A Study of DMLs as a Full Range Speaker

I'd be surprised if you could get away with less than 12" wide and and still get down to 100 Hz, at least with any of the 5 ply plywoods I'm familiar with, which are all nominally 1/4" thick (typically closer to about 0.20" actually thickness).
I too would like to use a smaller panel (and still get down to 100 Hz)
Thank you Eric. Yes, we all would like small panels that go lower than 100Hz. I just wanted an indication of the size limits of the plywoods. Using a high aspect ratio (3:1 or 4:1) with a 12” wide panel brings me back to the size of the Tall Blondes.

I really appreciate your experiments, measurements, findings, and contributions on this thread.
 
  • Like
Reactions: 1 users
IT ONLY JUST OCCURED TO ME >
There many, many numerous reports of how DML's sound great but don't necessarily measure well.
The fast attack but slower decay is exactly analogous to Reverb, and a fundamental property of the panel/plate.
In recording, short time reverb is used to enhance sounds, and also create a 'space'.
This can explain how DML's may not be technically accurate, yet sound truly great !
The following info. of "Plate Reverb" is surely something to think about :)
https://www.vintagedigital.com.au/emt-140-plate-reverb/
In some ways that is correct. There are basically two types of effects, those that introduce delays, and those that directly modify the waveform.

In the first group you have delay effects obviously, but also reverb, chorus, phaser and flanger. Also EQ in a sense, which is a frequency specific delay.

In the second group you have compressor, expander, distortion, saturation and waveshaping.

Since the plate will emit the same frequency from multiple points you will hear multiple versions of the same signal with a slight delay, which is in essence what most delay based effects will be doing, just with different number of copies with different timing.

I would say a good plate have more of a chorus effect, which is a few copies of the signals with slightly different delay. Reverb is a burst with lots of delays.

I don't think slapping an effect on your master is a good idea though. That should be done during production or mastering if it benefits a track.

But of course DML is not an effect you can add to the signal. Adding a reverb to the mix will not be a substitute. DML is an effect that has to be created in physical space, and works in a way that our hearing and even shape of our ears are constructed to handle.

Perhaps using some psychoacoustic modelling listened to on headphones could recreate the sound of DML, but in general, unlike your regular chorus or reverb, DML effect can result in a sound that no producer or mastering engineer can achieve.
 
The reason for damping is that the spectrogram of an undamped panel will look something like this:

View attachment 1256997

Now I'm not saying that the spectrograms of your panels look like this. Just that if you have a truly undamped panel, it will indeed look like this.
And what I mean by "look like this" is that at particular frequencies the spectrogram will show ringing long after the signal that created them has stopped. The ringing frequencies are shown by the vertical lines in the spectrogram. In this particular panel, the strongest of these happens to be at around 260 Hz, but there are others near 150 Hz, 560 hz, and about 1.2 kHz, among others.
This ringing happens not because this panel has "problems", but rather because it is simply the nature of an undamped vibrating panel to do so. For such an undamped panel, once excited, it will tend to continue vibrating at those frequencies. Some of those frequencies will be better at radiating sound than others, so those will be likely to stand out in the spectrogram. In this particular plate, the mode corresponding to the ringing at 260 Hz is this one (below) which happens to be one of those that is particularly good at radiating sound.

View attachment 1257010

That said, one thing that I've observed is that when ringing occurs, it is worse at lower frequencies than it is at high frequencies. I usually don't see it much in spectrograms above 500 or 1000 Hz (the example panel is a bit of an exception). Is it because modal density (natural frequencies per octave) increases rapidly with frequency? Is it because at higher frequencies the panel's internal damping is more effective? Or does the viscous resistance of the air against which the panel is pushing provide more effective damping at high frequency? All of the above? Other? I'm not sure. But regardless, panels seem to act as if they are naturally better damped at high frequencies than low. So one thing that might really work, is simply use a high pass filter to avoid driving the panel at the worst of the ringing frequencies.

Another solution is to have damping. Damping doesn't eliminate standing waves (modes). In theory I suppose it could, but in practice it's really hard. Damping just spreads out the resonance frequencies so they don't happen at such well defined frequencies, but instead are spread out over a range of frequencies. Damping could be at the perimeter, or in the panel, or distributed along the panel. It could be damping you added on purpose, or just happened to be in the panel, or even be added inadvertently. But regardless of it's nature, damping is the most reliable way to eliminate ringing in a panel, if you would like to use it in the lower frequency range, where ringing is most likely to happen.

Eric
Eric.
Can you describe to me , a way of physically damping those resonances in your spectrogram, in a way that does not affect all of the other frequencies on the panel?
I personally prefer as little damping as possible but sometimes it is necessary on certain materials.
I usually use a few weights or mounting points, shapes, to sort out these problems, or throw them in the bin, if they are that bad.
A good panel material should have a clean clear sound with no discernable ringing.
Did you read the white papers by Mapp and Azima ?
Were they of any help to you?
Steve.
 
As can be seen in the first picture on the left, I used an angled length of wood to help break up the standing wave on this ply panel.
The second picture shows how I reduced the standing wave on a long narrow panel.
I have posted pictures somewhere on this forum of this panel.
It showed the panel clamped at the bottom, but with an angled brace clamped in a sturdy vice.
You could also cut or brace the top, or both .
I felt that clamping the bottom of the tall thin panel gave a good sound and a sturdy base for supporting the panel.
So solved two problems for the price of one👍
Steve.
 

Attachments

  • 20230911_181753.jpg
    20230911_181753.jpg
    225.7 KB · Views: 98
  • 20220926_001638.jpg
    20220926_001638.jpg
    245.8 KB · Views: 99
  • Like
Reactions: 1 user
Steve, How is your current state of the art panel configured? I know you do all kinds of experimenting with canvas, card stock, etc. - but if I wanted to build and hear what you consider your best effort - exactly what would I build? Thanks, Bruce.
Hello Bruce.
EPS,( NOT XPS) has been my favourite for many many years .
But it needs a lot of delicate care and attention to get it sounding its best.
The same can be said for the other panels, as far as I am concerned, they are all good sounding panels and will not disappoint.
Especially as they are usually cheap or free.
There is only one, that I know of , that sounds good straight from the get go, all you have to do is slap an exciter on it.
There is possibly a way to improve this panel ,I believe.
This sample was sent to me by JohnnoG, but it is very expensive and probably hard to get hold of.
I will have to pm John to see how he is getting on, but I know he is very busy.
It probably depends on how you intend to use them in your environment ?
Steve.
 
Steve - My environment is 14’ x 18’ x 8’ carpeted, acoustical tile ceiling. I want all genres of music to sound right, but the most challenging source example is Sheffield’s Drum and Track Record, Jim Keltner at live levels. I want listening to that to put a smile on my friend’s face! I have high density EPS and a wire cutter to make any thickness. I am not limited by panel size. Will your preferred 10-watt exciter be enough? I have an electronic crossover/equalizer, infinite baffle sub and lots of power. I need dimensions, PVA or epoxy, suspension method, etc. I’ve read and valued so much of your expertise, here and at AC, but I still don’t know what you’d specifically recommend for my needs. Thanks, Bruce.
 
Due to my lack of talent for working with double-sided tape, here's my 14" birch with frame.

birchfr2.jpg
birchfr1.jpg


I ripped 1/4" strips of XPS, which I have plenty of, and glued to the inside of each one of those long boards, and then sandwich the edges of the plywood between the xps, tightening with bolts. I need to get more bolts, there's enough to do the basic job, but, it's a delicate balance to constrain the board enough not to buzz but without killing its relatively epic bass (for dml). The board might look straight but definitely isn't when trying to sandwich it, and so I think at least 4 bolts on each side, and then gradually tighten -- however, might not matter much because at the end of the day, crossing over at the trusty 100hz and letting the sub do its thing may sound the best anyway.

I did a long left-channel comparison with one my xps panels connected in series to my dml tweeter, both powered by the same version of amp and can easily toggle either off and on. More impressive than the bass is the highs, the articulation is off the charts; it might be too much. It also has more depth than my xps. The bad -- cold and bright. Lacking in mids. If I had to do a final build tonight, Id rip 10" off one xps panel and mount it to the front of the birch. Together they sound very good.

Coming from PE in a few days is a test mic. Today my -fhe-4 exciters came. So tomorrow I can try with the same exciter Eric uses. If I can figure out REW, I guess I can see exactly what frequencies are missing and go from there.
 
bdjohns.
Thanks ,it's nice to know that someone is interested in my designs 😀
I usually only use the lower grades of eps , but ziggy loved his large HD eps panels.
Although I did manage to get some samples of 40cm square panels .
The HD should have a better hf to 20k.
Do you have a block of eps , if so what size?
EPS would be very good for drum reproduction, it's like having the drum kit in your room.
Maybe pm me for discussions.

These exciters seem similar and are a good price at the moment.
They are 25watt exciters.
Now you have made me buy 4 of them 🙄😅

https://www.soundimports.eu/en/dayton-audio-daex25fhe-4.html


Steve.
 
  • Thank You
  • Like
Reactions: 1 users
Due to my lack of talent for working with double-sided tape, here's my 14" birch with frame.

View attachment 1258031 View attachment 1258032

I ripped 1/4" strips of XPS, which I have plenty of, and glued to the inside of each one of those long boards, and then sandwich the edges of the plywood between the xps, tightening with bolts. I need to get more bolts, there's enough to do the basic job, but, it's a delicate balance to constrain the board enough not to buzz but without killing its relatively epic bass (for dml). The board might look straight but definitely isn't when trying to sandwich it, and so I think at least 4 bolts on each side, and then gradually tighten -- however, might not matter much because at the end of the day, crossing over at the trusty 100hz and letting the sub do its thing may sound the best anyway.

I did a long left-channel comparison with one my xps panels connected in series to my dml tweeter, both powered by the same version of amp and can easily toggle either off and on. More impressive than the bass is the highs, the articulation is off the charts; it might be too much. It also has more depth than my xps. The bad -- cold and bright. Lacking in mids. If I had to do a final build tonight, Id rip 10" off one xps panel and mount it to the front of the birch. Together they sound very good.

Coming from PE in a few days is a test mic. Today my -fhe-4 exciters came. So tomorrow I can try with the same exciter Eric uses. If I can figure out REW, I guess I can see exactly what frequencies are missing and go from there.
Jmproject.
This looks very similar to this old patent.
He uses this method to vary the damping applied to the panel.
Cohen, was way ahead of his time.
If we had paid more attention to his design instead of cones, where would we be now ?

https://patents.google.com/patent/U...=priority:19610101&oq=inventor:+A+Cohen++1961

1961 👍
Steve.
 
  • Like
Reactions: 2 users
Thanks for the historical note, Spedge. I hope his worked better than mine. Actually, it's good for the job of getting the bass just right, but not sure beyond that. But, I await a mic and we'll see if I can figure out how to do a test.

my boatload of fhe-4's came in. On the panel in the recent pic, I put one of those on in place of the thruster and then later wired it and the thruster in series, and the sum total of those experiments is that it all sounds about the same to me.

Because I just have that one birch panel and the others are somewhat torn apart, I wanted to get some stereo going. So, I took the 10" x 48" piece of birch left over from cutting the main panel, inserted it sideways and about 1" deep into the frame of the panel in that last pic and tightened the bolds lightly. Then put a fhe-4 near either end. Unbelievable how good even that sounds with the right sub mix and a little help from a couple small junker xps panels. There is a use case here -- I need a light, long, thin panel that I can put up behind my recliner in the shop here for rear surrounds on occasion, since I found a place for the Denon here rather than haul it back to the house after the testing over the holidays.

Just the single 10" x 48" piece of birch though with 2 exciters sounds better than the rears I paid $110 a piece for in the 2000s and its portable.
 
Jmproject.
I can't say I am a fan of tall thin panels, but maybe you should try my idea of only clamping the panel in a vice at the bottom , leaving the rest of the panel to vibrate.
I would recommend a couple of sturdy pieces of wood to clamp in the vice each side of the bottom of the panel.
I recently showed a picture of a tall panel with the bracing at an angle in red.
The angle will remove most of the standing waves bouncing up and down the panel, dispersing them.
I would probably not use foam damping in the clamping, to increase the reflections, not reduce them.
Although it would be easy to compare the two, or to use as variable clamping pressure, as in the patent?
Steve.
 
Hello Bruce.
EPS,( NOT XPS) has been my favourite for many many years .
But it needs a lot of delicate care and attention to get it sounding its best.
The same can be said for the other panels, as far as I am concerned, they are all good sounding panels and will not disappoint.
Especially as they are usually cheap or free.
There is only one, that I know of , that sounds good straight from the get go, all you have to do is slap an exciter on it.
There is possibly a way to improve this panel ,I believe.
This sample was sent to me by JohnnoG, but it is very expensive and probably hard to get hold of.
I will have to pm John to see how he is getting on, but I know he is very busy.
It probably depends on how you intend to use them in your environment ?
Steve.
Hi Steve
I've been revisiting some 12mm thick high density EPS panel material and although I'll persist for a while, I find it has a large amount of harsh self noise at reasonable listening levels... almost like vibrations within the depth of the panel... Have you found this with your lower density panels, and if so, how did you treat it?
At this stage I haven't coated it, but have put masking tape in a few spots, mainly directly over the exciter position to see if it tones it down... Marginally is all I'd say

Eucy