I also think DML seems surprisingly forgiving both to what amplifier is used and crossover settings. So you probably don't need anything advanced for crossovers, but I think it can be nice to EQ the plates, so a DSP unit can be practical.
You can get decent DSP boards very cheap, like the Alldsp, if you want to integrate it in a build.
I use a Omnitronic DXO-24 E in my studio. It is cheap and has no fan, so quiet, and can handle x-over and EQ fine.
MiniDSP would offer more DSP power and options in a similar price range, but with RCA instead of balanced IO.
I just got one of these for my sound system rack, and extremely happy with it:
https://www.alibaba.com/product-det...a2756.order-detail-ta-ta-b.0.0.6d702fc2GQEcKk
Lost of filter options including FIR, every input can be either analog or digital AES/EBU, sounds great and nice to use both from hardware and software...but does come with a fan. Not super loud, but to use in a home situation I think it would need to be modified.
You can get decent DSP boards very cheap, like the Alldsp, if you want to integrate it in a build.
I use a Omnitronic DXO-24 E in my studio. It is cheap and has no fan, so quiet, and can handle x-over and EQ fine.
MiniDSP would offer more DSP power and options in a similar price range, but with RCA instead of balanced IO.
I just got one of these for my sound system rack, and extremely happy with it:
https://www.alibaba.com/product-det...a2756.order-detail-ta-ta-b.0.0.6d702fc2GQEcKk
Lost of filter options including FIR, every input can be either analog or digital AES/EBU, sounds great and nice to use both from hardware and software...but does come with a fan. Not super loud, but to use in a home situation I think it would need to be modified.
In a separate message, Homeswing (Christian) posed the following question.
What I have observed is that increasing the size of a panel mainly influences the low frequency, with larger panels providing increased output at low frequency. Here are two examples.
First are results for three different XPS panels (14 mm thick). One 4.5" x 6", the next 9"x12", and the third 18"x24". Note each is the same aspect ratio and twice as long in both directions as the previous. All where hung "freely" using only two strips of tape, and the exciter (DEAX25VT-4) was placed in the center. Note that at low frequencies the output from the larger panel is always stronger. However, between about 1 and 2 kHz they start to converge.
A second result is for two 3-ply (3 mm thick) birch panels, one 10"x16", and the second is 24" x 32". In this case the panels were slightly different aspect ratio, but the larger one is still roughly four times the area of the smaller one. Again, the panels were hung freely, this time with the exciter placed at the 40/40 position on each. As in the set above, the larger panel provides more output at lower frequencies, but the frequency response tends to converge at higher frquencies (in this case around 500 Hz).
Eric
- For a given material, we know that the sensitivity doesn't change drastically according to the surface. One remaining question is : "is there a change in the mid to treble balance when the area is changed?". In other words, does the FR remain with the same shape to the HF?
What I have observed is that increasing the size of a panel mainly influences the low frequency, with larger panels providing increased output at low frequency. Here are two examples.
First are results for three different XPS panels (14 mm thick). One 4.5" x 6", the next 9"x12", and the third 18"x24". Note each is the same aspect ratio and twice as long in both directions as the previous. All where hung "freely" using only two strips of tape, and the exciter (DEAX25VT-4) was placed in the center. Note that at low frequencies the output from the larger panel is always stronger. However, between about 1 and 2 kHz they start to converge.
A second result is for two 3-ply (3 mm thick) birch panels, one 10"x16", and the second is 24" x 32". In this case the panels were slightly different aspect ratio, but the larger one is still roughly four times the area of the smaller one. Again, the panels were hung freely, this time with the exciter placed at the 40/40 position on each. As in the set above, the larger panel provides more output at lower frequencies, but the frequency response tends to converge at higher frquencies (in this case around 500 Hz).
Eric
one clarification on the crossover question: anyone have an opinion on whether cutting out low frequencies results in less risk to melting the exciter at high volumes? I think I have plenty of headroom for more volume should I need it, but would be nice to know that there's that little bit more protection -- although one active crossover costs the same as a handful of exciters.
If I recall, Eric uses his panels in a DJ capacity and really drives them hard -- curious if he thinks it's worth it.
If I recall, Eric uses his panels in a DJ capacity and really drives them hard -- curious if he thinks it's worth it.
Hi Eric, thank you - it’s the first time we see the use of the DAEX25VT exciter in this thread. You seem to prefer 25mm exciters.
Just a casual observation, scrutinizing and comparing exciter specs (without actually experimenting with any of these exciters) – the DAEX25TP-4 also deserves a try. The specs are very close to the 25VT, except for a lower inductance (0.10mH vs 0.24mH). You could possibly get a better HF response because of this. Just cut off those silly crooked arms. I am considering using the 25TP for my next build.
Just a casual observation, scrutinizing and comparing exciter specs (without actually experimenting with any of these exciters) – the DAEX25TP-4 also deserves a try. The specs are very close to the 25VT, except for a lower inductance (0.10mH vs 0.24mH). You could possibly get a better HF response because of this. Just cut off those silly crooked arms. I am considering using the 25TP for my next build.
Using a HPF makes a big difference even if the panels will roll off a lot of the low end naturally. At sub frequencies it seems like very little energy becomes sound and most of it just heats the exciter, so using a HPF will increase the power you can feed them significantly even at a frequency where they hardly affect the sound.
One important aspect when it comes to low end reproduction is suspension. I found that free hanging plates tends to give a bad response in the low end, it is like the plate swings around with the low frequencies so it phase cancels itself. With smaller plates and good suspension you get more of a piston effect in the bass region, and it will sound both louder and tighter.
Sorry guys - I see the DAEX25VT has featured many times before. My search failed me. Maybe a typo.
Loeb,
No doubt. I used free hanging panels for that set of tests only because it was a lot easier than making frames of multiple sizes. I think you are right about the cancellation idea with free hanging panels. For a free hanging panel, the first flexural mode looks like this (end-on view) with the center of the panel pushing air one direction while the ends push air the other direction. In the far field, the two will largely offset each other.
But with suspension (as below), the first flexural mode has the whole panel going in the same direction at the same time, so no cancellation. I do have a pet peeve against calling it "pistonic" as it clearly involves bending, but I suspect the radiation characteristics are similar to that of a rigid piston as you suggest.
And one thing I didn't realize for a long time is that the power radiated by the fundamental mode (M=1, N=1) is a huge fraction of the total radiated power, even at frequencies far, far above it's own natural frequency (at least for a well supported plate), as shown in the figure below from Hambric.
https://www.researchgate.net/public...und-Structure_Interaction#fullTextFileContent
Eric
Leob,
Since I started doing impedance measurments, one thing I've been wondering about is whether or not adding a brace to support the exciter significantly reduces the generation of excess heat at low frequencies. The reason I wonder about that is because I've observed that when I add a brace, the big impedance hump associated with the exciter's own mass and spider completely disappears. See the impedance plots below for the same speaker with and without a brace/spine supporting it. If the impedance hump is eliminated, wouldn't that suggst a lot less heat generation too? What do you think?
Eric
Thanks Leob. Since parts express has their 15% off sale I decided I couldn't let it go so I just ordered the rolls 2-way stereo x-over.
I have a couple new subs on the way from PE and it soon may no longer be the case that the panels have as much headroom over the subs.
2 of my four thrusters have braces. I didn't build braces for the other two because it's kind of a pain in the *** and I couldn't hear any difference between the two. I guess I'm getting my x-over just in time to save the other two thrusters from that 30 hz spike.
---
Unrelated to the above, a couple of thought fragments from the exchanges with Mister Audio. Someone (Veleric?) mentioned frames; I'd say definitely try it that way. With DML (as probably any dyi audio) it unfortunately isn't as cheap as advertised because you've got to do your own R&D and see for yourself on a lot of points. I went the DML path based on the infamous Tech Ingredients video -- the main draw was good cheap sound in a lightweight or wall-mount package to fit my space constraints. I've had studio monitors (very mid-grade) that are heavy, and were unused in 13+ years because too bulky to incorporate into my desk area so just relied on headphones. So I went into DML with a lot of hope and was surprised over how good it sounded throughout -- there were some issues though. The one improvement that sealed the deal and I went from temp panels I was fooling around with to shooting for a final build was the frame. First thing I hit it with was one of those audiophile jazz tracks with a big deep cello, and I now had a big full-range speaker. It also got rid of the ringing problem. One of my go-to test tracks is "Colour to the Moon" by Allan Taylor. It opens with a questionable artistic flourish of mic noise as it fades into the guitars. It was ear piercing, and I'd brace myself for that part as the song starts out. The frame also fixed that, and a little added help by spray painting them. All kinds of additional benefits: XPS panels don't come perfectly flat but the frame fixes that, I could build multi-sub-panel panels and move them around to different positions in the room etc..
ETA: when I say big full-range speaker, that only counts for acoustic music. They can stand alone to a point, but I listen to all kinds of music including synthwave and they actually sound fantastic with synthwave but yeah, definitely need a sub for that kind of stuff.
I have a couple new subs on the way from PE and it soon may no longer be the case that the panels have as much headroom over the subs.
2 of my four thrusters have braces. I didn't build braces for the other two because it's kind of a pain in the *** and I couldn't hear any difference between the two. I guess I'm getting my x-over just in time to save the other two thrusters from that 30 hz spike.
---
Unrelated to the above, a couple of thought fragments from the exchanges with Mister Audio. Someone (Veleric?) mentioned frames; I'd say definitely try it that way. With DML (as probably any dyi audio) it unfortunately isn't as cheap as advertised because you've got to do your own R&D and see for yourself on a lot of points. I went the DML path based on the infamous Tech Ingredients video -- the main draw was good cheap sound in a lightweight or wall-mount package to fit my space constraints. I've had studio monitors (very mid-grade) that are heavy, and were unused in 13+ years because too bulky to incorporate into my desk area so just relied on headphones. So I went into DML with a lot of hope and was surprised over how good it sounded throughout -- there were some issues though. The one improvement that sealed the deal and I went from temp panels I was fooling around with to shooting for a final build was the frame. First thing I hit it with was one of those audiophile jazz tracks with a big deep cello, and I now had a big full-range speaker. It also got rid of the ringing problem. One of my go-to test tracks is "Colour to the Moon" by Allan Taylor. It opens with a questionable artistic flourish of mic noise as it fades into the guitars. It was ear piercing, and I'd brace myself for that part as the song starts out. The frame also fixed that, and a little added help by spray painting them. All kinds of additional benefits: XPS panels don't come perfectly flat but the frame fixes that, I could build multi-sub-panel panels and move them around to different positions in the room etc..
ETA: when I say big full-range speaker, that only counts for acoustic music. They can stand alone to a point, but I listen to all kinds of music including synthwave and they actually sound fantastic with synthwave but yeah, definitely need a sub for that kind of stuff.
Last edited:
The impedance peak without a brace actually reduces heat (and power consumption) at very low frequencies.
Another thing I would mention is, so long as your amp isn't low power, the 4 ohm ratings give you the ideal
opportunity to use two in series, thereby giving you twice the power handling plus a nice 8 ohm load.
Also, a simple 6dB/first order single capacitor (* made from 3) filter can further increase 'power protection'.
* low cost capacitor made from 2x 1000uF 63V in series (negative to negative) plus a 47uF Bipolar across that.
Interesting...cannot say I understand what is going on really, and in practice I always have a HPF higher than exciters FS when pushing the plates.
But instinctively I would say that bracing will give you a bit better bass response, and at 35hz you will get more acoustic power, but also use more electric power with the lower impedance. So my guess is that unless you turn it down to match the level, the exciter would become warmer with the bracing.
No doubt that using multiple exciters in various positions will provide interesting results. But what exactly do you mean by "symmetrical drive" and "dead center resonance"? I think those are new terms here. Can you explain why (or if) that is preferable?
Eric
If you have only one exciter mounted dead-center, is sets up/generates a 'common mode' resonance.
I have seen this in other peoples measurements.
This is why many believe in a particular ratio of exciter position offset. (I think it's 2/5 - 3/5 side to bottom)
However, this 'position ratio' leads to very asymmetric drive of the panel.
So, mounting two exciters vertically above and below center you can create symmetry.
I have seen this in other peoples measurements.
This is why many believe in a particular ratio of exciter position offset. (I think it's 2/5 - 3/5 side to bottom)
However, this 'position ratio' leads to very asymmetric drive of the panel.
So, mounting two exciters vertically above and below center you can create symmetry.
Mister Audio,
I'm still not sure what you mean by "common mode" or "dead center" resonance. Neither term is familiar to me. Can you explain?
It sounds like you might be talking about the fundamental mode, but I'm not sure. That would look like this for a free panel (first image) or simply supported panel (second image).
I'm not sure exactly why you would not want to drive that mode, but placing exciter(s) at the 2/5 location will still drive that mode pretty effectively. To avoid driving it, in the case of the first image (free panel), you could place the exciters along the blue vertical stripes (nodal lines) which are at about 22% of the panel length from each end (close to 1/5, rather than 2/5). Another way, which would "work" with either mounting, would be to place two exciters in symmetrical locations, but wired out of phase.
But again, I'm still not sure I understand what you mean by common mode resonance, and if you really mean to eliminate it, (or moderate it?), and for what purpose? Can you clarify?
Eric
I think the point when tuning a DML speaker is to distribute the modes to make an even response. If the fundamental mode is very strong you will get a narrow response, so it makes sense trying to tame that.
However, somewhat to my surprise, exciter placement doesn't really make that much difference. Material, suspension and dimensions all seem to have much more impact.
However, somewhat to my surprise, exciter placement doesn't really make that much difference. Material, suspension and dimensions all seem to have much more impact.
You're both onto what I was trying to say. I have been using my own terms/words to explain all I have read & watched.
( I actually have a degree of confusion between modes and nodes, although related )
The one thing I have noticed is a very strong consensus that a single exciter should not be placed dead center.
It would be good if someone else could 'jump in' and expand upon this topic of exciter location.
PS.
I would never use twin exciters out of phase.