• Disclaimer: This Vendor's Forum is a paid-for commercial area. Unlike the rest of diyAudio, the Vendor has complete control of what may or may not be posted in this forum. If you wish to discuss technical matters outside the bounds of what is permitted by the Vendor, please use the non-commercial areas of diyAudio to do so.

Geddes Bandpass Subs and the Multi-sub approach

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Green Glue is just Melamine glue, identical to the stuff from Franklin. Liquid Nails for Subfloors claims to "never harden". When left out for years, I know that it does, but thats in open air and I'm told in a sealed situation like between dry-wall will not harden. Standard Liquid Nails is intended to harden - you have to get the Sub-Floor stuff.
 
Melamine huh, I thought I read Urethane. I relooked it up, Latex Polymer is the only ingredient listed, other than water. Wonder what I was looking at. Maybe I made up the urethane bit, or maybe its changed. Melamine glue is listed as Polyvinyl acetate, is that the same thing, or were you just saying they were similar products? Curious, because if what you say is true, that sounds like useful information. I'm sure a 5 gallon container of Franklin Melamine glue is cheaper than 5 gallons of green glue.

I don't recall where you said you get your urethane from, but I know any source I've found is expensive. Is it just the urethane for mold pouring and such? Is it a urethane foam (doesn't feel it)?
 
The polyurethane is extremely expensive - yes. Its over $100 a gallon. I use Innovative Polymers (Innovative Polymers, Inc.)

The Melamine glue just looks and acts like the Green glue, but I was not sure of what the ingredients were, just that Green Glue was far too expensive and Melamine glue was quite reasonable and works well. Thats what I would use if I do any wall structures again.

I use polyurethane foam to seal cracks and holes, but not as the CLD layer.
 
How much will a good CLD wall treatment attenuate bass ringing? For example, would it improve my situation? Here is my LF room response as a waterfall plot... The signal decays by about half (in terms of db) in something like 200-300 ms, but some of the modes keep ringing ring for another 300 ms or so.

An externally hosted image should be here but it was not working when we last tested it.
 
While I hate to speak for Dr. Geddes ... pjpoes, it was not specificaly aimed to Dr. Geddes. Thanks for the comments.

... Sounds like they made the CLD wall too stiff. It really has to be quite flimsy to work right - think or a hanging sheet of cloth - very limp. Most often that I have seen people try this technique they make it far too stiff and it kills the effect.

Yes. The stiffness seems too strong. That is why I have questioned if damping 0.4 is even better than nothing.
I asked workers to do CLD wall with half of studs (every 4' not 2'). I told them that I don't mind the cracks in joints between boards. Didn't help.
For design I have used empiric formula (found in A.Everest book) Fo= (1/2*pi)*SQRT(ro*c*c/m/d) and for center frequency about 50Hz and double 1/2'' gypsum board gives me 3'' gap. Gap was filled with 40mm glass insulation. Maybe it was not a good idea according to Toole's book. Because it softens resonance of the wall therefore the efficiency. Unfortunately I read it later.


I use 1/2" drywall and historically used Liquid nails. My polyurethene mixture would work better, but I have not used that on a wall structure. the Liquid Nails (for Subfloors) is spread about 1/4" with a flooring trowl. I attach the first layer to RC1 resilient channel, which is attached to the studs, then the second layer is just glued on the first. You have to hold the layers together with screws until set, then just remove the screws.
Resilient channel was not in stock here - very special here. Workers used 1/5" foam pads for supports.
 
How much will a good CLD wall treatment attenuate bass ringing? For example, would it improve my situation? Here is my LF room response as a waterfall plot... The signal decays by about half (in terms of db) in something like 200-300 ms, but some of the modes keep ringing ring for another 300 ms or so.

An externally hosted image should be here but it was not working when we last tested it.

That is why it is made - attenuate the bass ringing. Consequently it softens resonances, making them broader and more overlapped. There is a difference if you try to flatten out narrow 10dB peaks and 15 dB gaps spaced well appart or 5dB "wiggles" which rings 1/3 of time.
Btw. Your measurement is not spatialy averaged - there are separate reflections showing up which will be averaged out if done properly and only resonances (modes) will stay. Point of interest of CLD wall treatment is only up to 200Hz.
 
Last edited:
Resilient channel was not in stock here - very special here. Workers used 1/5" foam pads for supports.

Well that would completely defeat the whole technique. Resilient channel is not an option, it is a requirement. Its what makes the whole thing work. In Thailand they did not have resilient channel either, but we fabicated some from metal studs which seem to be available all over the world. I can see no way that the standard contractor would ever do this kind of wall right. they just don't "get it" and end up messing it all up. I have never seen a contractor do it even marginally acceptable.
 
Ok let me rant for a quick moment. Also let me add, this is directed at nobody on this forum. When Doug posted his waterfall plot, it reminded me of a pet peeve of mine. I see on so many forums talking about room dampening and modal ringing that suggest that special DSP eq's can some how reduce it. I don't see how. Also, they show waterfall plots that they suggest prove reduce modal ringing. Ok so this is my rant. My understanding of the concept of dampening is that its not an overall amplitude reduction, its that the amplitude is steadily reduced over time.

Ok so here is the problem I find with the posted graphs that they suggest is offering dampening, they show overall amplitude reduction, which of course will show up over time too, and then taught "See it dampened the modes". I'm sorry, but thats just eq, and thats all an EQ can do. Many of the LF absorbers out there do offer dampening, just not nearly as much as claimed, and in my opinion, not enough to be useful.

Dr. Geddes a room acoustics white paper that addresses this aspect of eq vs dampening would be a good one (assuming at least some of my argument is correct).
 
In Floyd Toole's book on loudspeakers and rooms he talks about the functonal equivalence of notch filters at modal peaks and bass absorption in a room. Is there any real difference in the time signature of the decay that can be achieved by these two different approaches?

I think it might be possible to achieve a faster decay at bass frequencies by having one sub out of phase with another. The WF plot I generated with my subs operating that way looked cleaner than the one I showed here where both subs are in phase with the mains. It wasn't a controlled experiment, however.
 
Matt

When one is looking at a waterfall, then this is a transient picture most times derived from some steady state measurement. The logic behind what you say is "sound" (Pun intended) in that is that no damping can affect the sound in the room before this sound reaches the damping. This much is obvious. This means that the direct field cannot be affected by room damping and hence any lowering of the sound of the direct field due to room damping must be incorrect. But at low enough frequencies it becomes impossible to detect or seperate the direct field from the reverberant field and only the steady state can be measured with any certainty. Hence it is reasonable in the modal region to see a lowering of the initial response due to the sbsorption. But if someone shows a lowering of the intial CSD above the Schroeder frequency then something is wrong, and its usually a failure of the assumption of determinism - that the sound field is stable and stationary at those frequencies. Its not. Take the same measurement several times on different days - temp and humidity - and you will get different results in the details and especially at supposed "modal peaks" (which do not exist in the statistical region). Take enough measurements at these frequencies and eventually you will get the data you want - just throw away all the other examples!

But Doug is also correct in what he says for "modal" frequencies, but alas its not true at statistcal frequencies. Thats the whole problem - the direct field and the reverberant field are different in the HF region and correct one does not necessarily or actually cannot, correct the other.
 
DCX question

I will be building my Abbeys soon and I'm thinking generally about the subs set up. The DCX manual seems to indicate that it can be set to sum the L and R stereo inputs you feed it so as to produce mono outputs. Is that correct?

(I want to feed each of my 3 subs with a mono signal for cable management reasons.)
 
I will be building my Abbeys soon and I'm thinking generally about the subs set up. The DCX manual seems to indicate that it can be set to sum the L and R stereo inputs you feed it so as to produce mono outputs. Is that correct?

(I want to feed each of my 3 subs with a mono signal for cable management reasons.)
Yes, that is correct, I use mine for 3 subs and apply crossover filters individually.
 
Sealed vs. ported vs. PR vs. bandpass

Hello. I am also interested in trying out a multisub approach. If I remember correctly, Dr. Geddes recommended using multiple subwoofers for 40-80Hz and one or two bigger subs for frequencies below 40Hz.

I wonder, what will I lose if I simply build three or four small subs and hope that they will cover the range from 20Hz to 80Hz? I think one reason to use separate subs for 20-40Hz is to ease the placement of the many subs that cover 40-80Hz, because that way they can be smaller. But if I build all my subs small, then this is not a problem.

I simply need to settle on a driver and a type of enclosure. There is a continuum from 1) dipole, 2) sealed, 3) ported, 4) passive radiator, 5) bandpass with one ported and one sealed chamber to 6) bandpass with both chambers ported. These are in the assumed order from "tight, clean and articulate" bass to less "clean and articulate" bass.

I need to know, if people rated these in a blind listening test, would dipole always come on top and would dual ported bandpass always end up last? Or are one or two of these always different from the rest? Do I gain or lose perceived quality if I go from ported to passive radiator or from ported to bandpass?

My second stated goal is to go as low as possible. Do any of these enclosure types allow that without eq? I know that the sealed box often matches the room gain. But I also know that many would rather tune a "house curve" than listen to a sealed box plus room gain resulting in a flat 20-80Hz response. Instead, a flat 20-80Hz anechoic response plus room gain often results in a kind of "house curve".

I also assume that driver harmonic distortion is not a problem per se, but power compression is. And to get less of that, I should rather use pro drivers than anything offered by Seas? To state my problem another way is, if it is possible to get good bass from multiple 1 cf cubes with any enclosure type at all? If my starting point is four Seas L26ROY with one passive radiator for each, where would you go and what would you do if you were me?
 
I would elliminate the dipole as its efficiency is not sufficient in the frequency range you are talking about. Then, as far as I am concerned, the rest are all about the same. I don't buy your "ranking" of subjective appeal. I use only bandpass subs and I find that they work just fine. Dual ported is a tough enclosure to get right. Easiest is closed box, but requires a big box to get down to 20-30 Hz. Passive radiators are great, if you can get a good radiator at a "reasonable" price (not what I have seen).

I would say that 1 cf cubes is not going to work very well as that is just too small. Bass takes power and size and nothing that I know of can get arround this simple fact.

My subs go much higher than 80 Hz. One goes all the way up to about 130 Hz. The big ULF only goes to about 80 Hz.
 
100 liters seems enough for push-pull to get decent 30Hz at 91dB. 200-300 liters for higher sensitivity like ULF-18. There's another small solution - tapered transmission line - for example 1:10. Usually there's more subjective bass extension than simulated response suggest (room gain).
 

Attachments

  • Sub.gif
    Sub.gif
    52.9 KB · Views: 485
  • Transmissionline and U-frame.gif
    Transmissionline and U-frame.gif
    12.3 KB · Views: 475
The Melamine glue just looks and acts like the Green glue, but I was not sure of what the ingredients were, just that Green Glue was far too expensive and Melamine glue was quite reasonable and works well. Thats what I would use if I do any wall structures again.

I use polyurethane foam to seal cracks and holes, but not as the CLD layer.

This CLD detour in the thread has caught my interest, so I'm making some simple tests and observations to compare Melamine Glue to Green Glue. I hope to compare some other adhesives/sealants later as well.

Green Glue indeed has done a great job of documenting its performance. I've no doubt that melamine glue will also "work well". What I want to figure out is 1) is GG essentially the same stuff as melamine glue?--does melamine glue damp as well as GG? And 2) if not, I'd like to determine some point of diminishing returns. I'm pretty sure that point is well below the present cost for GG, and somewhere above the cost for cheap latex caulk.

Context: I'm simply observing the glues - No formula discovery or CLD performance conclusions at this stage. I'm no chemist. I do have some experience in room acoustics. In the end, I'm shooting for best bang for the buck backed by at least a little bit of science. 'Just a guy documenting practical differences or similarities between these glues.

So far I've simply laid down similar "lines" of each glue, and also adhered some materials together with the two glue types.
The "lines" I've laid were to obseve such things as viscosity, skin time, cure time, etc. I put the lines on a piece of flexible vinyl, so I can observe flexibility & elasticity after cure, etc. Each line was about 1/4" wide/round, and 4 inches long.

Observations over a 28 hour period: The glues are not the same. They smell somewhat similar, but are distinguishable from one another. 'Hard to put useful layman's words on this aspect.
Viscosity: The GG is more viscous than the Melamine. The MM glue consistency is similar to white/yellow wood glue, while GG has the "gooeyness" of a runny sillicone adhesive. Within the first few hours, the MM line spread and flattened to about 150% width. The GG didn't change much, maybe 110%.

Skin: The melamine glue skinned nicely in a couple hours, and was smooth & dry to the touch after about 4 hours; The GG is still tacky after 28 hours :eek:as I type this.

I've also adhered some pieces of vinyl together with the glues to observe shear behavior (difficult to observe using small pieces of drywall-- too thick relative to the sample size; can't proportionately flex like a full sheet of drywall will).

I'll have to wait until the GG is cured before offering any more observations on cure time, shear, bond etc.

-- Mark
 
I'm not sure Dr. Geddes was saying they were the same thing. I had thought that too at first, but I think he was just suggesting it to be as effective as GG. I think to really prove or disprove that would require tests of STC and damping properties. Damping could arguably be tested with an accelerometer I would think. STC would not be feasible on a small scale to the same level that a lab would test it. However I wouldn't be surprised if you could approximate the results on a small scale. Make a sound proof box, that is as sealed and "soundproof" as possible on 5 of it's 6 sides. Leave it open on the 6th side. I would want the box to be in the range of 4'x4'x4' or so. suspend the test microphone inside the box, a good 2' from the front of the chamber. Test the box open, with a sound source, and see what the response is. This is the baseline, you will zero out to this. Then connect a single 1/2" piece of drywall, with screws to the front, and test again. This is your "business as usual" type test, the baseline that any old home would have. Then test each of your test solutions, and compare the sound transmission loss seen with each.

It's not a perfect test, it won't equal what Green Glue, for instance has done, but it would at least give you something to consider. I think doing it right, like that, would cost you more than just buying some Green Glue, so if you are testing to see if Melamine Glue is cheaper, it would be a pretty silly thing to do. On the other hand, if you want to try a product that is cheaper, that a PhD Acoustical Physicist feels confident would work well enough, then use the Melamine glue.
 
I never said that they were the same thing. It doesn't really matter does it? It's function that matters and I suspect that they do work the same or similar.

It's like the screen in my HT. It's a bed sheet. Is it "identical" to "micro-perf" (or whatever the new buzz word in screens is?) No, of course not, its a cotton fabric not a vinyle screen material! Does it work the same? No, not actually, it works better. Getting the best function at the lowest cost is what its about.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.