And here is how I implemented the spine. This is a simple 3/4" piece of plywood cut to width, and then attached to two corners of the frame and the back of the exciter with a combination of PVA adhesive and stacked strips of MLV. Once the adhesive dries I am going to test the result with a quick pink noise RTA measurement and then install them.
If anyone else does this, the frames look to me like they could be HDPE so they should be flame-treated first or the adhesive may not bond well.
If anyone else does this, the frames look to me like they could be HDPE so they should be flame-treated first or the adhesive may not bond well.
Last edited:
Nice. I think a spine is a good idea, but for prototyping a spine doesn't usually matter much, it's only over time that the surround will start to creep and sag. But If I were you I'd want to test that no matter what anyone said.
Normally, I put the spine across the shortest span, to maximize its stiffness. Minor point I think.
Eric
Haha, you would think that would know that wouldn't you! I certainly should know it, but sadly I do not! I did not want to cut pieces from the panel to assess the moduli. My best guess is that the moduli in the primary axes are similar. But to be sure it is an important question.
My working theory is that it is ideal if the frequency of the 1,3 mode doesn't exceed that of the 1,1 mode by more than about a factor of 1.3.
If the long direction stiffness is higher, you need a higher aspect ratio, and if the short direction has a higher modulus, you can get away with a lower aspect ratio.
Eric
Thank you Eric. This the basis of a design rule. Yes you are right, I was hoping you had made some measurements of the material characteristics which need to cut some samples... Here I am waiting to find again some additional poplar plywood to do it (shortage in the local DIY store). You have already mentioned that the 5 ply you have has similar stiffness in both direction, which is not the case of a 3 ply.
Thank you for the answer about the influence of the stiffness in the 2 directions.
For a material with different stiffness in the 2 directions, do you have an advice in which direction to have the higher one? One possibility is to have it in the direction of the smaller dimension (ie width) so that the height is shorter than the 3.4:1 or more ratio.
If we go a bit further : have you identified some possible calculation prediction of the ratio of the 1,3 mode to the 1,1 mode? If I remember, you already pointed in a previous post that the kind of suspension you have (3M tape) gives a result between simply supported and clamped models.
I have an other question, sorry... There is a phenomena which has not been evaluated for now by the theoretical side which is the rear to front wave cancellation. I wonder in your test what belongs to the mode spacing and to the rear to front cancellation. Have you evaluated this high ratio design in different positions in room or even outdoor? An alternative to outdoor which is not easy to do might be a kind of half space test where the panel is just in the opening of a sliding window or door, the rear wave going outside (or, I am thinking to it while writing) to another room.
In other words, I wonder if there is a relation between the panel width and the targeted frequencies due to that. Not sure to be clear...
Christian
Hello Deude_Mann
My knowledge of acronym and materials is too short...
MLV = Mass Loaded Vinyl?
HDPE = High density Poly Ethylene?
From the pictures, your canvas seems not made from the classical coton or linen fabric. Is it?
Christian
While looking around for info from "scientific" areas I came across some rambling about aspect ratio. It mentioned NXT's magic one which it seems was 0.94 but when commercial samples were measured more like 0.9 was found or around this figure. Very high aspect ratios also mentioned but it seemed to be in a dismissive manner. Nothing of much use in it so didn't keep a link also hearsay.
Another source, a thesis started by lets look at commercial practice and chose
Packard Bell PC speakers. These are oblong and use corex with the ribs running along the long axis. The ratio is ~0.675. Things can be assumed about that in terms of stiffness in each direction but it would be best to measure it. These were produced under the NXT patents. They are roughly A5 sized. - similar to 1/2 letter in USA terms. As very cheap used I bought some. Within there range the panels appear to produce a rather clear sound. Amp though poor as is the woofer that comes with them. Volume adequate for me but but not far off clipping. Tone appears to just blend the woofer in via it's volume control. Not good enough to judge suitability for this use really. 3w too each panel and 6 to the woofer. RMS mentioned but have doubts about that.Hi Eric
It strikes me that it would have been instructive to do chladni tests on each version to see the low frequency patterns and how they change with the change in aspect ratio
Could still be useful with the final panel to actually see the region which generates the 125 Hz tone ... May lead to a solution for the dip or at least a better understanding of why it happens
Eucy
Okay thats pretty good and your Dire Straights one a few pages back #9889 is crystal clear too. 😱
I tried playing the same dire straights track and its certainly nowhere close to yours over my 3mm MDF / twin exciter units.
I dont have a decent mic to record mine yet so that will have to wait.
Pity you dont disclose your methods here as is your right , persevere persevere i shall.
Yes sorry MLV is mass loaded vinyl though technically the stuff I showed is butyl
HDPE is correct
I looked at that and honestly I do not know. It may not be. I did have that nasty ring at 1500 Hz, and the sound is OK not great, so perhaps that could explain why.
OK all I am about to go big, really big, on building another DML. Here is where I am headed...
The large size is the main deviation from the norm in the name of experimentation. I am filling a large 30'x40'x12' space, so even those dimensions sound huge, the panels are not huge relative to the room. I think there will be some interesting results as well with a 3mm ply panel this large, and the large amount of compliance and damping the surround will provide.
Basic build procedure:
I'm opening the floor for comments. I plan to start fabrication tomorrow.
- 3mm hardwood ply panel
- Dayton Audio DAEX25Q exciter
- 1x4 yellow pine frame
- Spine between the back of the exciter and frame
- 3.86:1 aspect ratio (same as the tall blonde)
- Panel dimensions will be large at 7.72 feet x 2 feet (2353 mm x 610 mm)
- Square corners (not rounded)
- No panel surface treatments, not yet anyway. This may change depending on the initial result.
The large size is the main deviation from the norm in the name of experimentation. I am filling a large 30'x40'x12' space, so even those dimensions sound huge, the panels are not huge relative to the room. I think there will be some interesting results as well with a 3mm ply panel this large, and the large amount of compliance and damping the surround will provide.
Basic build procedure:
- Build a frame (glued and screwed), cut the plywood, attach the EPDM surround to the frame.
- Place the panel on top of the EPDM on the frame, with the frame resting on top of towels on top of sawhorses.
- Use pink noise, an RTA, and my ears to find a good spot for the exciter. I'll be aiming for flat response and bandwidth.
- If I find a spot that looks and sounds good, I'll attach the EPDM to the panel, then build and attach the spine to the exciter. Then test some more.
- If I don't find a good spot, or get any really nasty resonances, or some other bad anomaly, I'll stop and regroup.
I'm opening the floor for comments. I plan to start fabrication tomorrow.
At least you are one of the few that can use your own ears to judge the sound quality of a recording.
MDF in my opinion is not a ideal DML material. I said it before in one of my other posts that an ideal DML material will have the least amount of filtering, which is higher density 1/2inch EPS.
The mic I use is on my cell phone that I got almost for free as a promotional offer so its most likely the bottom of the line.
Keep an open mind and dont over think DML's as they are more similar to conventional cone drivers then not as you cant defy certain physics. Dont get caught up in all the technical aspect of DML's instead use your own ears and hands on experience to tell you what works and what doesnt work. Look at patents that have actually accomplished building a good sounding DML panel. Good luck in your DML journey.
Go big or go home!
Here are a few of my thoughts.
Are you sure you don't want to play around a bit with other panel materials first?
What I'm thinking is this: to fill that room you are going to need some serious output, and I'm not sure plywood is going to have it. Size doesn't really give you much more output, if any. Probably you are going to need 4 exciters on each panel, I suspect, and I don't know if even that will be enough.
I'm thinking you might want to give 1/2" PS foam (Dow Foamular, the pink stuff) a try. Maybe make a smaller speaker with each plywood and the PS and see what you think then. Personally, PS foam is not my favorite, but others love it so don't be over-influenced by me. PS foam is really efficient (most efficient I know of) and would have the best chance of filling that room with sound.
When you hear how much difference in efficiency there is between plywood and foam you may reconsider.
To be honest I don't think going that big is going to gain you much. Fun to try, for sure, so go for it if you want, but unless you've tried smaller you won't really know if it does or not. For sure you are in a size range that I've never played in. I normally design for fundamental frequency on the order of 100 Hz. With your panel material and size I expect your fundamental might even be below 10Hz. I'm not saying that's bad, just that it's so different I don't know what to expect. On the plus side, the by the time you get up to 100 Hz, the modal density will be very high, which ought to be good. I'll be interested to see.
Also, I don't think testing different exciter positions before the panel is attached to the foam is going to tell you much. It may actually mislead you. When you just rest the plywood on the foam, it won't likely contact the foam firmly all around, and you will have very ill defined boundary conditions. Hence the best location in that situation probably won't be the best location after the panel is actually mounted.
Good luck. No matter what it should be fun.
Eric
I wondered about this... Hmmm... The idea was to avoid full tilt attachment to the panel until I had reasonable confidence in the result.
As far as size... I was going for conciseness because trying to contextualize all I have read would be verbose by default. But I'll try to keep it brief. Reasons for going big:
- I am not so much seeking output efficiency, but for the output to behave more like a DML and less like a point source by scaling it to the room size. It seems like a really small panel in a really large room would lose the DML qualities, in particular more even response throughout the room.
- Deeper extension would be another benefit but I have a powerful sub in place already so that is not critical (sort of); I am annoyed when I can hear bass notes jumping back and forth between sub and mains and so based on experience I like to cross over no higher than 100 Hz and prefer lower when possible. I've played with DSP crossover settings and time alignment for hours trying to get this to where I like it. Deep bass extension makes that easier.
- To me it seems that the DML qualities will be more pronounced if the panel is free to form waves (like free hanging) but to also have significant edge damping which seems to be reasonably established as desirable. A panel that is very thin with respect to its size, attached to a rigid frame with compliant foam, seems like a good way to achieve these based on all I have read and understand.
Not trying to be argumentative at all, just fleshing out my assumptions. Tone is hard in text.
Last edited:
OK, so I bit.
I just played the above recording while split screen on PC switching back and forth (which only took 1 time) between the above recording and this recording of the same song which is supposed to be the best on YouTube:
The song is : Once Upon A Time In The West - Dire Straights
Talk about night and day, we're talking galaxy distance here. The magic carpet recording is heavily balanced to the left and sounds like it's coming from a cheap....something? To his credit, the uploader did say he recorded it with a bargain basement free phone, which there is absolutely no doubt about.
I have this album (Communique) in FLAC and tested the higher grade youtube video against that and it sounds damn good.
I'm sure those speakers sound very good in person, amazing even as Audiofrenzy has been at this for years and years and has common sense and tons of trial and error under his belt. I've personally paid attention to his techniques and trust the ones he's shared and a few he let slip.
I would love to hear those panels, EPS is my favorite right after my "magic foam panels" of course. Even so, hearing them in person, personal preference would rule and no two people are the same.
Point being though, one can't expect someone to hear what they hear on their equipment. Especially when recorded with crap.
Todd, that should be a given as mostly everyone knows that the better the recorder the better the sound quality that is why studios spend hundreds of thousands if not millions of dollars on recording equipment.
The purpose of my vids or anyones sound clips is so that one can get the gist of how it sounds.
I made another recording but this time i used some techniques to make the recording sound better see if you can tell the difference.
The purpose of my vids or anyones sound clips is so that one can get the gist of how it sounds.
I made another recording but this time i used some techniques to make the recording sound better see if you can tell the difference.
Last edited:
Nice!
HOWEVER....
That size panel will have to have serious damping everywhere to prevent ringing. Somebody mentioned Chladni patterns recently, and it might be a good idea to use these to position your exciters. RTA's are just as informative, but I think they are more hit-and-miss because you cannot actually see the nodes and anti-nodes on the panel itself. Using both techniques at the same time is ideal.
Secondly, the bass in a very large panel might be floppy, both physically and aurally. It will not have the transient punch that a dedicated, front-loaded sub will have. A heavier panel will also not translate into good transients. Maybe a low-density panel with high self-damping will work better? I'm suspecting that you might also get away with a thicker core. This description seems to point to EPS contrary to Eric's suggestion above. I'm not advocating, I'm just hypothesising.
I have tried Twinwall skinned with Kraft paper, and also Twinwall skinned with 5mm EPS. Both measured pretty nice, but not robust enough (considering very rough handling) for live PA purposes.
My next iteration will be done with expanding builder's foam (and with various skins) so that I can vary the thickness of the core across the panel. I'm hoping that I can create "mechanical" cross-overs and use the appropriate drivers where they exhibit their best bandwidths. More on that anon.
Can you clarify please , The 1/2" EPS you refer to , is that the same as polystyrene that is used as packaging when you unbox the average appliance these days , or is it the heavier density EPS grade that has a hardened skin on both sides ?
I have seen both labelled as "EPS" in suppliers i go to so i need to clarify which it is.
Edited ** PS , I have a supplier with "EPS sheet" in 13 kg/3M , 19kg/3M and 24 kg/3M showing on their website.
I have the sense that i should be testing thinner lightweight panels with a stiff facing material .
My list of options is now;
2.5 mm thick Poplar Ply with a 0.4 mm bonded walnut or Blackwood veneer on the front.
2 mm Balsa panels with a 0.5 mm Carbon Fibre sheet bonded on the back and front.
6 - 12 mm EPS bonded with canvas photo front and -as yet undecided rear skin for fully transmitting the sound wave energy.
6 - 12 mm EPS bonded front and back with a 0.4 mm Blackwood or Maple veneer.
What opinions do people have for pro's and cons ?
Selection of the adhesive type could create large variations in the above.
Last edited:
It sounds a little tinny to me in places for a finger plucker which he is. My PC speakers are too close together for good stereo yet it's very apparent in that video. Having talked to some one about mobile phone recording apps I was informed about what I could do to improves a trial recording I made. I also I had a look around at the apps that are available. Dolby On looks interesting as it can record uncompressed however is will also stereo stretch and do a number of other things.
Mind you some might say what - record in wav!!!! Given web limitations I suspect it wouldn't be too bad.The point I am making is that the phone used is only part of the story. An owner of a phone could have £1,000's worth of audio precessing gear around not just a specific app in their phone.
@Audiofrenzy Any chance of a link to the dire straights post - or anybody. That is more like the music I would prefer to judge results with.