The claim was that lowering the vibrations of the horn walls lowers the HD. That is impossible. Sure the horn loading will reduce the cone excursion and that will have some effect on the HD, but that was not the claim that I objected to. I already said why I wouldn't show any supporting measurements. I offered my opinion, if that is not good enough then so be it.
I explained how earlier and you agreed with it! It is one of the distortion mechanisms in coaxials.
Gedlee,
If adding CLD to horn walls makes the measured HD go down substantially, and you say it is impossible to reduce HD by reducing horn wall vibrations, can you please explain what the mechanism causing this observable fact is? Really, I am trying to understand this and willing to hear what the source of this anomaly as you are describing it is, if it is not the vibration reduction. I think we both agree that a horn doesn't change the intrinsic HD of a driver. Let's take the tone down a few notches and start over. I am listening.
If adding CLD to horn walls makes the measured HD go down substantially, and you say it is impossible to reduce HD by reducing horn wall vibrations, can you please explain what the mechanism causing this observable fact is? Really, I am trying to understand this and willing to hear what the source of this anomaly as you are describing it is, if it is not the vibration reduction. I think we both agree that a horn doesn't change the intrinsic HD of a driver. Let's take the tone down a few notches and start over. I am listening.
You mean your post #345 and Earl's reply #347? I hadn't realised you were referring to coaxials! (presumably the Tannoy type where the large cone forms an extension to the small waveguide of the tweeter, and the HD occurs in the band of crossover frequencies where both drivers radiate the same frequencies?) I was thinking you referred to a super-soft horn wall like xrd's undamped 5mm foam, 'pumping' at the same frequency as the drive unit, and was trying to visualise how applying the CLD layers might stiffen the wall and reduce HD. I couldn't quite think it through and gave up. 😱
I was referring to linear acoustics which applies to everything. The coaxial thing is interesting because I would not have expected this to be the dominant form of distortion created by the movement of the midrange cone and the tweeter output but Andrew Jones (post #10 here) said that it was. And he should know.
Perhaps it's best to simply restate the issue- I think we may be stuck on language.
"CLD substantially reduced the wideband resonances of the walls of my horn, which showed up as distortion in my measurement."
They're not necessarily harmonic distortion as we know it, but the way they manifest is similar enough to be measured as such.
I can say that foamcore, used with CLD, is fantastic. I use layered foamcore with soft adhesives as a turntable platform, and the large front panels of my surrounds in the bedroom are two foamcore layers, fastened around the edges, with some soft gasketing distributed between. Not as good as a full sheet elastomeric bond, I suppose (though there's some pre-tension by my method) but the resonance on an 20"x28" or so panel is not perceptible, and it was easy to achieve.
"CLD substantially reduced the wideband resonances of the walls of my horn, which showed up as distortion in my measurement."
They're not necessarily harmonic distortion as we know it, but the way they manifest is similar enough to be measured as such.
I can say that foamcore, used with CLD, is fantastic. I use layered foamcore with soft adhesives as a turntable platform, and the large front panels of my surrounds in the bedroom are two foamcore layers, fastened around the edges, with some soft gasketing distributed between. Not as good as a full sheet elastomeric bond, I suppose (though there's some pre-tension by my method) but the resonance on an 20"x28" or so panel is not perceptible, and it was easy to achieve.
Measurement error or interpretation error.
And if you believe otherwise, by what mechanism can a vibrating horn create harmonics? We went through this already and there isn't one unless the excursions are enormous.
I think that Andrew is talking about a different thing (modulation distortion does not create harmonics of the fundamental,) and the cone in a coaxial does have significant motion. But if Andrew is saying that most of the distortion in a coax is due to the cone motion modulation then I would not accept that either. The coaxes that I have measured just do not have more distortion that drivers on rigid waveguides - far worse FR yes, but not THD.
Last edited:
I can accept this possibility and even mentioned it earlier, but the contention seems to remain, that it was harmonics that were reduce and not just broadband chaotic noise from a flimsy construction.
The best cabinet material ????
What are the chances that something approaching a general consensus on what the best cabinet material is, will ever be reached ?
Audiophile conversation:
Q) What's the best box ?
A) No box at all !
Q) okay, what do we do for bass response?
A) Use a box.
Q) What if, we could build a system to where there was no excitement of the cabinet walls to begin with ?
A) okay, let's do that !
What are the chances that something approaching a general consensus on what the best cabinet material is, will ever be reached ?
Audiophile conversation:
Q) What's the best box ?
A) No box at all !
Q) okay, what do we do for bass response?
A) Use a box.
Q) What if, we could build a system to where there was no excitement of the cabinet walls to begin with ?
A) okay, let's do that !
I agreed that if the excursions were excessive that such a thing could happen in theory, but I doubted that the excursions in a horn could ever get that high. Also there is the fact that breathing types of nonlinearities like this tend to create sub-harmonics and we didn't see that in the plots.
An interesting joke, but I think that there are substantial ways to alleviate the enclosure vibrations to the point where they are not audible. That, to me, is well worth discussing and would not lead to a null result. I have spent decades looking at this problem and it certainly has not resulted in a null result.
Last edited:
What I have gathered so far:
1) Absorption of internal waves - Adequate stuffing
Question - What are the better absorption materials? (Fiberglass insulation?) How much is enough?
2) Reducing extraneous cabinet vibration - Bracing the driver by the front frame as well as the magnet. Cross bracing the cabinet to ensure stiffness.
Question - What materials provide optimum stiffness? What kind of bracing works best?
3) Damping material resonance - Adding CLD layers to cabinet walls and internal braces.
Question - What is the best material for CLD? How best to apply CLD? Is it about using a CLD that acts also as an adhesive? Does nailing or screwing CLD layers decrease effectiveness? When using CLD on internal braces you want those braces real tight to the cabinet to do their job providing support, but are we also supposed to squeeze some damping compound in between?
In a quick nutshell, that's where I stand on the topic.
Back to studying for exams
1) Absorption of internal waves - Adequate stuffing
Question - What are the better absorption materials? (Fiberglass insulation?) How much is enough?
2) Reducing extraneous cabinet vibration - Bracing the driver by the front frame as well as the magnet. Cross bracing the cabinet to ensure stiffness.
Question - What materials provide optimum stiffness? What kind of bracing works best?
3) Damping material resonance - Adding CLD layers to cabinet walls and internal braces.
Question - What is the best material for CLD? How best to apply CLD? Is it about using a CLD that acts also as an adhesive? Does nailing or screwing CLD layers decrease effectiveness? When using CLD on internal braces you want those braces real tight to the cabinet to do their job providing support, but are we also supposed to squeeze some damping compound in between?
In a quick nutshell, that's where I stand on the topic.
Back to studying for exams
Last edited:
Earl, you once said that strongly coupling the front and rear panels was a worthwhile endeavour and I've been trying to reconcile the mechanism.
Assuming that each was optimally CLD'ed and loosely coupled at the edges to the side panels, where mass thereby doubles and stiffness goes up by a factor of 8. Fr would *2. Motion potential might reduce, damping potential would, I don't know go both ways?.. and radiation would become dipole, cancelling at the low end dependent on the cabinet dimensions.
Seeing as the DIYer can take a cabinet in an number of ways, can this all be pinned down to something? Say frequency dependent or not? Just practical use of available stiffness or a combination of the above?
Assuming that each was optimally CLD'ed and loosely coupled at the edges to the side panels, where mass thereby doubles and stiffness goes up by a factor of 8. Fr would *2. Motion potential might reduce, damping potential would, I don't know go both ways?.. and radiation would become dipole, cancelling at the low end dependent on the cabinet dimensions.
Seeing as the DIYer can take a cabinet in an number of ways, can this all be pinned down to something? Say frequency dependent or not? Just practical use of available stiffness or a combination of the above?
1) I did an AES paper on this topic back ion the early 80's. We found that a loose weave made from old socks and stuff used in automotive - we further shredded it - worked best. Next was loose fiberglass and/or pillow stuffing. "fill loosely" is the key-word.
2) I am not sure that I want to connect the magnet to anything in the enclosure, but a very firm front baffle is necessary. I use a 2" CLD panel that is very stiff and damped. Cross bracing is essential - should be damped as well. I use polyurethane which has a stiffness/weight ratio that is very high and it is well damped for the panels and the braces, although I have used Oak braces before I started damping the braces themselves.
3)CLD requires two panel of near equal stiffness. The inner layer is a well damped adhesive between the two outer sheets. The adhesive material is critical and usually proprietary. I use a filled polyurethane rubber. You must not "short-out" the adhesive layer or the effectiveness drops substantially. In cross-bracing you have to find what works for you. I use a combination of rigid connection and floating - it's based on my knowledge of machinery isolation in noise control.
2) I am not sure that I want to connect the magnet to anything in the enclosure, but a very firm front baffle is necessary. I use a 2" CLD panel that is very stiff and damped. Cross bracing is essential - should be damped as well. I use polyurethane which has a stiffness/weight ratio that is very high and it is well damped for the panels and the braces, although I have used Oak braces before I started damping the braces themselves.
3)CLD requires two panel of near equal stiffness. The inner layer is a well damped adhesive between the two outer sheets. The adhesive material is critical and usually proprietary. I use a filled polyurethane rubber. You must not "short-out" the adhesive layer or the effectiveness drops substantially. In cross-bracing you have to find what works for you. I use a combination of rigid connection and floating - it's based on my knowledge of machinery isolation in noise control.
My principle has been to create a point in the center of the cabinet that attaches as many sides as possible to this one point. The vectoring of forces will want this point to remain stationary. By forcing all panel motion through a damper to this central point causes all panels to be well damped. This is a form of what is called skyhook damping. A damper to some fixed point in space rather than some point also in motion on the body. Skyhook damping is always more effective.
Everything else derives from this principle, although in detail the device just might look rather complex. It was complex to make at first. It took several stages of assembly. Then I got smarter and now I just cast a bunch of solid pieces into a single mold with the damping compound and wha-la! complex damper piece in just a few minutes. That plus months of trial and error getting it to work!!
Thanks Earl 🙂
1) When using a thick front baffle, would you suspect running into a problem of reflections back into the driver basket ventilation holes?
2) Would you suggest lining all the internal walls with fiberglass and then a loose fill within. Or just a loose fill? How does fiberglass alter internal volume?
3) How soon do CLD layers hit a point of diminishing returns?
Last edited:
Mr geddes
what's "wha-la" or do you mean "voila" the french term for "there it is"
could you elaborate on why you would not connect the magnet to anything in the cabinet?
what's "wha-la" or do you mean "voila" the french term for "there it is"
could you elaborate on why you would not connect the magnet to anything in the cabinet?
Last edited:
To build on that, what about this idea I have thought of. Some speakers have magnets that can fit into pipes, abs, pvc, etc. What if a pipe connected to the back of the magnet made it's way to the center of the speaker, where a connector branched it off in the 5 other directions of the cabinet. And say you filled those pipes with sand, or some lighter filler. And attached it to each wall with a CLD compound.
Yes, no?
A chosen ground point. Cabinet design is no different to designing an amplifier as I've been seeing it.
Regarding choosing the best point for this, I could see this being simple for a spherical enclosure or one with that level of symmetry. Some cabinets might theoretically call for a rigid elongated ground point which may be fraught with modes where it cannot be held still in all ways, instead practically calling for a single point where the compromises are manifested in the attachment points/methods..which brings some questions, especially in a non box like enclosure.
Should a ground point be located symmetrically. Can it be located outside the box if the focus lands there? If it doesn't? Can it be made from an array of points if the risks are considered?