Damping makes any material sound better but ply wood material depending on thickness and type is usually over damped making it heavier hence less effecient. Also most people prefer the sound of "WOOD TONES" as its a natural material made from natural raw compounds. My ears also prefer wood tones. Wood is also harder/stiffer then EPS/XPS so it is a bit more accurate. Its a combination of being very light, hard/stiff yet able to flex/bend a little for a ideal sounding panel. There are compromises to each material you just have to pick the one that suits your taste.
Ideal materials are EPS/XPS, Honey Comb composites and certain types of wood.
I would have assumed yawl read that thread before. Zygadr was a pioneer for starting DML topics. He even started one on here under the name Ziggy.
PIEZO NXT type panel
Zygadr's panel of choice was high density (high compression strength) EPS, the same material I use.
So here's the deal with DML's. NXT was the first company to try to purely define the DML and its particular needs mathematically for design purposes. Unfortunately other than the info you can get from their patents I dare say the mathematics of it all is beyond the scope of most of us. But...
There's really no need for the math. The beauty of DML's is that they can very inexpensively and quickly have proper dimensions, exciter placements tested out without knowing any of the math or numbers.
Just take a panel, lay it across 4 chairs at the very tips of the panel corners. Then connect the exciter to a source and play sound through it. Now simply run the exciter over the top of the panel and listen carefully for the sound levels. You will soon find the resonant modal points (I.e., where the sound is loudest and sounds best). There you go, mark that point and glue your exciter there. While your at it also mark down the other points that are very resonant. Keep those in mind as potential dampening/clamping points or points for multiple exciters if your going that route. Easy peasy and no need to remember ratios and such...and If you decide you want to market your invention, there's no issues with needing an NXT license because you didn't use their info...BTW, that's pretty much how Podium got away without needing an NXT license. They were able to show that they didn't use any NXT derived numbers...they did it simply by hand and ear
BTW, from what some have said and I tend to agree, the panel size ratios to me seem to not make such a big deal really. The end point is to have as many resonant modal points as possible on a panel. That should give the smoothest and widest FR.
There's really no need for the math. The beauty of DML's is that they can very inexpensively and quickly have proper dimensions, exciter placements tested out without knowing any of the math or numbers.
Just take a panel, lay it across 4 chairs at the very tips of the panel corners. Then connect the exciter to a source and play sound through it. Now simply run the exciter over the top of the panel and listen carefully for the sound levels. You will soon find the resonant modal points (I.e., where the sound is loudest and sounds best). There you go, mark that point and glue your exciter there. While your at it also mark down the other points that are very resonant. Keep those in mind as potential dampening/clamping points or points for multiple exciters if your going that route. Easy peasy and no need to remember ratios and such...and If you decide you want to market your invention, there's no issues with needing an NXT license because you didn't use their info...BTW, that's pretty much how Podium got away without needing an NXT license. They were able to show that they didn't use any NXT derived numbers...they did it simply by hand and ear
BTW, from what some have said and I tend to agree, the panel size ratios to me seem to not make such a big deal really. The end point is to have as many resonant modal points as possible on a panel. That should give the smoothest and widest FR.
The non-cancellation effect for the DML's sound field has another very interesting and useful property. The sound pressure level decrease as distance increases away from the speaker is about half that of a piston/cone directional speaker. I.e., at any seating position the sound volume will be louder than you might expect. As well, because of the increased dispersion of the DML sound field the room effects as well seem to be decreased (not eliminated but definitely lessened) so room acoustics don't seem to be so harsh and placement needent be so critical (as long as its well away from wall). As I said there seems to be no sweet spot. You can sit at either end of your couch and be happy. You don't need to sit at one optimal location with your head in a vice. Some find this unusual and not pleasing, but that's just more of being habitually used to hearing beaming all the time. It takes time to get used to the DML sound field. But I think its very well worth it. I'm still just experimenting with it all and the more I hear it the more I hate my old cone speakers. There's just something special to being literally immersed in the sound field...but then I've been playing with this stuff since the early 80's and the experiments that Polk did with their original SRS type of speaker...Sound field manipulation make all the difference to me.
As to OB's lots here feel that DML's combined with OB Sub's is the only way to go...I'll reserve judgement until I can make/hear a DML sub
Last edited:
There's a lot to read when it comes to DML's, I'm just getting started! A lot is great stuff and a lot is BS and a lot is difficult to understand so sometimes you have to read through several times to know what to ignore and what to focus on!
Eric
Yes, the first TI video the guy rattles off a list of about 24 different materials they tested (that included carbon fiber and aramid materials in plain and composite forms) and yet the best material they found was:
1. XPS overall the best overall sound and frequency range
2. End Grain Balsa had better/full lower mid-range but XPS had better highs and overall sound
3. Honeycomb cardboard, not corrugated!
4. Resonant spruce was also found to be quite good but impractically expensive.
What I found interesting was none of the high tech materials and composites were found to be as good as plain old XPS. Yet most people in these forums still lay claim to composites as being somehow superior.
Then in the second video he found the ceiling tile which had good bass and almost as good highs as XPS, but maybe most importantly was also the cheapest ($5 for a 2'x4' sheet). Its only problem was its heavy and thus SPL's were lower and needed a more powerful exciter(s). They also decided that the end grain balsa, though good was expensive and hard to find as well. So their final speakers were a mixture of the XPS and Ceiling Tile.
Now this is where I take exception to their findings. They only test specific size of material and only used 2 different exciters for their testing. That to me is not a resounding qualification to say one panel is better than another. There are just too many other variables involved. In the end I think they did a great job in what they did, but they should have been a good bit more open about all their testing methods. None the less, they probably did more to excite the DML interest in DIYer's since Ziggy started this all back in 2008...RIP Ziggy, your legacy lives on...
This is what TI tested in their first video:
steel
glass
Aluminum
Brass
Poly-carbonate
Acrylic
Pine
oak
resonant spruce
fiber glass
gypsum (dry wall)
end grain balsa
corrugated cardboard
coraflute (corrugated plastic)
Nomex/carbon fiber honey comb composite
Carbon Fiber/urethane composite
Carbon fiber
Polyethylene
Polyurethane
Polystyrene XPS and EPS
honeycomb cardboard
and they added ceiling tile in the 2nd video.
Someone was asking about honeycomb cardboard, this is relatively cheap for a good sized sheet and TI recomemded it as almost as good as XPS: Save On Discount Utrecht Honeycomb Board Sheets & More Corrugated Cardboard at Utrecht
steel
glass
Aluminum
Brass
Poly-carbonate
Acrylic
Pine
oak
resonant spruce
fiber glass
gypsum (dry wall)
end grain balsa
corrugated cardboard
coraflute (corrugated plastic)
Nomex/carbon fiber honey comb composite
Carbon Fiber/urethane composite
Carbon fiber
Polyethylene
Polyurethane
Polystyrene XPS and EPS
honeycomb cardboard
and they added ceiling tile in the 2nd video.
Someone was asking about honeycomb cardboard, this is relatively cheap for a good sized sheet and TI recomemded it as almost as good as XPS: Save On Discount Utrecht Honeycomb Board Sheets & More Corrugated Cardboard at Utrecht
I for one would take Tech ingredients vids like a grain of salt. He says EPS is "LOUSY" but praises XPS. What he doesn't realize is that there are different densities of EPS as well as XPS. The XPS commonly found in home depot or lowes are of the higher density grades while most common found EPS at home depot and lowes are of the lower density grades. You cant honestly compare the two if one has a lower density grade. Ziggy loved it when he found the higher density grades of EPS.
What T.I. also doesn't understand is that different thickness is materials will also produce a significant change in sound. For example like I stated before the panel material with the most bass is the aluminium panel for the mere fact that it is the thinnest of all the panels. He over looks this fact by saying the aluminum panel is the worst sounding panel because it rings but doesn't realize it has the best bass because of its the thinnest panel. If he would have tried 1/2 or 1/4inch XPS he would of realized the difference instead of using the 1inch thick XPS. In his first video he says one can use weights and in his second video he disregards the weights. LOL
Paper honey comb composites are used in the Podium speakers. They are coated with phenolic resin to make them stiffer/higher compression strength most likely higher strength then EPS or XPS hence more accurate. If not mistaken the honey comb is 5mm thick so it can flex/bend a little. Thicker materials less bend while thinner materials have greater flex/bend. Down fall is it cost more and the process of knowing how to coat it with the right amount of resin and or how to apply mylar skin is a very hard task. There is a DML topic on here of the use Honeycomb composite and what he went through to build it. For me its too much cost and work.
Carbon fiber + nomex honeycomb sandwich DML panel construction project
If done right Honeycomb is the best but expensive which is the main reason most wont attempt to experiment on that type of material. Next in line is high grades density/compression of EPS and XPS which have all around good sound. IMO EPS is better sounding then XPS when comparing equal density grades.
Then there are certain woods but they are usually more expensive then EPS/XPS. Main down fall with most woods or heavier material is LESS EFFICEINCY. Also I feel that the heaviness of wood reduces certain frequencies like the mids. You might get better high frequency from wood at the cost of reduced mids. But at the end of the day its all personal preference.
What T.I. also doesn't understand is that different thickness is materials will also produce a significant change in sound. For example like I stated before the panel material with the most bass is the aluminium panel for the mere fact that it is the thinnest of all the panels. He over looks this fact by saying the aluminum panel is the worst sounding panel because it rings but doesn't realize it has the best bass because of its the thinnest panel. If he would have tried 1/2 or 1/4inch XPS he would of realized the difference instead of using the 1inch thick XPS. In his first video he says one can use weights and in his second video he disregards the weights. LOL
Paper honey comb composites are used in the Podium speakers. They are coated with phenolic resin to make them stiffer/higher compression strength most likely higher strength then EPS or XPS hence more accurate. If not mistaken the honey comb is 5mm thick so it can flex/bend a little. Thicker materials less bend while thinner materials have greater flex/bend. Down fall is it cost more and the process of knowing how to coat it with the right amount of resin and or how to apply mylar skin is a very hard task. There is a DML topic on here of the use Honeycomb composite and what he went through to build it. For me its too much cost and work.
Carbon fiber + nomex honeycomb sandwich DML panel construction project
If done right Honeycomb is the best but expensive which is the main reason most wont attempt to experiment on that type of material. Next in line is high grades density/compression of EPS and XPS which have all around good sound. IMO EPS is better sounding then XPS when comparing equal density grades.
Then there are certain woods but they are usually more expensive then EPS/XPS. Main down fall with most woods or heavier material is LESS EFFICEINCY. Also I feel that the heaviness of wood reduces certain frequencies like the mids. You might get better high frequency from wood at the cost of reduced mids. But at the end of the day its all personal preference.
I'll agree with you on all those points but one...the use of weights. They didnt disregard the weights, they simply found another compromise they felt was better. With the weights they were able to reduce the resonances by at best 30% but ended up creating other resonance points in the process. But what they found in the second video is that by using 2 sheets of the same material and simply adjusting their shape and exciter position the peaks and dips in the FR would be sympathetic and tend to cancel themselves out and they were able to achieve better than 50% reduction in resonances and a smoother overall FR. So they dropped the weights because they felt they found a better method...such is the way of compromises...but at what costs? Their new compromises mean they had to build double the panels at double the costs and taking up significantly more space such that all those panels wouldn't fit in an average size room nor would anyone want to look at them...I would have gone back and tried other dampening methods...felt, or coatings like glue/rosin/epoxy/shellac/etc...maybe do a little panel carving. etc...or hell, maybe just trying a framing method. But hey, thats just me talking out loud...they did the work I didn't, so I cut them some slack. At any rate, they gave us a good start, but not really anything that wasn't already available in the decade long threads we already have. But it was instructive and entertaining and brought more people here...so its all good
I'll agree with you on all those points but one...the use of weights. They didnt disregard the weights, they simply found another compromise they felt was better. With the weights they were able to reduce the resonances by at best 30% but ended up creating other resonance points in the process. But what they found in the second video is that by using 2 sheets of the same material and simply adjusting their shape and exciter position the peaks and dips in the FR would be sympathetic and tend to cancel themselves out and they were able to achieve better than 50% reduction in resonances and a smoother overall FR. So they dropped the weights because they felt they found a better method...such is the way of compromises...but at what costs? Their new compromises mean they had to build double the panels at double the costs and taking up significantly more space such that all those panels wouldn't fit in an average size room nor would anyone want to look at them...I would have gone back and tried other dampening methods...felt, or coatings like glue/rosin/epoxy/shellac/etc...maybe do a little panel carving. etc...or hell, maybe just trying a framing method. But hey, thats just me talking out loud...they did the work I didn't, so I cut them some slack. At any rate, they gave us a good start, but not really anything that wasn't already available in the decade long threads we already have. But it was instructive and entertaining and brought more people here...so its all good
That's because "HE DOESNT KNOW HOW TO USE THE WEIGHTS PROPERLY".
Secondly what gave him the idea to use metal washers? Anything you attach to the panel will resonate and make sound. Metal will make a harsh sound. I've tried it and metal is not ideal to be used as weights.
Discarding weights to double the panels which not only increases cost but is impractical in most average sized homes because almost no one will have the space for 4+4 huge panels is ridiculous. lol It would of been a lot more practical to stick to the weights even if it had more resonance.
It was entertaining but a lot of faults that is why I said take it like a grain of salt because for the mere fact he is only a BEGINNER himself into DML technology. I don't care what physics degree he has since he is just starting out he doesn't know much. I even wonder what gave him the idea to hang the panels from some "STRING"? if I remember correctly he said DML idea came from Rich M. at P.E. tech talk project gallery yet Rich never hung his panels by a string. hmmm lol Even his exciter placement probably came from P.E. guide to exciters. He even cut two holes in the panels to hang the strings from using some sort of plastic bushings. I guess he feels cutting holes in panels just to hang strings will have no effect on the sound. lol Anytime you attach, modify or change anything about the panel its sound will change. In the comment section of youtube he tells people that if you paint your panels with spray paint it wont change the sound that is FALSE. Paint one panel and don't paint the other and do a A/B comparison and there will be a slight difference in sound. If you paint dots over one panel and not the other there will be a difference in sound. Just putting a piece of tape on one panel will change the sound. The reason why I tell people to cut the legs off of the exciters that come with the legs is because anything other then the exciters voice coil touching the panel will change the sound. OH my bad just rambling. LOL
Good sounding exciters. Not in any particular order.
1.DAEX30HESF-4 (high power handling and excursion)
2. DAEX25FHE-4 (best bang for the buck all around exciter)
3.Tectonic TEAX19C01-8 (smaller exciter with low power handling but good sounding)
4. DAEX25SHF-4 (highest spl of all exciters but expensive)
5. DAEX32EP-4 (high power handling all around good sound)
6. DAEX19CT-4 (really good high frequency)
7. TEAX25C10-8/SP (all around good sounding exciter but on the expensive side)
8. DAEX13CT-8 (good sounding exciter with good highs but very small so spl and power handling is very low)
9. DAEX32U-4 Ultra (Good sounding exciter but there have been some with reliability issues)
If anyone feels I left out a good sounding exciter please let me know.
General information of exciters from my own experience is that the larger more powerful exciters have more bass output while smaller exciters have less bass output but better high frequencies.
13mm and under exciters seem to lack bass and midrange but shine in the higher frequencies. 19mm exciters seem to have the best balance of bass, mids and highs. 25mm exciters seem to have good bass, mids and highs but can be bit shouty in the mids. 30mm and up exciters have the best bass, some shouty mids but lack in the higher frequencies.
1.DAEX30HESF-4 (high power handling and excursion)
2. DAEX25FHE-4 (best bang for the buck all around exciter)
3.Tectonic TEAX19C01-8 (smaller exciter with low power handling but good sounding)
4. DAEX25SHF-4 (highest spl of all exciters but expensive)
5. DAEX32EP-4 (high power handling all around good sound)
6. DAEX19CT-4 (really good high frequency)
7. TEAX25C10-8/SP (all around good sounding exciter but on the expensive side)
8. DAEX13CT-8 (good sounding exciter with good highs but very small so spl and power handling is very low)
9. DAEX32U-4 Ultra (Good sounding exciter but there have been some with reliability issues)
If anyone feels I left out a good sounding exciter please let me know.
General information of exciters from my own experience is that the larger more powerful exciters have more bass output while smaller exciters have less bass output but better high frequencies.
13mm and under exciters seem to lack bass and midrange but shine in the higher frequencies. 19mm exciters seem to have the best balance of bass, mids and highs. 25mm exciters seem to have good bass, mids and highs but can be bit shouty in the mids. 30mm and up exciters have the best bass, some shouty mids but lack in the higher frequencies.
He said he tried polystyrene both the expanded and the extruded and the expanded is lousy and it was hard to get the exciters to stick to them.
I find that hard to believe as I have bought the same EPS from home depot and never had a problem with the exciters sticking to them. Now if he is trying to "REMOVE" the exciters off the EPS then yes the beads will start to fall off due to the sticky tap adhering to it.
Last edited:
Has anyone on here looked into printing a small panel and testing them in various different designs. Materials like polywood have a density of 0.85g/cm3 so you can find materials which should keep weight down.
You could change your front panel thickness, say 0.4, 0.6, 0.8 and 1mm
Then use different reinforcement structures such as rectilinear, honeycomb, diamond, triangular. Maybe try some custom layouts too. Would help build up what works well and what doesn't.
Not sure if any of you have seen my work on my own full range speaker driver from scratch. Would any off you be interested in exciter? What criteria would you look for?
Paul
You could change your front panel thickness, say 0.4, 0.6, 0.8 and 1mm
Then use different reinforcement structures such as rectilinear, honeycomb, diamond, triangular. Maybe try some custom layouts too. Would help build up what works well and what doesn't.
Not sure if any of you have seen my work on my own full range speaker driver from scratch. Would any off you be interested in exciter? What criteria would you look for?
Paul
Has anybody tried MDF sealer as opposed to diluted PVA or shellac sealer for sealing XPS / EPS board? I only ask as I have both PVA and MDF sealer to hand, but no shellac based sealer.
My exciters and ultra high density EPS should arrive this week, so some more materials are needed.
I'm still trying to get my head around the best way of affixing the panels. They will sitting on top of OB bass bins, so will probably be frame based with a spine/spline to help secure the exciter. Planned XO is probably in the 150-200hz region using a miniDSP with LR24 XO plus EQ as needed. Panel size will be around 40x54cm (silver ratio or thereabouts).
My exciters and ultra high density EPS should arrive this week, so some more materials are needed.
I'm still trying to get my head around the best way of affixing the panels. They will sitting on top of OB bass bins, so will probably be frame based with a spine/spline to help secure the exciter. Planned XO is probably in the 150-200hz region using a miniDSP with LR24 XO plus EQ as needed. Panel size will be around 40x54cm (silver ratio or thereabouts).
EPS is very sensitive to corrosive substances which is why PVC glue is recommended for not being corrosive.
Frames should be wider then your panels thickness. 1inch thick panel should be in a 2-4inch width frame.
Do not connect panel material directly to frame. It needs a damping material between them like frost king foam.
Last edited:
Okay, let's get down to physics here. Who really understands the physics?
If I want to maximize the sound radiated from my panels, with a given exciter, what panel material properties (or combination of properties) do I want to maximize (minimize, or optimize?). Yes, I know, stiff and light. But I want a more specific answer. What is the mathematical description of the property I want to optimize?
I often see the quantity B/m mentioned, or (B/m)^0.5, where B is Bending Stiffness (per unit width) and m is panel mass per unit area.
As I understand it, the speed of a transverse wave is proportional to (B/m)^0.5, and sound radiation is maximized when the transverse wave speed approaches the speed of sound. Hence, increasing B/m increases wave speed, and hence increases sound radiation.
But I also often see the quantity B/m^3 mentioned. Several patents refer to minimum values of B/m^3 for good performance. What is the basis of the B/m^3 quantity (instead of just B/m)? And is it a better predictor of performance than B/m?
Let me be more specific. For the purposes of this question, let's say panel size is not limited. That is,for the purposes of this discussion, that I'm willing to make the panel as large as it needs to be to have a sufficient modal density down to a particular frequency (say, 100Hz). Assume also, that I have a single exciter (or maybe two) of the sort commonly available (like DMLBES's list above). Assume also for the sake of this question that I know how to support the panel and locate the exciter in reasonably good locations.
Given all that, what is the mathematical expression of the material property I need to maximize, minimize, or otherwise optimize, in order to maximize output?
Is it B/m? B/m^3? Other?
Thanks,
Eric
If I want to maximize the sound radiated from my panels, with a given exciter, what panel material properties (or combination of properties) do I want to maximize (minimize, or optimize?). Yes, I know, stiff and light. But I want a more specific answer. What is the mathematical description of the property I want to optimize?
I often see the quantity B/m mentioned, or (B/m)^0.5, where B is Bending Stiffness (per unit width) and m is panel mass per unit area.
As I understand it, the speed of a transverse wave is proportional to (B/m)^0.5, and sound radiation is maximized when the transverse wave speed approaches the speed of sound. Hence, increasing B/m increases wave speed, and hence increases sound radiation.
But I also often see the quantity B/m^3 mentioned. Several patents refer to minimum values of B/m^3 for good performance. What is the basis of the B/m^3 quantity (instead of just B/m)? And is it a better predictor of performance than B/m?
Let me be more specific. For the purposes of this question, let's say panel size is not limited. That is,for the purposes of this discussion, that I'm willing to make the panel as large as it needs to be to have a sufficient modal density down to a particular frequency (say, 100Hz). Assume also, that I have a single exciter (or maybe two) of the sort commonly available (like DMLBES's list above). Assume also for the sake of this question that I know how to support the panel and locate the exciter in reasonably good locations.
Given all that, what is the mathematical expression of the material property I need to maximize, minimize, or otherwise optimize, in order to maximize output?
Is it B/m? B/m^3? Other?
Thanks,
Eric
Last edited:
I can't answer that question directly, but I'm sure its probably buried in some research paper that been posted before in these threads. But lets not focus on just one factor.
According to Henry Azima, PhD (one of the original designers of NXT panel theory) DML's are highly deterministic and you can very accurately determine a panels suitability via 5 parameters:
1. The size and shape of the panel (it can be curved in one or more plane)
2. The position of the exciter(s)
3. The bending stiffness.
4. The surface density
5. The internal damping of the panel material.
So we know the first two parameters from one of the NXT patents. I'm betting that the other 3 parameters for various materials can be found through google or panal material manufactures websites. The rest is probably going to be guess work (trial and error).
However, the set of equations needed I beleive are proprietary to NXT and thus not available without a License from the current patent holder. NXT made software , Sound Designer, that included all the experimental data needed to make these determination that you would receive with the License. I know there was at least one person in one of the DML forum threads years ago that said they had the NXT software and might be able to help...though I would be cautious of licensing liabilities. BTW, you can also find the software NXT used for free as well, however without the NXT extensions and proprietary data. (that can be found in one of the DML threads as well)
Trying to determine the correct equations and parameters took NXT Researchers millions of dollars over several years to figure out. The best most of us can do is simply understand the parameters and make best guesses as to what seems to work best based on all the research that is available.
By the way, a paper published by Azima (more of a marketing paper for NXT, but also published in Audio magazine in the 1990's (yes the Audio article is available via google as well) gives the above info as well as a few tips on the how and why a DML works. Its easy reading and may give you more insights:
21st-NXT/Azima
According to Henry Azima, PhD (one of the original designers of NXT panel theory) DML's are highly deterministic and you can very accurately determine a panels suitability via 5 parameters:
1. The size and shape of the panel (it can be curved in one or more plane)
2. The position of the exciter(s)
3. The bending stiffness.
4. The surface density
5. The internal damping of the panel material.
So we know the first two parameters from one of the NXT patents. I'm betting that the other 3 parameters for various materials can be found through google or panal material manufactures websites. The rest is probably going to be guess work (trial and error).
However, the set of equations needed I beleive are proprietary to NXT and thus not available without a License from the current patent holder. NXT made software , Sound Designer, that included all the experimental data needed to make these determination that you would receive with the License. I know there was at least one person in one of the DML forum threads years ago that said they had the NXT software and might be able to help...though I would be cautious of licensing liabilities. BTW, you can also find the software NXT used for free as well, however without the NXT extensions and proprietary data. (that can be found in one of the DML threads as well)
Trying to determine the correct equations and parameters took NXT Researchers millions of dollars over several years to figure out. The best most of us can do is simply understand the parameters and make best guesses as to what seems to work best based on all the research that is available.
By the way, a paper published by Azima (more of a marketing paper for NXT, but also published in Audio magazine in the 1990's (yes the Audio article is available via google as well) gives the above info as well as a few tips on the how and why a DML works. Its easy reading and may give you more insights:
21st-NXT/Azima