Hi Ben, Guitar luthier here. Great work. About the bracing, it's has multiple functions that you didn't took into consideration, especially the main one : helping the membrane to resist to the string tension (about 80kg for a steel-string guitar). So it needs to be extra extra stiff to avoid complete deformation of the top under tension. Using a thicker membrane make it too heavy to dissipate string energy efficiently so we use bracing to get stiffness and avoid extra weight. In your DML speaker you don't need them and adding them probably accentuate some resonance mode (I think you said around 200hz). If I understood well, a DML needs to have as few resonance peaks as possible. Another function is to compensate to orthotropic properties of wood which is much stiffer along the grain that across is. So bracing is designed to make it a bit stiffer in the cross-grain direction and balance the speed of propagation of the waves in the top. You wouldn"t need that either as acrylic is isotropic and homogeneous.
Hey Luke thanks very much. Yeah you’re right it it similar- especially the way ML supports their panels with a conventional sun/woofer. Appreciate the kind words 🙏
Thanks a Stack, Greetings from ChCh, NZ .... I really like your 'Simplistic', yet Very in Depth approach to the whole DML Issue .... Brilliant Education and Entertainment ..... Very Best to You and Yours from NZ
Since the speakers I like at my LAS are $4,000, which I can't afford BTW, i decided to investigate some diy options such as what you are doing. Thanks for sharing your work!
@@TypicalBenYT ive gone through just about 20 different ideas. Only to realize my first 3 combinations were best . Doing a full build this summer with added subwoofers Love the clarity of the system . Just needs to have that gut feel of bass. If and when I get to build my dream home . I have an entire system I've designed to fill the living room. It will be hidden within the design of the room.
Regarding resonance - I seem to get the best results (both frequency response and faster decay) by using them, compared to free hanging panels. In my testing on acrylic panels (1200x400mm and 600x400mm) the best results have been with the spacers along each edge at ~250mm intervals - resulting in 5/side on the large panel and 3/side on the smaller one.
Thanks Craig. That was my thinking for trying out the mat design. Of course there are many different ways to brace but I just picked the one that seemed to be most common in the guitar.
@@TypicalBenYT Some other RUclips diy audio people have said dml speakers don't image. I have found my dml diy speakers image incredibly well, or at least on par with any other diy speakers. What have your experiences been like with sound stage imaging?
@@sebastianallen I'd agree with your assessment. I find them to be pretty good but not exceptional. Having said that, I haven't had a great room to listen in, which I suspect is about half of the equation! I had a friend over a few months ago - someone who has a pretty nice system in the $x,xxx range - and he was shocked at how present the vocals sounded (I think we were listening to Radiohead at the time) and how they took 'centre stage' with the rest of the band wrapped around them.
@@TypicalBenYT Agreed. It doesn't make sense from a physics perspective why well designed DML speakers like you are after couldn't sound stage/image as well as other conical speakers. I love your idea of a hybrid DML speaker, but I'm actually going the opposite direction, and wanting to add a normal tweeter, rather than the low end. But I'll be a bit coy, and not share my design yet publicly. Happy to chat in some other medium if you're curious to hear more. With my design, I'm actually getting too much bass. Not doing anything crazy with amps or anything, mostly normal power.
first of all, you did a good job explaining the results. If I get you right, using the acrylic panels, a DML speaker has better freq. response when the board if fastened at 6 points (4 corners and 2 sides) compared to a free hanging design. correct?
That was my experience with the larger sized panels I tried (900x400mm). With the 600x400mm panels it didn't make enough of a difference so I only fastened the corners and supported the sides with some soft foam.
I've watched a bunch today and I've had a thought since your going to use plexiglass why not use an acrylic mirror it should function and be aestetic as well
've watched a bunch today and I've had a thought since your going to use plexiglass why not use an acrylic mirror it should function and be aestetic as well
You're right there's no reason it has to be transparent. Acrylic panels also come in a variety of solid and opalescent finishes which could be made to look good depending on the frame color and the room I guess. Mine actually aren't going to be transparent either - I'm just using these for my testing.
Hi Pete thanks a lot. The mic is the Dayton IMM6. It’s meant for smartphones I believe but my laptop has the same TRRS 3.5mm jack and it works great. I don’t remember exactly where I bought the spacers but they’re similar to this: www.bunnings.com.au/surface-gard-22mm-black-round-rubber-bumper-6-pack_p0262227?region_id=118057&gbraid=0AAAAADtbEB8hYLm0pS1dW9asJIW-CZtoS&gbraid=0AAAAADtbEB8hYLm0pS1dW9asJIW-CZtoS&gclid=CjwKCAjwyryUBhBSEiwAGN5OCE8VjE6vaey7FWF_4e4gH1nhhMfB-B_95QW9xZtRLx83s4JlQezvaxoCV9UQAvD_BwE&gclsrc=aw.ds
Hi! I'm using the same thruster exciter and got very similar frequency test curves with acrylic as you did with your earlier tests with the bamboo, with the same high area at the 12-13k area and then stead dropping from there on, but in these tests you have no such problems with the high frequencies dropping and the curves look a lot different anyway, is that all room mode effect / testing done in different room? I was wondering why can't I get the higher frequencies to play better..? You might know :)
I wish I had a better answer for you but I think my measurements (the HF portion specifically) are flawed. The graph shows what should be a very harsh sounding sloping high frequency response but in reality the speakers don't sound harsh at all. I ran a test recently with the same measurement setup but with the volume turned all the way down (no output) and ran some sweeps. I found that I still got an increasing response in the HF area (and nothing in the LF). I suspect it's some HF noise being generated in my equipment somewhere but I haven't had a chance to troubleshoot it properly yet. My guess is that when I track down the problem and fix it, I'll go back to getting HF similar to yours. I have two other thoughts on the HF 1. For 'full range' flattish response up to 20khz, I suspect a tweeter will be required. Tweeters are extremely lightweight compared to a panel and I just don't think it's possible to accurately recreate that extremely fine movement. My next attempt will definitely be a 3 way design. 2. It might not matter all that much. My finished hybrid speakers sound good to my ears anyway. I mostly likely can't hear 20khz at all. If you're able to reduce the big ramp at 12-13khz but maintain reasonable response to 15+ that might be good enough.
Good job ! A question ..what is a 3/4 centered position ? The other thing ..measurements ,you don't have to tunnel or simulate acoustically dead chamber ..the reason is -your are not listening your speakers in those environments. The beauty of active speaker system is - you matching filters (PEQ and FIR) to actual room and this is a reason why you don't have headache about millions money cost "frozen" snake oil charged cables and different amplifiers .All that because you not using hard filters for speakers. Just do measurements wright ,do right FIR and PEQ filtering implementing ,get good amplifier and you are the king.
I'm referring to the placement of the exciter on the panel - at the 3/4 point on the long edge and in the center of the short edge. I get what you mean about the 'live room' measurements, but I'm not planning to use active filters. I'm hoping to build a suitable passive crossover for these speakers, just because I want to have the experience of doing so! In that context I think removing outside influences (room reflections) makes more sense, as what I really need are accurate measurements of individual driver behavior.
@@TypicalBenYT Hi ,it is pain in back side to communicate using You tube comments,because they are deleting everything...so I will tray small steps at the time. You know ,communication like during WW II in concentration cell type.
Thanks! Yeah they are - It seems to be a property of the acrylic as it was similar in my previous attempt. With those speakers I had to roll of the treble control about 1/3 to make it listenable. My next task with these is the crossover, so as well as dealing with the relationship between woofer and panel, I'm planning to roll off the high end electrically to try to make it flat. At this stage my bigger concern is overall smoothness of the response, as peaks/dips in the middle are harder to manage in the crossover.
@@TypicalBenYT Thanks for the reply, we'll be waiting. I settled on the tried and true high grade EPS for my final panels. .......As of right now anyway.
@@toddincabo For anyone in Australia , The XPS foam from "Bunnings" is an el-cheapo lightweight version , I tried it and its too soft for DML Speakers. Guys in the UK and USA who used XPS found that 11 kg per square metre is better , but 15 kg/2m is better again. Our lightweight XPS just does not transmit clean soundwaves even if you coat it with PVA , fibreglass epoxy coating _might work_ but i recommend just going to 3 mm laminex , its crystal clear if you use the right exciters. There is also decent evidence that carbon fibre sheets over 2 mm balsa works very well.
Yeah this would be a really cool look, but I'm sure sound quality would take a big hit. With the panel having a dipole radiation panel, mounting it up against a wall won't produce the best sound. With my floorstanding ones I can pull them out (3 - 5 ft, like an open baffle speaker) and they really come alive.
You said early in the video that lighter = better because of stored energy is a bad bad thing. You the immediately started adding mass which would therefore increase said stored energy and were disappointed in the results. Were you surprised or shouldn't you have seen that one coming from a mile away?
Having more mass in the panel (undamped) = bad yeah, I stand by that. The idea with these experiments was to reduce the overall mass of the panel while controlling the resonance. Both the balsa and the magnets are very light. An inherently stiffer, lighter material would be ideal. I like acrylic for the aesthetics though.
Bracing struts in (well made) acoustic guitars are far more finessed than the ones you made. They taper distally and often have concave recesses and fullers to help maintain rigidity in specific directions but flexible in others. Maybe just put exciters in various acoustic instruments to get a real full range sound! (mute the strings first though)
Yeah you're not wrong - my bracing was pretty clumsy (too thick as well I think). It's definitely something I'd like to look into further in the future.
I have been doing this for over 4 + years, and have way better performance with my own hand built system, that also includes my foils (cables) and connectors, that have only 0.65 gr total pure copper instead of the usual industry used around 35+ grams per each connector. You'll be the judge of it, as the tunes will speak for themselves. ruclips.net/video/0pBGxVLLXDY/видео.html
Hi Ben, Guitar luthier here. Great work. About the bracing, it's has multiple functions that you didn't took into consideration, especially the main one : helping the membrane to resist to the string tension (about 80kg for a steel-string guitar). So it needs to be extra extra stiff to avoid complete deformation of the top under tension. Using a thicker membrane make it too heavy to dissipate string energy efficiently so we use bracing to get stiffness and avoid extra weight.
In your DML speaker you don't need them and adding them probably accentuate some resonance mode (I think you said around 200hz). If I understood well, a DML needs to have as few resonance peaks as possible.
Another function is to compensate to orthotropic properties of wood which is much stiffer along the grain that across is. So bracing is designed to make it a bit stiffer in the cross-grain direction and balance the speed of propagation of the waves in the top. You wouldn"t need that either as acrylic is isotropic and homogeneous.
Great points, thank you! Every day is a school day.
It's like a new way to diy a Martin Logan style speaker. But VERY cheap to do and not nearly as complicated. Awesome video man
Hey Luke thanks very much. Yeah you’re right it it similar- especially the way ML supports their panels with a conventional sun/woofer. Appreciate the kind words 🙏
Thanks a Stack, Greetings from ChCh, NZ .... I really like your 'Simplistic', yet Very in Depth approach to the whole DML Issue .... Brilliant Education and Entertainment ..... Very Best to You and Yours from NZ
Since the speakers I like at my LAS are $4,000, which I can't afford BTW, i decided to investigate some diy options such as what you are doing. Thanks for sharing your work!
No problem, hope your found it helpful in some way!
I've been working on a system design for years. This is awesome
Well done
Thanks a lot. How is your system coming along?
@@TypicalBenYT ive gone through just about 20 different ideas. Only to realize my first 3 combinations were best .
Doing a full build this summer with added subwoofers
Love the clarity of the system . Just needs to have that gut feel of bass.
If and when I get to build my dream home . I have an entire system I've designed to fill the living room. It will be hidden within the design of the room.
Excellent work! Real science and real data ='s real provable end results. Cheers man 🍻.
Thanks a lot! Yep trying things and testing can be slow but should hopefully lead to a decent result.
You are correct. Conventional wisdom isn't always wisdom.
Awesome. I have enormous respect for your approach-methodical and rigorous. How do you find the time? Love your work.
well done
I love your evidence based approach. What material are your spacers made of? How does the spacer material affect resonance?
Hi thanks a lot! The spacers are a soft-ish rubber: tinyurl.com/3y7475h2
Regarding resonance - I seem to get the best results (both frequency response and faster decay) by using them, compared to free hanging panels.
In my testing on acrylic panels (1200x400mm and 600x400mm) the best results have been with the spacers along each edge at ~250mm intervals - resulting in 5/side on the large panel and 3/side on the smaller one.
@@TypicalBenYT Thanks again, and thanks for sharing your experiments with us. What type of material are the spacers made from?
Cool. Reminds me of acoustic guitar bracing.
Thanks Craig. That was my thinking for trying out the mat design. Of course there are many different ways to brace but I just picked the one that seemed to be most common in the guitar.
@@TypicalBenYT Some other RUclips diy audio people have said dml speakers don't image. I have found my dml diy speakers image incredibly well, or at least on par with any other diy speakers. What have your experiences been like with sound stage imaging?
@@sebastianallen I'd agree with your assessment. I find them to be pretty good but not exceptional. Having said that, I haven't had a great room to listen in, which I suspect is about half of the equation!
I had a friend over a few months ago - someone who has a pretty nice system in the $x,xxx range - and he was shocked at how present the vocals sounded (I think we were listening to Radiohead at the time) and how they took 'centre stage' with the rest of the band wrapped around them.
@@TypicalBenYT Agreed. It doesn't make sense from a physics perspective why well designed DML speakers like you are after couldn't sound stage/image as well as other conical speakers. I love your idea of a hybrid DML speaker, but I'm actually going the opposite direction, and wanting to add a normal tweeter, rather than the low end. But I'll be a bit coy, and not share my design yet publicly. Happy to chat in some other medium if you're curious to hear more. With my design, I'm actually getting too much bass. Not doing anything crazy with amps or anything, mostly normal power.
first of all, you did a good job explaining the results. If I get you right, using the acrylic panels, a DML speaker has better freq. response when the board if fastened at 6 points (4 corners and 2 sides) compared to a free hanging design. correct?
That was my experience with the larger sized panels I tried (900x400mm). With the 600x400mm panels it didn't make enough of a difference so I only fastened the corners and supported the sides with some soft foam.
I'll be making a few if this works out. looks simple enough.
I've watched a bunch today
and I've had a thought
since your going to use plexiglass why not use an acrylic mirror it should function and be aestetic
as well
've watched a bunch today
and I've had a thought
since your going to use plexiglass why not use an acrylic mirror it should function and be aestetic
as well
You're right there's no reason it has to be transparent. Acrylic panels also come in a variety of solid and opalescent finishes which could be made to look good depending on the frame color and the room I guess. Mine actually aren't going to be transparent either - I'm just using these for my testing.
@@TypicalBenYT pun intended
SOUNDS great 🤗👍😂
Hi Ben great videos.. what mic are you using for REW and what are the spacers you use for the panel ? Thanks and greets to Oz from Alaska
Hi Pete thanks a lot. The mic is the Dayton IMM6. It’s meant for smartphones I believe but my laptop has the same TRRS 3.5mm jack and it works great.
I don’t remember exactly where I bought the spacers but they’re similar to this:
www.bunnings.com.au/surface-gard-22mm-black-round-rubber-bumper-6-pack_p0262227?region_id=118057&gbraid=0AAAAADtbEB8hYLm0pS1dW9asJIW-CZtoS&gbraid=0AAAAADtbEB8hYLm0pS1dW9asJIW-CZtoS&gclid=CjwKCAjwyryUBhBSEiwAGN5OCE8VjE6vaey7FWF_4e4gH1nhhMfB-B_95QW9xZtRLx83s4JlQezvaxoCV9UQAvD_BwE&gclsrc=aw.ds
I’ve always wanted to visit Alaska… maybe one day I guess.
Hi! I'm using the same thruster exciter and got very similar frequency test curves with acrylic as you did with your earlier tests with the bamboo, with the same high area at the 12-13k area and then stead dropping from there on, but in these tests you have no such problems with the high frequencies dropping and the curves look a lot different anyway, is that all room mode effect / testing done in different room? I was wondering why can't I get the higher frequencies to play better..? You might know :)
I wish I had a better answer for you but I think my measurements (the HF portion specifically) are flawed. The graph shows what should be a very harsh sounding sloping high frequency response but in reality the speakers don't sound harsh at all.
I ran a test recently with the same measurement setup but with the volume turned all the way down (no output) and ran some sweeps. I found that I still got an increasing response in the HF area (and nothing in the LF). I suspect it's some HF noise being generated in my equipment somewhere but I haven't had a chance to troubleshoot it properly yet. My guess is that when I track down the problem and fix it, I'll go back to getting HF similar to yours.
I have two other thoughts on the HF
1. For 'full range' flattish response up to 20khz, I suspect a tweeter will be required. Tweeters are extremely lightweight compared to a panel and I just don't think it's possible to accurately recreate that extremely fine movement. My next attempt will definitely be a 3 way design.
2. It might not matter all that much. My finished hybrid speakers sound good to my ears anyway. I mostly likely can't hear 20khz at all. If you're able to reduce the big ramp at 12-13khz but maintain reasonable response to 15+ that might be good enough.
Isn't it possible that the way the weights contact that table damages the results?
Good job ! A question ..what is a 3/4 centered position ? The other thing ..measurements ,you don't have to tunnel or simulate acoustically dead chamber ..the reason is -your are not listening your speakers in those environments. The beauty of active speaker system is - you matching filters (PEQ and FIR) to actual room and this is a reason why you don't have headache about millions money cost "frozen" snake oil charged cables and different amplifiers .All that because you not using hard filters for speakers. Just do measurements wright ,do right FIR and PEQ filtering implementing ,get good amplifier and you are the king.
I'm referring to the placement of the exciter on the panel - at the 3/4 point on the long edge and in the center of the short edge. I get what you mean about the 'live room' measurements, but I'm not planning to use active filters. I'm hoping to build a suitable passive crossover for these speakers, just because I want to have the experience of doing so! In that context I think removing outside influences (room reflections) makes more sense, as what I really need are accurate measurements of individual driver behavior.
@@TypicalBenYT Hi ,it is pain in back side to communicate using You tube comments,because they are deleting everything...so I will tray small steps at the time. You know ,communication like during WW II in concentration cell type.
accurate measurements....what mic you are using ?
Kick ass man, nice work. So those highs aren't too piercing? Looking forward to more, thanks.
Thanks! Yeah they are - It seems to be a property of the acrylic as it was similar in my previous attempt. With those speakers I had to roll of the treble control about 1/3 to make it listenable.
My next task with these is the crossover, so as well as dealing with the relationship between woofer and panel, I'm planning to roll off the high end electrically to try to make it flat. At this stage my bigger concern is overall smoothness of the response, as peaks/dips in the middle are harder to manage in the crossover.
@@TypicalBenYT Thanks for the reply, we'll be waiting. I settled on the tried and true high grade EPS for my final panels. .......As of right now anyway.
@@toddincabo For anyone in Australia , The XPS foam from "Bunnings" is an el-cheapo lightweight version , I tried it and its too soft for DML Speakers.
Guys in the UK and USA who used XPS found that 11 kg per square metre is better , but 15 kg/2m is better again.
Our lightweight XPS just does not transmit clean soundwaves even if you coat it with PVA , fibreglass epoxy coating _might work_ but i recommend just going to 3 mm laminex , its crystal clear if you use the right exciters.
There is also decent evidence that carbon fibre sheets over 2 mm balsa works very well.
A DML sound wall center channel arround the TV !
High tek look and it wil hide the cables Amigo.
Yeah this would be a really cool look, but I'm sure sound quality would take a big hit. With the panel having a dipole radiation panel, mounting it up against a wall won't produce the best sound. With my floorstanding ones I can pull them out (3 - 5 ft, like an open baffle speaker) and they really come alive.
nice ty ty
Thanks Phillip!
You said early in the video that lighter = better because of stored energy is a bad bad thing. You the immediately started adding mass which would therefore increase said stored energy and were disappointed in the results. Were you surprised or shouldn't you have seen that one coming from a mile away?
Having more mass in the panel (undamped) = bad yeah, I stand by that. The idea with these experiments was to reduce the overall mass of the panel while controlling the resonance. Both the balsa and the magnets are very light.
An inherently stiffer, lighter material would be ideal. I like acrylic for the aesthetics though.
Bracing struts in (well made) acoustic guitars are far more finessed than the ones you made.
They taper distally and often have concave recesses and fullers to help maintain rigidity in specific directions but flexible in others.
Maybe just put exciters in various acoustic instruments to get a real full range sound!
(mute the strings first though)
Yeah you're not wrong - my bracing was pretty clumsy (too thick as well I think). It's definitely something I'd like to look into further in the future.
Have you see Tech Ingredients video? ruclips.net/video/GslJ8Hf4WwE/видео.html
I have been doing this for over 4 + years, and have way better performance with my own hand built system, that also includes my foils (cables) and connectors, that have only 0.65 gr total pure copper instead of the usual industry used around 35+ grams per each connector. You'll be the judge of it, as the tunes will speak for themselves. ruclips.net/video/0pBGxVLLXDY/видео.html