I simply cut the rear fibreglass lighthouse down to a 11.5" diameter throat and mounted the Vitavox drivers in an enclosure on a flange. I tried it with and without the top on the enclosure - preferring top cover off marginally. The PM2As did a very good job and measured well. However the K15/40s subjectively have more power and dynamic range. All the drivers I have tried look much the same in the bass with a sweep tone - obviously horn determined. But I expect it is all about transients in the bass / lower mids. I am building pp300B amps to see if I can push it out further with more headroom. The long horn had a lot worse time alignment and worked best with the dbx digital cross over bi-amp set up I was using then. I like the simplicity of a single amp - I run the mid horns off a lower tap on the OT. All very crude stuff - I make no claims.
An externally hosted image should be here but it was not working when we last tested it.
There was a discussion some pages ago about mechanical resistance (Rms) and how it affects low level dynamic.
Using a pair of drivers will yield to a "virtual driver" with two times the Rms. So in theory in could reduce low level dynamic, while of course increasing high level dynamic and headroom...
Speaker cones are close to acoustically transparent, with only a few dB of loss. (How much acoustic isolation would you expect from a wall made of paper?)
If the two (or more) drivers share a common rear chamber, the out-of-phase radiation from one driver will go through the cone of the other driver, with a bit of additional time delay thanks to the transit time between the two drivers. The net result is degraded time response and (fairly small) ripples in the frequency response. Subjectively, the midrange is muddled and confused-sounding, as you would expect with time-delayed artifacts in the 1 mSec region.
The solution is to have an isolated chamber for each driver ... although the bass vents can be shared, since the contemplated vent system has both vent damping and a folded path from the inside of the enclosure to the outside.
This was a lesson-learned from the Ariel ... an early version with a shared chamber sounded quite bad, while measuring pretty well. The rear-wave crosstalk isn't obvious on FR measurements unless you're looking for it, while the time-domain degradation was somewhat more obvious and clear with respect to the phase of the delayed rear-wave. (One of the things to always remember with FR measurements is the phase and time information has been discarded ... and phase and time information can be useful for diagnostic purposes.)
Although it's easy to abuse time-domain information for marketing purposes, it is an excellent diagnostic for tracking down colorations that aren't apparent on FR graphs. When loudspeakers use a common shared rear chamber (frequently seen in commercial line-source speakers), the delayed and inverted "echo" of the first-arrival sound is revealed in the time-domain graph ... provided the driver is clean enough to reproduce a recognizable impulse, and the response is not cluttered with box-edge diffraction.
If the two (or more) drivers share a common rear chamber, the out-of-phase radiation from one driver will go through the cone of the other driver, with a bit of additional time delay thanks to the transit time between the two drivers. The net result is degraded time response and (fairly small) ripples in the frequency response. Subjectively, the midrange is muddled and confused-sounding, as you would expect with time-delayed artifacts in the 1 mSec region.
The solution is to have an isolated chamber for each driver ... although the bass vents can be shared, since the contemplated vent system has both vent damping and a folded path from the inside of the enclosure to the outside.
This was a lesson-learned from the Ariel ... an early version with a shared chamber sounded quite bad, while measuring pretty well. The rear-wave crosstalk isn't obvious on FR measurements unless you're looking for it, while the time-domain degradation was somewhat more obvious and clear with respect to the phase of the delayed rear-wave. (One of the things to always remember with FR measurements is the phase and time information has been discarded ... and phase and time information can be useful for diagnostic purposes.)
Although it's easy to abuse time-domain information for marketing purposes, it is an excellent diagnostic for tracking down colorations that aren't apparent on FR graphs. When loudspeakers use a common shared rear chamber (frequently seen in commercial line-source speakers), the delayed and inverted "echo" of the first-arrival sound is revealed in the time-domain graph ... provided the driver is clean enough to reproduce a recognizable impulse, and the response is not cluttered with box-edge diffraction.
Last edited:
If this were true, how could a speaker radiate any sound at all? The front would just cancel with the back. Modeling a diaphragm as acoustically opaque gives very accurate results. How then could what you say be true?
That's what I think is another big advantage of the Open Baffle. The rear wave doesn't bounce back out thru the woofer cone. With light cones like the Altec, it can be noticeable and take some work to avoid. The work is worth it.
The Open Baffle just lets that rear wave bounce allover the room, first.
It does sound different.
The Open Baffle just lets that rear wave bounce allover the room, first.
It does sound different.
I'd like to see some measurements that validate this effect. I tried and was unable to do so.
I attempted to measure this effect by using a back on a cabinet and then piping the back wave away to "infinity". I was unable to measure a difference that could not be accounted for by just a lumped volume box stiffening the cone which happens with an opaque cone not a transparent one. If it were acoustically transparent then there would be no stiffening from the box at all.
To be sure there is some leakage, but 80-90% as was claimed is simply not reasonable. Based on my tests I'd guess it at less than 10% or some 10 dB down. I would have to have been that low or the results would have been obvious.
I think that I would use the term semi-transparent to describe the cone material. Different materials will have differences in the amount of transparency to perpendicular waves hitting the cone. This will have much to do with the internal damping properties of the material and the actual diaphragm thickness. As much as we should worry about the reflective waves inside an enclosure we also have to look at the noise produced by the spider, this can contribute a large portion of the sound on the back side of the diaphragm. Hasn't it been shown many times that the frame contours themselves, the flat sections of frame that are directly behind the cone can affect the sound radiation from the speaker, this is one reason to minimize and change the cross-sectional width of the frame member to help to ameliorate this phenomena. This is not only related to the loading as seen by the cone but also the reflective surface area that directs the sounds back out the front of the cone.
Pano,
Sounds stupid but just place another speaker behind the cone in question and if you know the spl output you can measure the emissivity with a microphone placed in front of the cone in question. What is the attenuation of the spl level and what is the FR that you can measure. I think you could be surprised. The rear speaker should be inside a closed box enclosure with no vent facing the back of the cone under test.
Sounds stupid but just place another speaker behind the cone in question and if you know the spl output you can measure the emissivity with a microphone placed in front of the cone in question. What is the attenuation of the spl level and what is the FR that you can measure. I think you could be surprised. The rear speaker should be inside a closed box enclosure with no vent facing the back of the cone under test.
Another good post, Lynn. I think that to get the right answer, you've got to ask the right question!
I'm feeling generous this hot afternoon and gonna lay it on a plate for you folks at diyaudio.
"Which is better in a common enclosure, 16 ohm series or 4 ohm parallel wired D'Appolito bass units?"
You see, you gotta ask the right question. It's easy. I say no more.
System7,
I have never seen a D'appolito configuration that didn't cause its own problems in the vertical polar response. This is just a given with this type of drive layout. the need to stay on the high frequency devices vertical axis has been shown so many times. I would never use this driver arrangement, the vertical comb filtering is always there if you look for it.
I have never seen a D'appolito configuration that didn't cause its own problems in the vertical polar response. This is just a given with this type of drive layout. the need to stay on the high frequency devices vertical axis has been shown so many times. I would never use this driver arrangement, the vertical comb filtering is always there if you look for it.
Mate, you engaged your brain for precisely 8 minutes there on a terrifically difficult problem.
I really don't care what you think of D'Appolito's and comb filtering. You imply the Ariel is a pile of poo, which I would doubt. Lobing is a mess with two ways and the D'Appolito addresses it.
Answer the question and learn something: "Which is better in a common enclosure, 16 ohm series or 4 ohm parallel wired D'Appolito bass units?"
It's a good question.
System7,
Leaving out the D'appolito configuration for now and the impedance differences I would always use a paralleled set of drivers over a series connection. Make your own assumptions why I say that also.
And I did not say the Ariel was a pile of poo, you assumed that. I only said I do not agree with the D'appolito configuration in respect to the polar response off axis of the center device.
Leaving out the D'appolito configuration for now and the impedance differences I would always use a paralleled set of drivers over a series connection. Make your own assumptions why I say that also.
And I did not say the Ariel was a pile of poo, you assumed that. I only said I do not agree with the D'appolito configuration in respect to the polar response off axis of the center device.
Last edited:
I am not sure but that wasn't the point was it?
I can only say that it is more than 10 dB down, it could be 100 for all I know, I could not resolve any better with my simple experiment. What experiments do you know of that quantify it any better? So based on what I know it may or may not be a minor problem. I tend to believe that it is a minor to negligible problem, certainly not one worth obsessing over. But then that's based on data, not subjective listening. I am sure you hear it and that's all that counts, right?
You would need to glue the cone in place for this to be valid, but if you did that it would be a decent test.
Are we talking about the back wave from a cone being reflected by the cabinet walls and coming back out from the cone? If so, then this is easily remedied by absorption inside the cabinet. Basically you need to absorb down to half wavelength the largest dimension.
http://www.diyaudio.com/forums/multi-way/215133-box-colourations-really-22.html#post3086482
And why would this be a problem only for the dual woofer configuration and not a single woofer?
http://www.diyaudio.com/forums/multi-way/215133-box-colourations-really-22.html#post3086482
And why would this be a problem only for the dual woofer configuration and not a single woofer?
ra7,
And how do you create a 1/4 wave absorption inside an enclosure with anything reasonable? To say that it is easy to remove any reflective waveforms that would be able to pass through the cone membrane seems a little more than realistic? If the cone can produce an FS of say 35hz how can you stop that with any shape or absorbent material. You may be able to attenuate the waveform, but to remove them enough to stop the transfer through the cone is another matter completely.
And how do you create a 1/4 wave absorption inside an enclosure with anything reasonable? To say that it is easy to remove any reflective waveforms that would be able to pass through the cone membrane seems a little more than realistic? If the cone can produce an FS of say 35hz how can you stop that with any shape or absorbent material. You may be able to attenuate the waveform, but to remove them enough to stop the transfer through the cone is another matter completely.