Andre Bellwood
Freq response without eq looks amazing. Nice.
Could you post a few more pictures of the construction. It's maybe in another thread, but why the 100 mm hole + plate fixed with glue? Why not fixating the plate directly to the panel without the hole?
Exactly whic Tectonic exciter are you using?
Wondering if it is better than the Dayton ones, and are causing the better freq response.
Could you elaborate a bit more on the sound, and in comparison to other speakers ..
Thanks for posting
Freq response without eq looks amazing. Nice.
Could you post a few more pictures of the construction. It's maybe in another thread, but why the 100 mm hole + plate fixed with glue? Why not fixating the plate directly to the panel without the hole?
Exactly whic Tectonic exciter are you using?
Wondering if it is better than the Dayton ones, and are causing the better freq response.
Could you elaborate a bit more on the sound, and in comparison to other speakers ..
Thanks for posting
Edit: .. ok I think I get it. The plate in the whole is actually responsible for the much better high freq response as this is the only part moving at higher freq. Towards lower freq, the panel will start moving as well.
I would think you thereby also have a more "normal" off axis response with reduction towards higher freq.
It's kind of a 2 way mechanical system
I would think you thereby also have a more "normal" off axis response with reduction towards higher freq.
It's kind of a 2 way mechanical system
Guys, apologies for being here a little haphazardly.
Firstly: Here in darkest Africa (Johannesburg) there's a thing called 'Load Shedding.' The national electricity supplier has been plagued with all kinds of mismanagement, incompetency and corruption, resulting in a electricity grid that can supply only a limited amount of power at a time. And it simply gets switched off from time-to-time, and then we twiddle our thumbs for a few hours...
Be that as it may.
I have not plotted the curves that are below the noise floor (6th - 9th harmonics.) The curves here are masked where they drop below the background noise. I have also not plotted the THD since that is on a previous post.
I'm not totally happy with these distortion curves though. The drivers I'm using (Tectonic TEAX32C30) have had to be raped in order to get rid of the protective rubber shrouds buzzing. I've also had to reinforce the feet that attach to the metal springs. And they're still buzzing a bit.
I'll use Daytons next, possibly the DAEX32EP-4.
Yes, I did suspect that the 100mm polycarb disc was acting like a piston. But I have measured off-axis responses and they're still better than any dedicated tweeter I've ever used. This implies that the disc itself is still acting like a DML panel. My theory is that the weight of that disc is a lot lower than the whole panel, and will therefore perform better at HF.
Once the wind dies down here, and IF there's mains power, then I'll do better off-axis measurements for the system.
Another thing, the disc was mounted behind the panel, on the back skin. In effect it was firing through a 100mm-dia, 20mm-deep cavity. I suspected that this cavity might be resonating at high frequencies, so this morning I re-mounted it on the front skin, and re-measured. No difference, except that quick off-axis measurements show a little improvement over the back-skin performance.
I have a friend close-by who's an Audiophile with a lot of nice gear. I took these panels to his listening room and we compared them. He was stunned at the sound quality. It rivals, if not exceeds anything he's ever had.
Listening with eyes closed, the sound stage is so well-defined that you want to reach out and touch the musos, you can hear the fingers of the bassist as he plays the upright. You swear that you can hear Diana Krall's epiglottis when she starts singing. Trumpets are incredibly lively. Drum solos are tight and transparent. Snares are particularly crisp. I was surprised at the bass extension. Okay, these are not full-size panels; I intend to add multiple drivers and make them 1200mm high for live use on-stage (one-band band with voice, bass and back-trax.)
On the whole, I cannot fault them.
WRT the photos, I WAS going to supplement the bass with 15" open-baffle cone speakers. But after auditioning the panels with and without the cones, I've decided to go without them, and will simply make the panels a bit longer.
Firstly: Here in darkest Africa (Johannesburg) there's a thing called 'Load Shedding.' The national electricity supplier has been plagued with all kinds of mismanagement, incompetency and corruption, resulting in a electricity grid that can supply only a limited amount of power at a time. And it simply gets switched off from time-to-time, and then we twiddle our thumbs for a few hours...
Be that as it may.
Here are all the distortion curves for the 'blue' panel.
I have not plotted the curves that are below the noise floor (6th - 9th harmonics.) The curves here are masked where they drop below the background noise. I have also not plotted the THD since that is on a previous post.
I'm not totally happy with these distortion curves though. The drivers I'm using (Tectonic TEAX32C30) have had to be raped in order to get rid of the protective rubber shrouds buzzing. I've also had to reinforce the feet that attach to the metal springs. And they're still buzzing a bit.
I'll use Daytons next, possibly the DAEX32EP-4.
Thanks Baldin,Edit: .. ok I think I get it. The plate in the whole is actually responsible for the much better high freq response as this is the only part moving at higher freq. Towards lower freq, the panel will start moving as well.
I would think you thereby also have a more "normal" off axis response with reduction towards higher freq.
It's kind of a 2 way mechanical system
Yes, I did suspect that the 100mm polycarb disc was acting like a piston. But I have measured off-axis responses and they're still better than any dedicated tweeter I've ever used. This implies that the disc itself is still acting like a DML panel. My theory is that the weight of that disc is a lot lower than the whole panel, and will therefore perform better at HF.
Once the wind dies down here, and IF there's mains power, then I'll do better off-axis measurements for the system.
Another thing, the disc was mounted behind the panel, on the back skin. In effect it was firing through a 100mm-dia, 20mm-deep cavity. I suspected that this cavity might be resonating at high frequencies, so this morning I re-mounted it on the front skin, and re-measured. No difference, except that quick off-axis measurements show a little improvement over the back-skin performance.
I have a friend close-by who's an Audiophile with a lot of nice gear. I took these panels to his listening room and we compared them. He was stunned at the sound quality. It rivals, if not exceeds anything he's ever had.
Listening with eyes closed, the sound stage is so well-defined that you want to reach out and touch the musos, you can hear the fingers of the bassist as he plays the upright. You swear that you can hear Diana Krall's epiglottis when she starts singing. Trumpets are incredibly lively. Drum solos are tight and transparent. Snares are particularly crisp. I was surprised at the bass extension. Okay, these are not full-size panels; I intend to add multiple drivers and make them 1200mm high for live use on-stage (one-band band with voice, bass and back-trax.)
On the whole, I cannot fault them.
WRT the photos, I WAS going to supplement the bass with 15" open-baffle cone speakers. But after auditioning the panels with and without the cones, I've decided to go without them, and will simply make the panels a bit longer.
Attachments
Last edited:
Pete, I think it was just common sense to me. The exciter would drive a lighter weight to a higher frequency, while the physical "impedance" of the larger panel would be coupled to the driver via legs, and operate at a lower frequency.
I've never actually thought about it in detail.
I think that fortunate happenstance had something to do with it too.
I'm sorry I don't understand what the exciter vibrates on only on that polycarbonate disc? and the rest of the panel what is it for? you could also put some photos in front and behind.
Thank you
The disc is fixed to the rest of the panel via "legs" of hot glue. You can see the black dots in the photo of the back of the panel where I marked the 60° positions before mounting the disc with hot glue to the panel.
Yes it's a bit messy, but the proof of concept seems to work nicely.
Disc density and stiffness will effect the speed of sound through the material and will dictate what size the disc dia should be.
I think you're a bit too modest to suggest those of us clamouring to squeeze info out of you have such "common sense". It appears to me to be at the very least very intuitive. I was going to ask in response to post #72 if you arrived at this performance incidentally or was it a matter of experimentation? Either way, if those graphs are accurate some actual further experimentation would rival these against anything.
Thanks Pete.
umm...
Okay. I have indeed been experimenting for a while with all kinds of materials, sandwiches, laminations, weights, damping, shapes, etc etc etc.
I loaded Tolvan's 'Basta' ( https://www.tolvan.com/basta/Basta.exe ) to experiment with edge diffraction on various baffle shapes, keeping in mind of course that such edge diffraction is only really applicable to pistons and fixed baffles, and not vibrating baffles; I used Basta to correlate the measurements I was getting from DML panels, and with speed of sound in various materials vs speed of sound in air. But Basta cannot easily model cut-outs in a 2-D plane, so I used SolidWorks to model modes and nodes in different materials at different shapes and frequencies. But that's not useful.
None of this told me anything concrete.
The only thing that really worked for me was to actually BUY different materials and cut funny shapes out of them to test for resonances, modes (horizontally-clamped panels with glitter sprinkled over them) and edge diffraction, and to see if they lined up with the simulations and models. None of them did. Funny shapes, although interesting, just did not work as predicted, and I don't have the time right now to chase down those rabbit trails to find out why, although I could make some good guesses.
However, I've learned that certain notches in FR are due to mundane old panel-size acoustics. For example, a 1,300mm x 500mm panel of polycarb (which has a speed of sound of approx 2270m/s) will have unavoidable notches at around 1,700hz and 870hz, (due to panel length) and also 4,500hz and 2,260hz, (due to the width.) Notches due to air speed-of-sound in the same panel, would be around 340hz, 260hz, 170hz & 130hz. This is not rocket science, and it's the only math that I can make fit with the measurements with the time available to me.
I also cut lines and curves into various panels to isolate different sections of the panels from each-other in order to force parts of the panels to resonate at different frequencies. It was somewhere long this line of enquiry while I was cutting various diameter concentric pieces out of the panels, where I started calculating the required resonant freq with material properties with specifically-required wavelengths, and eventually managed to smooth out the FR curve and simultaneously boost the higher frequencies.
KEEP IN MIND that the speed of a bending wave over a panel will depend on frequency and is NOT the same as the acoustic property. (A surface wave in water, for example, moves at a different speed to the acoustic speed of sound in water.) This adds a bit more complexity to the whole thing.
More complexity would obviously be due to panel resonances as a result of material properties (stiffness/bending ratios) themselves.
I'd really love to spend some time working out how all of these properties are interrelated and then produce a set of equations similar to Thiele-Small parameters. But I'd need to do a thesis to do it properly, and unfortunately time and money right now are very few, and I need to bring home bacon.
Last edited:
Sorry but maybe the translations distort the sense of what you want to say, but what did you tend to say for common sense and steal information ?, this is a sharing forum, if you publish things voluntarily it is normal that you can receive requests for information or other.
Apologies Pixel,
I'm not sure what you want to know, do you want a layman's step-by-step instructions to build such a system?
You do understand of course, that this is all experimental right now, and that there are no fixed rules at this stage of this design? There's more than enough information in what I've posted for anybody to duplicate what I've done so far.
so great that there are no recipes for this thing, but what i dont understand what do you want? I have done one on a specific topic, just because there are no exact protocols but only empirical ones like everyone who is here is of pure interest and within the limits of personal research. you've got enough of it