DIY Flat Panel Speaker Love

[LOUT]

I certainly won't argue about there being a lack of deep-bass from most DML's.

I think the main reason larger tweeters aren't common is because their beaming starts at lower and lower frequencies as conventional drivers get larger and larger.
Do DML panels also beam at high frequencies, and does their beaming start at lower frequencies if the DML panel size increases?

I've seen one or two of the larger "full-range" drivers with similar frequency-response and similar high-frequency output both with and without whizzer cones, so I'm starting to think whizzers might be more of a gimmick than a feature. I also haven't seen any whizzer equipped larger full-rangers that managed to avoid high-frequency beaming.

[Unbiasedsound]

My statements are not based on assumptions and guesses though, they are firmly based on what actually works and not just my self but from others before me like Bertagni, NXT/Tectonic. This tech is nothing new as its been around for a very long time, we are not trying to reinvent the wheel here. If you look at all the professional designs of Bending wave technology today you will see that they use DML for the mid to high frequencies because they know DML's are inadequate in the bass department. The very expensive Goebel speakers use a mid/high BW transducer while using conventional cone drivers to cover bass. Manger audio uses there patented Bending wave transducer for mid/highs combined with a conventional cone driver for bass. I use a DML panel for mids/highs combined with a powered sub for bass. Also all the DML/Bending wave transducers are "SMALL" meaning around 12inches or smaller. You will not see a commercially based Bending wave panel that is 24inches or larger and its not due to cosmetic reasons but actual physics. Why are high frequency drivers 1inch and under? Why not make a 10inch dome tweeter? because physics will dictate that it wont sound good which is why dome tweeters are around 1inch. The saying size does matter is very true when it comes to speaker drivers. Full range conventional cone drivers range from 4-8inch but most of them use a "WHIZZER" cone and or a hard type of plug in the middle to increase high frequency response because when the diaphragm is too large the highs are spread out more and the high frequencies become less prominent due to the midrange over powering the highs. The whizzer cone or hard phase plug focuses the high frequencies and prevents them from spreading to the outer edges of the surround material. Every size driver has a "optimal" frequency response as drivers are usually classified into 3-5 catagories. Tweeter, Midrange, Woofer, Subwoofer, Full range.

1-3inch= Tweeter

3-5inch= Midrange

5-1/5-8inches= mid bass

10-21inches=Bass



All I can say is experiment on your own to see if what I and some others have said is true or not.

[LOUT]

I feel like I'd prefer something like objective measurement and direct comparisons before settling upon some/many of those statements super firmly. I'm pretty sure the DML wave formation (flexing-larger/smaller-waves VS pumping-all-at-once) causes different driver diameter/size physics, and conventional drivers (and BMR) use a very flexible surround that doesn't greatly impede their movement rather than attaching more directly to the surrounding frame...so it might be important to follow that philosophy if attaching to a frame under that plan (except instead of just following pumping/Y-axis motion it'll need to non-impede the wave/flex motion which may be really difficult).

The comparison to an open-baffle is probably pretty sensible, but it might be important to keep in mind how much of the "back-side-wave" is SENT via the front of the DML because of the flexing/pattern simultaneously with the opposite pattern since the entire panel is flexing rather than pumping-all-at-once. Otherwise a strangely-shaped enclosure could theoretically help the DML's low-end quite a lot (which I'm pretty sure it doesn't).

This is all just my assumptions and guesses though, so plenty of salt is recommended (the sceptical kind, please...not the bitter/mean kind :P ).

[Unbiasedsound]

After over 6 years of experimenting with DML panels I have learned what a DML panels is. I see so many people on others sites thinking that a DML is a whole completely different animal but in reality its not.

Here is the truth, DML's are more similar to conventional cone drivers then not because of the fact that the physics that work for conventional cone drivers apply to DML as well.

A DML is basically like a open baffle speaker but instead of a cone diaphragm uses a flat diaphram.

This is the reason NXT/Tectonic went from DML panels to a BMR driver which is basically a similar concept but utilizing a conventional cone design using a steel structure to hold the magnet as well as the diaphragm material in place. This BMR concept is nothing new though as Dr. Jose Bertagni has utilized this so called BMR design way before NXT/Tectonic existed.

This is the reason why I highly recommend using a frame and surround to hold the DML diaphragm in place with a spine to hold the exciters magnet in place just like a conventional cone driver and or a BMR driver. This is one of the FIRST and the most BASIC STEP to building a good sounding DML panel.

It seems most people dont even want to follow these tried and true basic steps, instead want to hang there panels by a string because a youtuber with 2 months experience told them to. LOL

If you want to build a DML panel using a single exciter then you build it like you would a single full range conventional cone driver. If you want to use 2 exciters per panel you have to build them like you would a 2 way conventional cone driver utilizing cross overs. If you look at Bertagni speakers he even uses crossovers in his design as his latest designs use 2 or 3 way crossovers like the traditional conventional cone speakers. When using 2 or more exciters on the same panel crossovers can help to reduce cancellations but nothing can completely stop the waves from overlapping. The only way to do it is to use seperate panels the same way one would use SEPARATE drivers.

Even the size of the DML panels matter when it comes to physics. When you look at conventional cone full range drivers from fostex, tang band, etc. they mostly range from 4inch to 8inch. You are NOT going to see a 24inch full range driver, yet many DML panels are 24inches and over. This is the reason I went from larger DML panels to smaller DML panels as you cant defy physics.

To much people are over thinking the concept of DML designs thinking they have magical physical capabilities that defy physics when in fact they are very similar to conventional cone drivers.

[Unbiasedsound]

That guy in the video posted above didnt even do the basic water/glue mixture treatment.

Best shape is square or rectangle.

Also if one is using NPE caps with there DML panels using "by pass caps" will increase the sound quality of the NPE caps. I use 0.68uf and under Dayton or Jantzen metalized polypropolyne. I am going to try the dayton film and foil by pass caps next to see how they sound.

[Unbiasedsound]

Yes I know all those panels sound awful. lol

[Unbiasedsound]

Testing Viewer Suggestions on “The World's Best Speakers” from Tech Ingredients - YouTube

[badfish]

Just some love here. Added a minidsp 2x4 HD and brought my SVS cylinder up (from my home theatre to my garage with my basic panels). Used the DSP crossover to cut out 150hz and below and on my basic home depot panels and played around. bunch with with the EQ. Sound I've achieved at this point trends to very questionably awesome until you remember I'm driving two 22$ insulation panels with a 70 buck amp from amazon. No question 10k systems sound better, and the addition of a 400 DSP and a 1000$ sub make a huge difference, but the mid/high are excellent and the value is mind numbing. Can't wait to build some real panels.

Do notice some probably clipping with my amzn 50W amp.. Might try something with a tad more headroom.

[bilidru]

I've assembled my first panels. Finally I went with 4mm thick plywood of 40,5 x 40,5 cm for the speakers. I thought they had more deep tones than the 2cm thick xps panels. (Didn't get thinner ones in the stores) I also added a 2cm thick EPS of 55x45cm as a subwoofer.and I'm happy with the result.

[LOUT]

Agreed about the $5, I'm mostly being kinda sarcastic and bitter about the higher price just to get one that doesn't look like it'll fall apart if you sneeze in its direction. I remember innitially feeling like the exciters were a really neat way to get a cheap speaker when I originally bought them to try. Now that I've heard what a $8+$13 tweeter+woofer in a small box can sound like, the exciters have sat in storage for a while.
This thread has rekindled my interest, but the traditional speakers are tough competition.
Luckily (silver lining) my main PE project parts are all on back-order for at least a month, so I have an excuse to play with the exciters and BassShakers while I wait.

[LOUT]

I think a lot of my original playing around was using different wood surfaces, so I'll have to remember to try some XPS at that size and think about a trip to a hardware store for EPS if I feel ambitious. The bass response did seem deeper than I remembered it being. I'll also need to experiment with bracing the fullrange exciter which makes sense for helping it reach more of its potential instead of simply relying on its own weightVSmomentum for moving the panel (like you and Kornbread already explained).

[Unbiasedsound]

I'm pretty sure I have a DAEX30HESF-4 I can play around with. I remember it being one of my favorites among the exciters I tried a few years ago. I certainly don't remember it reaching anywhere near 40hz though...I'll obviously need to try some new surfaces/materials. What material and size were you using that let it get that low?

This thread is making me excited to dig out the little transducer collection that I haven't played with in quite a while. I think the cheaper DAEX25FHE-4 was suprisingly decent for the price...and it looks like it might sound similar to the Thruster/DAEX32EP-4 albeit at 24w VS 40w power-handling.
I remember being pretty underwhelmed with the much more expensive DAEX25SHF-4 (though I don't think it was $40 when I bought it). I thought the HDN-8 was kinda garbage and didn't care much for the "SquareFrame" models I tried for whatever that's worth.

EDIT: Looks like I still have;
DAEX30HESF-4 (easily the most robust-looking of any fullrange exciter I've tried, with thicker tinsel leads and reinforced connection to voicecoil),
DAEX25FHE-4 (looks like a cheaper 30HESF-4, but seems to lack some of the HESF's sensitivity and also might lack some of the HESF's bass and treble),
DAEX25SHF-4 (still looks cheaper and more fragile than the HESF and mostly just sounds louder in the mid-frequency...like a louder 25FHE at 4X the price),
DAEX58FP (sounds better than I remember and no exposed+delicate tinsel. Thought it was a little quiet but then realized it's 8ohm...not bad).
I think the HDN-8 and coin-type low watt exciter I had broke. If I remember right, the HDN-8 partly fried (pushed a little too hard when it never got nearly loud enough) and the coin-type got slightly torn/damaged trying to seperate it from a surface it was stuck against with its strong adhesive.

I was using a 2ft.X4ft.X10mm EPS used with a frame and a spine with a sub amp.

[Kornbread]

Surely 2' of braided wire doesn't cost $5. They were others who brought my attention to the solid tinsel wire. They had the same issue. Maybe Dayton made a change, or maybe it's like you said, tiny and fragile.
Still have a large blue xps and treated cardboard panel framed and ready to go for the next exciter experiment, but, as much as I like messing with the odd stuff, a price tag close to $30 makes for stiff competition with some proven cone-n-dome budget drivers.
Still curious about what you find with panel placement. My guess is that the affect panel placement has on the bass will not be as dramatic as a monopole.

[LOUT]

The solid wire tinsel from the tab to vc was the failure point. I'd wager the small solid wire acts just like a wire coat hanger after it's been flexed several times. Add to this the increase in excursion the exciter was likely experiencing with the larger panels, and it literally only took a few minutes until self destruction. If only they would have splurged the extra $.02 and upgraded to a braided wire ...

Basically, this is the motor assembly of a typical cone-n-dome so yeah, you would think excursion would be an issue and something worth mentioning in the specs.

In the PE images of the "Thruster"/DAEX32EP-4, it looks like the tinsel IS braided wire...just really tiny/thin (maybe also too tight and weakly attached).
I think it's similar to most of the exciters I've used aside from the HESF, where it's still a flexible copper braid or strand, but stupidly thin.

The Thruster/EP-4

Still can't believe they kept that delicate build for the ~$40 DAEX25SHF-4 (although the current PE image looks to be a little better than mine). I thought maybe the HESF got more robust because of its higher wattage-handling, but that THRUSTER is also 40watts rated like the HESF but looks to be using the thin stuff.


The HESF-4

I haven't noticed any of the other Dayton fullrange exciters being built with more average speaker tinsel like this. Wonder if this is why the HESF costs $3-5 more.

[Kornbread]

Most of them have REALLY thin tinsel from the tabs to the VC though, and it's often tightly stretched against the spider?/support which doesn't inspire confidence.... So far they've been okay, but XMECH has to be a danger at some point, right?

The solid wire tinsel from the tab to vc was the failure point. I'd wager the small solid wire acts just like a wire coat hanger after it's been flexed several times. Add to this the increase in excursion the exciter was likely experiencing with the larger panels, and it literally only took a few minutes until self destruction. If only they would have splurged the extra $.02 and upgraded to a braided wire ...

Basically, this is the motor assembly of a typical cone-n-dome so yeah, you would think excursion would be an issue and something worth mentioning in the specs.