Acrylic DML Build 3 - Playing Music + Panel Comparison
Vložit
- čas přidán 12. 09. 2024
- 3D Printed DML build parts: www.typicalben....
My drivers finally arrived.
I hereby present my first pass at the rigid acrylic panels. There is still some work to be done, like bracing the exciter, adding a high pass and tidying up the wiring.
Still, I thought it worth a quick video just to show them in action and do a quick comparison with my existing bamboo plywood panels.
The new room is pretty lively so I'm not really able to show these off at their best - a video on room treatment and furnishing is sure to follow some time next year.
Anyway, hope the video is interesting - looking forward to all your comments and ideas!
![구름 찾기 52 ☆ [cloud ☁️] 보석 박힌 개 🐕 #art #그림 #painting](http://i.ytimg.com/vi/0dKUxPN33dU/mqdefault.jpg)
Another great video!!! Thanks for all your efforts in designing, building, recording, explaining and posting your videos. I truly appreciate watching and learning from your journey.
Thank you, glad you’re finding them interesting!
Great Effort and Video ! ! ! Thank you for all your dedication in designing, building, recording, explaining and posting this video. I truly appreciated watching and learning from your journey, without forgetting the part of the technical analysis that demonstrates scientific method ! Congratulations.
Hi Ben,
I have been watching a lot of DML built video's and I like your approach.
About 15 years ago I was deep into DIY Electrostatic Speakers , and I actually have a large hometheatre running my DIY hybrid ESL's in a surround configuration. They are great.
I have been looking at yours and others efforts and I see main flaws . Just like ESL's they are planars and they suffer in being a real full range speakers, in the lows mainly because of phase cancellation. Near feeld measurements will look good, but further away the cancelation will set in depending on the with of the panels and the room . In the highs it depends on the used materials ( the acrylic is better than the bamboo). In the lows , be carefull to use resonances to get as deep as possible. Full range ESL builders tend to do that also and it does not necessarally give the best response ( a bit muddy and less defined ). Mids ( from about 250Hz to 5kHz ) should be fantastic with these speakers. Dare to think of using a subwoofer. ( better 2, because it is far easier to get the dips and knots out of the room with 2)
BUT , I see many DIY ers suffer in searching and not finding that right frequency response. Tbh It does give a speaker the right signature.
And then I have looked ( just now ) at the AER Goldy....and they cheat....but in the way I would also do it to get real good results.
First : They have integrated a normal tweeter into their exciter. That is a point source capable of effortless playing well over 25kHz , without to much resonances. I have read on their website that it takes over as low as 2,5kHz ( seems a bit low)
Second : They use room equalization to get the required frequency response of the total system. To flatten the curve...to eliminate resonance peaks...to get a desired rolloff for a more warm sound....to scape a signature sound that matches the owner.
Using these techniques, they have their panel + tweeter playing in the comfort zone and getting the sound right.
Anyway : What can a DIY er do with this info ?
First : I would consider buying and integrating a tweeter ( with passive filtering) into the concept. If it is a small flat dome tweeter, you could glue it onto the acryl panel. ( They have it integrated in the exciter so not much difference ). Start cheap f.e. a dayton audio TD20F-4 (11€) with a small condensator (15microF or something) as high pass filter and hook it parrallel on the wires of the exciter
Second : I see that you are using a PC to play music. Install the free "equaliser apo 1.2.1" on your windows 10 system and you have a fantastic free tool to experiment with scaping a desired frequency room curve. Of course, only the sounds coming from your PC soundcard will be equilized using this technique, so if you would hook up another source on your amplifier ( cd player) it will not be equalized, just that you know that.
Online there are good pink noise generators that you can play through the soundcard and start equalizing until you get what you want.
You will need a measuring microphone to see the response curve. A behringer ECM8000 omnidirectional costs about 33€ and is perfect for that. I hook it to my behringer ultracurve pro deq2496 (that has a RTA to visualize the frequency response, but I am sure there is a way to connect the mic via a cheap micpreamp to the soundcard of your PC to visualize the measured curve. Up to you to find out, but it is possible. In my opinion , some measuring is always helpfull and needed to get to a good speaker.
Anyway, keep up the good work. After 15 years I am tempted to jump into a built
EDIT : I saw to late that you are already using measurements . fantastic. Please try equalizing as I described. Try also measuring in room response from the place where you listen using pink noise, so that you take into account also room accoustics and not just close mike measuring with a sweep
Thanks for the very detailed info!
Yes I've been using measurements to guide my project and gauge performance (though I'm still learning all of this). Initially I had my bamboo panels running through a mini dsp unit just to flatten the peaks and troughs a bit and I was sending the full range signal into a powered sub with the sub XO at about 150hz.
My plan for these acrylics was to make a 2.1 system (or 2.2!) but half way through the project I've moved house and don't have anywhere acceptable to physically put a sub! Right now I am considering 2 options:
1) Build a new pair using woofers
@@TypicalBenYT
Concidering subwoofers : In my hometheatre room I have the ability to place subs wherever I want, something a normal livingroom does not have, I do realize.
FIrst I placed my rather big 12" DIY adire audio shiva subs front .
-one woofer : a lot of peaks and knots and the knots were of course in the center of the room , my listening place. I didn't even need a measuring tool to hear it.
-2 woofers at the front left and right : a bit better but again a big knot ( hollow)at the center of the room and peaks ( pressure) in the corners ( boomy).
And then I moved the two subs along the right and left walls to about halfway the room,standing against the wall left and right of my listening seat : Perfect bassresponse in the room and not only in the center, but also walking and measuring on many places in the room. Very few knots and jus a little bit of gain in the corners.
I do use my AV receiver to cross over to the pannels at a rather high 150Hz, to avoid problems. You do NOT hear where the bass is coming from. It fills the room .
Measuring bass responce on a panel speaker close up does give you an idea about resonances of the panel ( good to know) but nothing about the real bassresponse in room. F.E. If I measure close mike to my ESL panel, it seems to have a straight frequency response down to 80Hz, but moving away from the panel, the phase cancelation sets in rather quick and I loose a lot of low end and depending on the width of the panel it will be at a higer or lower frequency. Also room interference will kick in. Using resonances/peaks in the behaviour of the panel ( that you measured close mike) to get a higher dB in room is never a good idea, because it will sound muddy and not controlled. So it is always better to just cut off (filter) the panel just above that resonance and let dedicated woofers take over in a controlled way. Using pink noise and measuring at the listening position in room, you then equalize the levels between sub and panel. In my opinion it is always nice to have a more slooping frequency response in room and not a clean straight one. ( so a bit more bass )
About the mid and high frequencies : My opinion is that it is fine to measure close mike ( 1m from the panel , that is).
You will see that if you measure with both panels on at the listening position, it may vary quick depending on the exact position of the microphone, because you will have a constant interaction ( cancelation and reinforcing of some frequencies from the left and the right source ) Don't spent time on trying to eq hights and dips out in the higher frequency band.
Still you do not want an overly ascending" in room at listening position" frequency response in the high end of the spectrum.
As a target you could aim at having a sloping response of around 5 a 6dB between 100Hz and 10kHz.
Don't be worried if it slopes even more between 10kHz and 20kHz, make sure it does not peak there.
Don't be worried to boost a little ifyou go under 100Hz (between 40Hz and 100Hz) but do not try to get a flat responce up to 20Hz in a normal room.
One last remark : Make sure you use ( a good source) of PINK noise to calibrate your speakers in room. Don't mistake it for white noise. Play it at a reasonable level ( say 80 - 85dB )
In pink noise , each octave has an equal amount of energy ( YOU NEED THIS )
In white noise , each frequency has equal power ( DO NOT USE IT TO CALIBRATE )
I like the acrylic panel better. I just finished my first acrylic experiment. Still tweaking because I used really thin pieces of acrylic. It's actually vibrating too much. Im only using 4 watt 4ohm exciters, 2 on each panel to get 8 ohns. I didn't see anyone using acrylic until today. I'm completely new to this. It's very satisfying . Ever tweak is an improvement so far😊.
ABSOLUTELY…FANTASTIC DML DESIGN YOU GOT THERE!!!!!!!😲😲😲 BOTH ACRYLIC & BAMBOO
Thanks a lot! I'm definitely more interested in the framed design now so the bamboo ones are donating themselves to science... or at least more testing with bracing, frame configurations etc
great test and demo, Ben!!! Learning a lot from your vids!
I like the bamboo. Warm sounding, but knowing what's there in the upper treble on the acrylic, I'd miss that if I went for bamboo in my next iteration. It's difficult isn't it, finding the ideal material? I still favour the XPS stuff. Cheap and cheerful, and if prepared well can give excellent results....
Yes absolutely. I actually have an idea for a multi panel floorstander with a boxed woofer on the bottom that might be better in my room (as there isn't really room for a separate sub to match these). It's something I'll probably look at next year - I'd still like to play with and improve these ones as much as possible.
I'll also have to bite the bullet and give XPS a go - I've resisted it so far because I don't like how it looks, but it does seem to be the go-to.
@@TypicalBenYT I too am not in favour of XPS visually, but we're after the sound in the end...
@@logotrikes My goal is actually to find the best compromise (for me) of sound and aesthetics to be honest. I don't have a dedicated listening room or anything so I want solutions that look good in my living room and that my wife doesn't hate!
@@TypicalBenYT I have very much the same issue Ben. The "wife acceptance factor" tends to steer me along a certain path. She literally hates my DML's but loves how they sound, so for her she has to accept them as they are...
Have you guys tried painting the foam panels the same color as the wall...at least that way they try to blend in instead of being part of a display. And if you do get happy with acrylic panels...have you considered using the opaque acrylic so the placement of exciter, etc isn’t as important?
Let me help you out a bit. The wood panel is denser then the plexi and the denser the material the more it hinders certain frequencies like the midrange which in turn will make the wood panel sound warmer, while the plexi will sound brighter clearer due to more vibrations passing through the material. Also when brightness is reduced the bass frequencies will sound more prominent.
Best materials are ones that are very "LIGHT" but rigid enough to hold its shape and able to bend a bit. This is the reason the two top materials for DML are 1/2-1/4 inch thick EPS and Honeycomb composite as pound for pound they are very light yet rigid enough to hold its shape with some bend ability. DML's produce sound by BENDING the material hence the term Bending wave technology.
If you take off the screws holding the plexi panel it will bend more therefore produce more bass.
If I take the screws off the panel... it will fall down!
In one of my previous videos I do show how much the acrylic flexed - originally I was only going to fix it at four points (corners) to allow it more freedom, but it sagged in the middle due to the length of the panel.
I then experimented with different number of screws and took measurements for each, ending up with a full 10 points of attachment, which is what you see here.
Reducing the number of attachment points didn't seem to increase bass response appreciably in my tests.
CZcams
Correction Ichigo - Bamboo is actually 30 or more percent lighter than acrylic for equal thicknesses. The loss of clarity in ply panels is caused by:
a/the glue between layers damping the resonance.
b/ the interlayer (filler layer) of plywoods is often of a different and softer/weaker wood than the outer skins and often visible gaps are present between strips making up the layer ->..more damping. This aspect could be checked by peeling apart a scrap piece.
c/ not all woods are suited to this acoustic task...maybe spruce laserply would give a better result.
Bamboo is a grass and because it is narrow by nature, the ply is made up of many narrow strips - even more glue.. It's been tried as a tonewood by some with apparently underwhelming results.
It would be interesting to see a comparison of Spruce ply - Another panel Ben??
Basically however - plywood as we commonly know it is not the best material for a DML panel.
Plywood is also not used in top end guitars for a reason.
Have you tried glueing a brace (like an archtop guitar) to make the panel more bass "friendly". I hope this idea is not nonsense, bit I'm fascinated by your patient research method
Hey Clem. Not yet, but I do actually have a pack of spruce guitar braces here for just that purpose! Due to the size of these panels they lacked stiffness so I ended up having to use more attachment points than I intended. My hope is that by using bracing, I'll be able to just attach the panel in the corners (less damping) and even use a thinner (lighter) panel, while retaining overall stiffness. I should be able to start experimenting with that in the next week or so.
And thanks I'm glad you're enjoying the videos!
I thumbs up just because you built something. 👍
Uh… thanks!
Thank you for sharing with us, Ben, very intersting project! What do you think about bass/low freq compensation I was brought up with clssic 2 stereo speakers that reproduce the full spectrum, an audiofile or at least hi fi experience, so I am reluctant to bring 3rd piece, the subwoofer. So, do you think there is a way to construct DML to have enought bass? Or, in your experience, you feel you already have enough bass? Please share your thoughts, thanks!
Hey Jack thanks a lot!
Regarding the bass issue - the general consensus is that to get panels to play lower, they need to be larger (the measurement of the longest dimension is directly related to the lowest frequency it will play). I haven't made a set of very large panels personally though i'd like to give it a go some time. I'd image that you'd need multiple exciters per panel in order to drive the mass of a large panel, and other aspects of performance such as decay might suffer from the increased mass.
Personally, I don't think my panels have enough low end on their own - I don't need slamming bass in my living room but they do lack a sense of weight that you get from a proper full range setup.
My next pair will incorporate a pair of cone woofers in the tower design - in fact I already have the drivers for it! Of course this introduces issues like having to build a crossover (or run active woofers etc) but I think that in practical terms, it's the best solution.
I have been waiting for this . . . Thank you!
No, thank YOU Greg - hope you found it interesting!
Another good and interesting video. As seems to be the norm, it raises more questions. I noticed the pillows in the corners behind the speakers and wondered why they were placed there. This made me think back how Paul Klipsch designed his corner horns to replicate the 16’ pipe on an organ with the last 8’ being along the wall of the corner placement. Can this be used with the DML speakers? Or would it cause more problems by being “out of sync”? Thanks again!
Yes it sure does! I’m aware that the room acoustics is pretty bad at the moment - the big cushions were just a token effort to bass trap the corners a bit. I expect they made next to no difference but it’s all I could do at the time!
Regarding the Klipsches, I can’t picture it, do you know where I can see this online at all?
@@TypicalBenYT I live close to Klipsch’s hometown and used to hangout at his factory so my knowledge is over 50 years old. I remember that he wanted his cornerhorns to replicate the best pipe organs with the longest pipe being sixteen feet long. The last four notes could not be heard by human ears but could be felt by the body. His folded horn design furnished the first eight feet built into the speaker but relied on the reflection off the corners to be the last eight feet. Using two cornerhorns with a LaScala middle speaker is unbelievable sound. Not sure that the successor company still produces these. Paul Klipsch was a unique person and way ahead of his time in speaker design.
Awesome. LYW but more than that I love your approach. Happy xmas!
I am thinking about using foam pipe insulation noodles along the acrylic panel edges and retaining within the frame by providing 1/4 or 1/3 round grooves cut all along the inside of the frame surround. The panel could be popped into the frame and one could experiment with using different lengths of noodles as well as position along each edge. This would provide a semi-rigid panel without screws or grommets.
Can you describe the grommets you used? What was the final location of the attachments?
This sounds like a great way to support the panel if you have the skills to do it well.
My grommets are similar to this (clear ones though of course in my case): www.digikey.com.au/en/products/detail/keystone-electronics/739/106230
hello good to see and hear a new post from you. I wanted to ask if you were playing in mono when you switched between the bamboo and acrylic panels? also the speaker which is in the solid corner (on the left of the system) will make more bass than the panel on the right due to the large opening in the room there. The two panels are very similar overall. I hope that you will have time to test and enjoy your speakers over the Holiday season. Thanks for posting.
Thanks! No it was the regular stereo mix - I guess for accuracy I should find a mono track to use next time, good comment!
The opening on the right is unfortunate isn’t it? I will be placing thick curtains along the left there to absorb as much as possible in to compensate. For now it’s worth noting that due to front/back wave cancellations, the sound is very weak 90 degrees off axis so the problem might not be AS bad as with box speakers.
@@TypicalBenYT Well that's life, our rooms are (for the most part) what they are and we all do the best that we can with what we have. I was able to get a pair of 4 watt 3/4" exciters to experiment with. A friend who has been experimenting with exciters for some time has told me to stay away from the large wattage exciters as the smaller one sound better (he is using 1" pink extruded 2'x'2' rounded corners with pipe insulation around the panel and low wattage 2" exciters). Thanks for posting and I do look forward to your experiments and measurements. Have a very happy Christmas.
Makes wonder what sheet carbon fiber would do. I bet well.
Hey Lou. Yep I’d imagine so. I’ve looked online but have only found quite small sheets before it starts to get expensive. Still… might be worth taking the plunge at some point.
@@TypicalBenYT You know this is the epitome of these experiments .... maybe in the future :)
Ben, what I'm really impressed with is your engineering and construction of the frames. I haven't seen anyone yet on YT go to that much trouble but I like it! Also, wondering why you didn't go with the 2/5ths and 3/5ths placements of the exciters. Was there a reason you found to do it in your way?
Thank you! In regards to the exciter positioning, I tested various positions including the 2/5 one, but the measurements were so close that I went with the centered position out of preference (aesthetics).
@@TypicalBenYT Thanks Ben!
i'm building a set of "bookshelf" sized panel using 2.5mm cedar ply. sounds great but i expect it to change slightly when i apply a gunstock oil finish
Hey Paul, that's interesting how big are the panels? I don't know much about oils etc but I did coat my bamboo panels with danish oil and actually didn't notice a great difference in sound.
@@TypicalBenYT300 high x 200 wide
oils will change the resonance of the board, hopefully for the better but here's hoping.
next project i think will be home made electrostatics
I would think the cedar will absorb much more of the oils than the bamboo. Bamboo historically resists absorbing...that was one of the challenges of early bamboo flooring...could not achieve enough variables or depth of stains or oils. They only had two colors originally...and they weren’t that different.
I've been playing around with some coroplast board like signs are made out of. Pretty impressive but wanting to try some carbon fiber panels next. My problem is making designing something to be portable in a way setup outside of my car etc when I want.
That’s pretty cool - is it for camping or something? Carbon fibre would be an interesting one for sure…
@@TypicalBenYT yes sir and I do ultra running event whereas my car could sometimes be my aid station on a loop so I'll have music playing and my diy amp project box power a sub too. I should have put together a video.
If anybody makes a small passive equalizer, so that audio in for each speaker goes into the equalizer, then into the exciter, couldn't just about any layout be tuned and tweaked into a fairly flat response? Then the only consideration would be resonances and general "timbre".
Well, my ear wasn't that far off: I guessed the acrylic had a harsh bump around 10k, and it turned out being 9k. The dropoff above that is definitely contributing to how sharply that bump comes across. That would be the first thing I EQ down if it were up to me.
With the bamboo being warmer and more effective at the low end, I'm curious to know what kind of sound you'd get by crossing over at, say, 500 Hz and putting all the low end toward the bamboo panel with some EQing to flatten it out. Obviously, that would mean four panels for a full-range stereo setup, and you would probably still need subs, but my guess is it would help mitigate that low-end rolloff.
Hey Josh - yeah not far off at all!
That spike in the acrylics didn’t really appear in my previous testing of the acrylics, instead I got a relatively smooth increase all the way to 20k. I used different exciters and the panel dimensions were slightly different (not that I think that would matter too much in the HF) so it’s interesting to find it now.
Of course I’m not running any EQ for these tests, but my plan is to tweak the soeakers, tweak the room, then tweak with EQ.
The complication with having 4 panels would be space of course, especially in my room. They also need to be quite large panels to hit lower frequencies.
My current thinking on how I’ll manage this is to get something like a 10” woofer and cross to the panel above 250hz. Previously I’ve used a standalone sub crossed at 150hz and while it sounds ok, it does still leave a bit of a dip in that 150 -250range. What do you think?
@@TypicalBenYT Yeah, from what I've seen in response curves, it always seems like the DML's drop off around 300 Hz or so. I haven't actually built any yet, myself, so I have no personal experience at the moment, but my experience with subs overall has probably been similar to your own judging from that 150 Hz crossover you mentioned. That's what gave me the thought of leveling them out a bit with EQ if possible. I dunno if it would work as well as we'd like. Like you say, though, it's also very much a matter of space, and building one GIANT panel as a subwoofer might as well involve just affixing a driver to the inside of the drywall!
@@JoshColletta agreed, I’ve always assumed that any good sounding panel setup I build would be 2.1. A pair of like 9’ panels with enough exciters might be acceptable but I don’t really want big tall panels either as this is our TV/living room etc
A DML sub is interesting if it can be wall mounted and disguised as art or decoration, but I wonder if it’s worth the hassle vs just adding conventional woofers!
Acrylic sounds better it has a good stereo balance.. the bamboo sounds a bit mono-muffled
HEY USE (2) exciters on the acrylic boards & wire them In series!!!!
Good timing. It just so happens that I have another pair of drivers right here and planning to try just that tomorrow!
…how…did you build your frame?, & how thick is is the acrylic baffle surface?
I actually ordered the frame made to measure from an online shop doing canvas art. I don’t have the woodworking skills to build them to the same standard (and square). The acrylic used is 4.5mm, panel dimensions 400x1100mm.
There’s more info on the construction in my video Acrylic Build 1 if you’re interested.
@@TypicalBenYT Yes …very interested!
Very good video but we could just try using different types of wood also because each type of wood is rated hard to soft. Each one has a different sounds.
Yes absolutely. Before I build any more speakers I want to experiment with a bunch of different combinations of materials, construction and drivers.
TRY using 2-3 exciters on each acrylic panel… wire each set in series
Have you tried thin carbon fiber panels? I'm about to do that next, but if you've done it already I'll pass :-)
No not yet. Carbon is on my 'to do' list but... well a lot of things are on that list! Would love to see your results if you try it though.
@@TypicalBenYT I've got six configurations I'll be trying and posting, including carbon fibre panels, acrylic, fabric, foam in a brand new configuration I have not seen anyone do, luan (like plywood but stronger and finer), lexan and a combination of a few. Results are so different I know I won't be satisfied unless I do it myself and with my own ears. I'm sticking to the manufacturer's suggestion, though and using panels in a 4 x 5 ratio, otherwise anything goes.
@@newunderthesun7353 super interesting, looking forward to seeing your results!
I also have built a dml panel from 1/4 acrylic and find it too harsh sounding. Which i attribute to placing near the wall and acrylic thickness. I wish to try the pink panels instead next
Hey Max, Interesting, I don't find these to be harsh sounding, though I guess there are a lot of variables at play.
Once reason I like building these as 'floor standing' is that you can bring them out from the wall.
Those hanging from the wall could be hung on a frame with differing length arms so they can be placed at the ideal spacing from the wall so they take advantage of the actual variables of the room.
Will you please recommend your best design to date - from Jan 5 2022
I'm actually really liking the sound of these rigid acrylic ones (construction is outlined in my previous videos). I liked the bamboo but having A/B tested them, I prefer these.
Balsa wood?
Do you have the thickness for Acrylic panels and a good suggestion for a source ??
Hey Sean - I'm using 4.5mm panels on these. I am keen to try thinner acrylic (maybe 3mm) with bracing as well. I bought these from 'Plastics Online' near the Gold Coast I could pick up, but there are a bunch of online shops providing acrylic cut to size as well.
Excellent video, thx. Could you please show harmonic distortion.
Hi, yes I’ll see what I can do. I’m not very familiar with this measurement but will give it a go.
The panel material will have its own sound.
I would expect the material with the least of its own sound would be the best candidate. Boring but more true.
You are not using ⅖ and ⅗ of the dimensions to cancel out most of distace related resonances.
Snazzy 🎉
🙋👍❤️