequency roll-off would start around 700 Hz.
If the green path is 2.5x the length of red, roll-off would start around 280 Hz.
At higher frequencies, the acoustic lens would have more influence, but at low frequencies I doubt it would provide much benefit.
Rough numbers, if you started with an overall panel width of 20 inches, D would be 10 inches (the red line), so low frequency roll-off would start around 700 Hz.
If the green path is 2.5x the length of red, roll-off would start around 280 Hz.
At higher frequencies, the acoustic lens would have more influence, but at low frequencies I doubt it would provide much benefit.
Last edited:
As Einstein said, you want your theory as simple as possible but not too simple.
Bafles have tops and bottoms and long edges. You need to think how the sound changes with each incremental location around the periphery. And that's just part of the analysis because you have all that sound bouncing off the ear wall in total phase chaos. It's the same non-sensical result you get from belief in comb filtering off adjacent walls that seems inaudible to me.
I suspect most dipole enthusiasts will tell you they think the sound is distinguishable from point source boxes and a lot better.
My guess is that if you applied the same micro-analysis to Rice-Kellogg drivers (AKA cones), you'd conclude they couldn't possibly sound right. And if your analysis started with the reed in the clarinet......
B.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.