-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Any interest in an ASR community speaker project?
- Thread starter 617
- Start date
OP
- Thread Starter
- #42
Here's some data.
I have simulated four woofers' maximum SPL at 2M distance. They require different amounts of power to get there, and the stereo effect and the room changes things somewhat, but this is a good guide.
The drivers are the Dayton RS125p, 150p, 180p and 225p, along with the Tang Band w4 1720. The Tang band is a 125mm diameter driver with a huge magnet and some ridiculous xmax, so I thought it might be an interesting comparison with the RS125.
Let's look at a specific case where we put the small drivers in vented boxes and the big drivers in sealed boxes and give them as much power as they can take before exceeding xmax at or above 60hz.
The results are:
125p= 92.9 db with 20W
w4= 93.4 db with 24W
150p= 94.9 db with 20W
180p= 100 db with 60W
225p= 101.5 db with 78W
Surprisingly, going from the 125mm to the 150mm driver doesn't give you a ton more output, but it does take less amplification to get there. Going to the RS150 (aluminum) gets you about 1db more output, which is tiny. Playing with different alignments does change things somewhat, but all the tiny drivers need to be vented to get the bass you need.
I looked into the use of small pro midrange drivers and they're not a great solution - they have very high Fs, so getting output below 100hz is a challenge, and they tend not to have very well behaved breakup.
So how much is enough? I would say 90 db at 2M is a decent enough target for a small speaker, but more output is not a bad thing, and making a small speaker which can go unusually loud due to a high pass filter seems like an unique proposition.
I have simulated four woofers' maximum SPL at 2M distance. They require different amounts of power to get there, and the stereo effect and the room changes things somewhat, but this is a good guide.
The drivers are the Dayton RS125p, 150p, 180p and 225p, along with the Tang Band w4 1720. The Tang band is a 125mm diameter driver with a huge magnet and some ridiculous xmax, so I thought it might be an interesting comparison with the RS125.
Let's look at a specific case where we put the small drivers in vented boxes and the big drivers in sealed boxes and give them as much power as they can take before exceeding xmax at or above 60hz.
The results are:
125p= 92.9 db with 20W
w4= 93.4 db with 24W
150p= 94.9 db with 20W
180p= 100 db with 60W
225p= 101.5 db with 78W
Surprisingly, going from the 125mm to the 150mm driver doesn't give you a ton more output, but it does take less amplification to get there. Going to the RS150 (aluminum) gets you about 1db more output, which is tiny. Playing with different alignments does change things somewhat, but all the tiny drivers need to be vented to get the bass you need.
I looked into the use of small pro midrange drivers and they're not a great solution - they have very high Fs, so getting output below 100hz is a challenge, and they tend not to have very well behaved breakup.
So how much is enough? I would say 90 db at 2M is a decent enough target for a small speaker, but more output is not a bad thing, and making a small speaker which can go unusually loud due to a high pass filter seems like an unique proposition.
Please read again, I wrote low cost/benefit ratio! There are hundreds of low cost 2-way desings in the web, we can't do any better at ASR. The cost includes dsp and amplifiers!
I built my first diy speakers in '70s - a 3-way kit and I still have them! Then in this millenium some 2-ways from respected designers, but they didn't sound or measure any better than my 500€/pair commercials. With minidsp and Hypex I could design unique 3 and 4-way speakers that are worlds better than what one can buy for 3000€ a piece!
Last edited:
Speedskater
Major Contributor
Speedskater said:
To even think about a passive crossover is falling into the old audiophile mix & match trap.
Any modern loudspeaker design deserves an active DSP crossover.
If you consider the total cost of amps and speakers, good luck with your passive DIY project.
To even think about a passive crossover is falling into the old audiophile mix & match trap.
Any modern loudspeaker design deserves an active DSP crossover.
Both Mackie and Behringer make very nice powered, active speakers in the $300 (range) each.
If you consider the total cost of amps and speakers, good luck with your passive DIY project.
If we're talking about a budget build, many people usually already have some sort of amplification in the form of an AVR or stereo receiver and if they don't used ones are often available dirt cheap at thrift stores so I don't think we really need to worry about additional cost.
I guess I'm missing whatever point you're trying to make. Sure, if you have absolutely nothing and are starting from scratch with only $300 to spend, there is no way DIY will ever compete with the economies of scale available to large-scale manufacturers. Even more so if you buy used powered monitors. But that isn't really the context of the discussion.
If you're looking at designing something that potentially has broad appeal - which is implied by the idea of a "community project" - you have to be able to satisfy a broad range of applications. Some folks are 2-channel only, some are HT, some need sub integration, some don't etc. As was covered previously, multichannel DSP systems available to the DIYer are fairly limited, and all of them have some level of constraint. What works well for a dedicated 2-channel music system might not work for a living room or HT system. Realistically, none aside from the Hypex Fusion amps or maybe the Minidsp SHD are likely to seamlessly slide into a 'conventional' system that has to be usable by family members - and both of those are expensive.
So, in that context if you're trying to propose something that is both inexpensive enough AND generic enough to attract somewhat broad interest, a passive design is the only thing that makes much sense. However, having seen this type of idea floated in other forums, it frequently fizzles out since it's tough to get quorum on a single design and so either it just fades or else everyone just works on their own design separately. You can already see that happening here - everyone throws out their pet ideas for the project, and they're all over the map.
So, as a personal project you're certainly free to go in whatever direction you like, and I wouldn't try to dissuade anyone from heading down the DSP route. I've been playing around with active DSP systems since the 90's, so I certainly feel comfortable saying I understand the benefits and trade-offs, but can also say that building a DSP system that 'just works' the way conventional gear does is much tougher than it might seem at first. This is why for my main system I've decided to design around the SHD and pursue a hybrid passive/active design using a passive M/T xover with DSP active between that and the bass drivers. This type of system puts you into a $2k-$2.5k system cost though, and so isn't something that would ever support any type of broad interest/engagement.
OP
- Thread Starter
- #48
As the person who began this thread, I'm not that interested in this debate; seems like people here don't understand the value of making something. I'm not trying to engineer someone's end game speakers or compete with JBL or Neumann, I'm trying to enable technically minded people to feel empowered to make something which achieves a high degree of sound quality for fun, and to help pass the time during what I'm sure will be a fucked up summer for a lot of people.
The goal here is to design something that those of us who like making things can work on together and send around across the country or world, to build community spirit, and to show everyone that ASR is the best audio community around.
If you're not interested in engineering a project which is small, fun and approachable, please do not derail the thread any more.
If you are interested in contributing something other than opinions, let me know what you like doing and are good at. I know for a fact that we have the tools between us to come up with a really interesting and refined design. Can you cut a baffle? Can you make a PCB? Can you donate parts? Can you make accurate measurements? Etc.
Please contribute opinions about topology and price point and, most importantly, what you can give to the effort.
The goal here is to design something that those of us who like making things can work on together and send around across the country or world, to build community spirit, and to show everyone that ASR is the best audio community around.
If you're not interested in engineering a project which is small, fun and approachable, please do not derail the thread any more.
If you are interested in contributing something other than opinions, let me know what you like doing and are good at. I know for a fact that we have the tools between us to come up with a really interesting and refined design. Can you cut a baffle? Can you make a PCB? Can you donate parts? Can you make accurate measurements? Etc.
Please contribute opinions about topology and price point and, most importantly, what you can give to the effort.
OP
- Thread Starter
- #49
Yes, as a design challenge, a small but very high output speaker going down to 70hz would be a really useful speaker and a lot of fun to build.
Hi all. Long-time lurker here. This is my first post. I have been diligently going through the multichannel music and speaker review threads, in preparation for getting back into DIY speaker design once I retire and move (not too long a wait, I hope).
I noted the preference many have for mch for music, even upmixed. (Auro 3D seems to have the vote for most effective upmixer.)
I also noted the reduced strength in preference between speaker models in mch compared to mono. That seems to imply that one does not need top-of-the-line speakers for mch. The question arises, what is the minimum requirement for excellent sound in mch?
The relevance to this thread is, what if ASR targeted minimum satisfactory mch as the use case for the community speaker project? In my gedanken experiment, I'm using a Denon X4500H with Auro 3D and Audessey XT32 (with the app). This simplifies crossover requirements to controlling acoustic slope and phase, and possibly driver resonances and impedance; frequency response tailoring can be left to EQ. My thoughts so far:
1) Absolute minimum - Vifa TC9. Will beam in high treble, needs a crossover to the subs around 200 Hz and will still be volume limited. Might work in small rooms with a fixed sweet spot. A foam core cabinet might work here - easy to hang on the walls. I lack the mechanical engineering chops to design a quiet foam core box, though.
1a) TC9 plus Peerless DX20. Crossover could be 1 cap at around 20K; this would improve the highs and correct the beaming, but at the cost of comb filtering. Requires modeling or testing. Still output-level limited.
2) Dayton DSA175. Proven low distortion and benign frequency response per Audio Express Test Bench, better than many ASR recommended speakers if the measurement protocols are comparable. (Note: I am pretty sure the distortion levels are at 94dB, not 104dB as stated in the Test bench article; probably a typo. 104 dB is the level they use for pro drivers and 94 dB for hifi drivers.) If the DSA175 is paired with a Peerless DA25TX00-08, the crossover could be in the range of 1300 Hz LR4, so the directivity should be smooth and wide. No waveguide needed because the DSA175 should not beam significantly at 1300 Hz. Too bad about the overly large faceplate on the DA25, but one cannot have everything. My suspicion is this is close to a minimum setup for high-end mch. A 3-way with a BMR 3" and a small tweeter would also work at the cost of additional crossover complexity (somewhat minimized by doing frequency response corrections in Audessey).
I read somewhere (cannot recall the source) that a desired characteristic for mch music is for the directivity to have a fairly sharp cutoff as one moves out of the listening area. If so, this looks interesting to use with the DSA175. Radian LT3 with waveguide:
https://radianaudio.com/products/lt3-planar-ribbon/?variant=20538618478654
Thoughts?
URLs:
TC9:
https://www.diymobileaudio.com/threads/vifa-tc9-3-midrange-widebander-review-measurements.170856/
https://www.diyaudio.com/forums/full-range/268626-vifa-tc9fd18-08-bang-buck.html
Peerless DX20:
https://hificompass.com/en/speakers/measurements/peerless/peerless-dx20bf00-04
DSA175:
https://audioxpress.com/article/test-bench-dayton-audio-dsa175-designer-series-6-5-woofer
DA25 polars in post #171:
http://www.htguide.com/forum/showthread.php?44456-Ardent-D-aka-Kurosawa-Jr-Winter-Camp/page5
I noted the preference many have for mch for music, even upmixed. (Auro 3D seems to have the vote for most effective upmixer.)
I also noted the reduced strength in preference between speaker models in mch compared to mono. That seems to imply that one does not need top-of-the-line speakers for mch. The question arises, what is the minimum requirement for excellent sound in mch?
The relevance to this thread is, what if ASR targeted minimum satisfactory mch as the use case for the community speaker project? In my gedanken experiment, I'm using a Denon X4500H with Auro 3D and Audessey XT32 (with the app). This simplifies crossover requirements to controlling acoustic slope and phase, and possibly driver resonances and impedance; frequency response tailoring can be left to EQ. My thoughts so far:
1) Absolute minimum - Vifa TC9. Will beam in high treble, needs a crossover to the subs around 200 Hz and will still be volume limited. Might work in small rooms with a fixed sweet spot. A foam core cabinet might work here - easy to hang on the walls. I lack the mechanical engineering chops to design a quiet foam core box, though.
1a) TC9 plus Peerless DX20. Crossover could be 1 cap at around 20K; this would improve the highs and correct the beaming, but at the cost of comb filtering. Requires modeling or testing. Still output-level limited.
2) Dayton DSA175. Proven low distortion and benign frequency response per Audio Express Test Bench, better than many ASR recommended speakers if the measurement protocols are comparable. (Note: I am pretty sure the distortion levels are at 94dB, not 104dB as stated in the Test bench article; probably a typo. 104 dB is the level they use for pro drivers and 94 dB for hifi drivers.) If the DSA175 is paired with a Peerless DA25TX00-08, the crossover could be in the range of 1300 Hz LR4, so the directivity should be smooth and wide. No waveguide needed because the DSA175 should not beam significantly at 1300 Hz. Too bad about the overly large faceplate on the DA25, but one cannot have everything. My suspicion is this is close to a minimum setup for high-end mch. A 3-way with a BMR 3" and a small tweeter would also work at the cost of additional crossover complexity (somewhat minimized by doing frequency response corrections in Audessey).
I read somewhere (cannot recall the source) that a desired characteristic for mch music is for the directivity to have a fairly sharp cutoff as one moves out of the listening area. If so, this looks interesting to use with the DSA175. Radian LT3 with waveguide:
https://radianaudio.com/products/lt3-planar-ribbon/?variant=20538618478654
Thoughts?
URLs:
TC9:
https://www.diymobileaudio.com/threads/vifa-tc9-3-midrange-widebander-review-measurements.170856/
https://www.diyaudio.com/forums/full-range/268626-vifa-tc9fd18-08-bang-buck.html
Peerless DX20:
https://hificompass.com/en/speakers/measurements/peerless/peerless-dx20bf00-04
DSA175:
https://audioxpress.com/article/test-bench-dayton-audio-dsa175-designer-series-6-5-woofer
DA25 polars in post #171:
http://www.htguide.com/forum/showthread.php?44456-Ardent-D-aka-Kurosawa-Jr-Winter-Camp/page5
- Joined
- Apr 24, 2019
- Messages
- 1,501
- Likes
- 2,822
TimW
Major Contributor
I don't use room correction outside of bass frequencies. I don't want to put effort into a speaker that doesn't sound good out of the gate. And I'm sure we all want to see Amir measure this thing and be amazed by the excellent results.
TimW
Major Contributor
I would like this speaker to have wide dispersion. So a waveguide tweeter does not interest me as much. We could try to get decent directivity matching in a two way by using an extremely robust tweeter, crossing over very low, and maybe modifying the baffle shape to affect tweeter directivity. I have eyed the Peerless DA25TX00-08 for a while and it looks promising to me. I have seen it paired with the Dayton Audio Esoteric series woofers which come in 5 1/2" and 7" sizes. Or we could just go with a 3 way.
A digital input only fully active bookshelf with a external subwoofer and some kind of onboard room correction would be a ASR speaker , ambitious maybe but we have all the skill sets in house to do it.
"Some kind of room correction" which should support active bookshelves and external subwofers should support 4 channel DAC, and something like that I don't see on the market. All I can see is Chinese manufacturers competing to produce 2 channel DACs with slightly better SINAD, which is simply a futile battle, but they don't seem to get it.
I was going to suggest the DA25Tx paired with the Anarchy woofer available via diysoundgroup. I think this has the opportunity to be close to a universal 'do everything' project. The Anarchy can be a bass monster if you are OK with a larger cabinet, but could also be used sealed in a small-ish cabinet for use as surrounds or with subs. The DA25Tx would allow an xover in the 1.6kHz range, helping mitigate the ~4kHz breakup peak (which is pretty mild for an AL driver) . Moderate cost - drivers would be about $250/pair.
Depending on how my personal experiments for my project go, I'm hoping to toss in a waveguide option for the DA25Tx, but that may not pan out.
As others have said, there are a bazillion 2-ways out there both commercial and DIY, so there's really not much new under the sun. This combo offers some interesting options, and doesn't break the bank.
I agree, you want 2 subs and bookshelves all driven digitally by a four channel DAC ."Some kind of room correction" which should support active bookshelves and external subwofers should support 4 channel DAC, and something like that I don't see on the market. All I can see is Chinese manufacturers competing to produce 2 channel DACs with slightly better SINAD, which is simply a futile battle, but they don't seem to get it.
The problem I'm seeing is that DAC manufacturers are not supporting this point of view. For that reason I'm thinking of integrating the sub I'm going to buy (hopefully soon) via my AVR processor which has decent specs and all the features I need to do that.
This idea may be worth splitting out into a separate thread.
If you're thinking the typical 'all in one' studio monitor form factor, this is pretty tough to do without custom electronics, which is an adventure all on it's own. Might be feasible if you're ok with a separate box for the electronics and amps (e.g. like the NHT Xd system did) since you could probably use off-the shelf bits, although still not without challenges. To be truly ASR you'd probably want to make it a riff on the DXT-Mon or other design with faultless directivity, but of course the DSP system is pretty generic and can be used for a whole host of projects.
The room correction piece would also be a bit of a trick if you're thinking 'automated' rather than 'use REW to export biquads into the system'. The only semi-automated open-source setup I know of is DRC-FIR which is old, cryptic and hardly a one-shot process. It's also inherently an FIR generator, which imposes constraints on the DSP architecture.
Off the top of my head, the 2 available ways to approach this would be a MiniDSP Nano Digi plus a few ToneBoards (room correction only via REW biquad export), or (more interestingly) maybe a Raspberry Pi 4 or embedded x86 board running a stripped down single-purpose linux along with a good USB audio interface (more work, more flexibility).
The work here is more in defining and implementing the DSP system, UI, remote control and measurement and filter design workflow rather than designing a single speaker, though. Whether it ultimately provides any value compared to the Hypex Fusion amps would be the big question.
That's where it's shite you have to use a crappy AVR , we can do better .
