• 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?

dwkdnvr

Senior Member
Joined
Nov 2, 2018
Messages
418
Likes
698
FWIW on the DSP amp front someone pointed these out the other day http://www.3e-audio.com/finished-amp/tpa32xx-all-in-one-amplifier/ . They're TPA3255/3251 amps with integrated ADAU1701. The TPA3251 in particular is a very good performer with the 3255 not far behind, meaning that the ADAU1701 itself is probably the limiting factor. They're just coming to market, but might provide an alternative to the Dayton amps for someone looking for an off-the-shelf option rather than a project. Not as attractive for building into a clean powered monitor style setup, though.

Looks like the same SigmaStudio programming workflow as the Dayton DSP amps, at least at first glance.
 
  • Like
Reactions: 617

HammerSandwich

Major Contributor
Forum Donor
Joined
Nov 22, 2018
Messages
1,137
Likes
1,497
If you'd like to do some more detailed woofer comparisons that would be a good idea. What data about the amp performance would you like to see?
Nothing. I just checked the TPA3116 datasheet, which essentially matches Dayton's specs. Within a dB or so, it breaks down like this...

On 19V, BTL stereo can make 2x 40W in 4Ω, so 12+Vrms. This is 13dBW, so 87dB speakers would just hit 100dB. Jump to a 24V PSU, and you get 60W | 16Vrms | 15dBW.

Switching to PBTL mono is not much help, because the rated loads drop to 2-3Ω. You're still looking at 13-16Vrms, and you've doubled your board cost. Maybe there's a really great 2Ω woofer out there? :D

Seems that 105dB peaks would require serious compromises elsewhere, so it's probably more realistic to target 100dB peaks with these Dayton amps but design the speakers to handle more.

And @dwkdnvr's TPA3255 suggestion could work.
 

somebodyelse

Major Contributor
Joined
Dec 5, 2018
Messages
3,682
Likes
2,962
Still waiting on details of how much of the DSP the 3E amp exposes - it's not clear how you'd tie the amps together for a 3 way crossover. 3e do have a separate DSP board that's better buffered than the separate Dayton/Sure/Wondom one though. The amp looks like it's got the same box header for the programming interface, but there's no obvious sign of headers for i2s for extra channels etc.
 
OP
617

617

Major Contributor
Forum Donor
Joined
Mar 18, 2019
Messages
2,404
Likes
5,296
Location
Somerville, MA
ted these out the other day http://www.3e-audio.com/finished-amp/tpa32xx-all-in-one-amplifier/ . They're TPA3255/3251 amps with integrated ADAU1701. The TPA3251 in particular is a very good performer with the 3255 not far behind, meaning that the ADAU1701 itself is probably the limiting factor. They're just coming to market, but might provide an alternative to the Dayton amps for someone looking for an off-the-shelf option rather than a project. Not as attractive for building into a clean powered monitor style setup, though.
Nothing. I just checked the TPA3116 datasheet, which essentially matches Dayton's specs. Within a dB or so, it breaks down like this...

On 19V, BTL stereo can make 2x 40W in 4Ω, so 12+Vrms. This is 13dBW, so 87dB speakers would just hit 100dB. Jump to a 24V PSU, and you get 60W | 16Vrms | 15dBW.

Switching to PBTL mono is not much help, because the rated loads drop to 2-3Ω. You're still looking at 13-16Vrms, and you've doubled your board cost. Maybe there's a really great 2Ω woofer out there? :D

Seems that 105dB peaks would require serious compromises elsewhere, so it's probably more realistic to target 100dB peaks with these Dayton amps but design the speakers to handle more.

And @dwkdnvr's TPA3255 suggestion could work.

There are a myriad of amps which would work. I would personally run an icepower 200W module and a 50x50 module if I wanted to upgrade. But the amps aren't really that important; I'm curious what the Daytons can actually do. The filter programming should be easy to apply to a variety of systems - peak here, notch there, LR2/4 crossovers, gain, maybe some delay, nothing fancy.

I agree that the speaker drivers should not be downgraded to match the amp's capabilities, but conversely I also don't think spending $200 dollars on a woofer which really needs a 200 dollar amp to sing makes much sense. Even if the Dayton amps are extremely anemic, with our current woofer selections, they will play quite loud if high passed at 80hz or whatever.
 

HammerSandwich

Major Contributor
Forum Donor
Joined
Nov 22, 2018
Messages
1,137
Likes
1,497
If you'd like to do some more detailed woofer comparisons that would be a good idea.
I figure I'll start with sealed boxes at 16Vrms, per earlier.

Does a 30L limit sound appropriate for this project? Always using Qtc 0.707 doesn't make a lot of sense, IMO, if box size varies 3:1. We all know that bigger boxes will give more bass output for a given amp (at least until Xmax), so there will be some room to adjust later. But let's start by looking at the upper limits for these amps in an attenuate-only DSP setup.

Make sense to everyone?
 
OP
617

617

Major Contributor
Forum Donor
Joined
Mar 18, 2019
Messages
2,404
Likes
5,296
Location
Somerville, MA
I figure I'll start with sealed boxes at 16Vrms, per earlier.

Does a 30L limit sound appropriate for this project? Always using Qtc 0.707 doesn't make a lot of sense, IMO, if box size varies 3:1. We all know that bigger boxes will give more bass output for a given amp (at least until Xmax), so there will be some room to adjust later. But let's start by looking at the upper limits for these amps in an attenuate-only DSP setup.

Make sense to everyone?
Do you have the spreadsheet of the woofers on dropbox?
I didn't simulate using a qtc since the active implementation makes it a bit irrelevant.

I think what I did was look at the xmax limited spl graph and change the simulated volume of each woofer for max output at 40 hz or so. This resulted in a ranged of volumes from .4 to 1.4 cubic feet (11L-23L). The outliers on the large side were things like the scan discovery and sb23 aluminum models which I don't think will do well in a tiny box.

The SPL at 30hz then varied between 89 and 95 db at 1m (single speaker, half space). I simulated vented and 10" woofers differently.

The best 8" under my price limit was the sb23MFCL which is a poly cone subwoofer, but it needs 150W to do its thing and costs 130 USD. The RS225p does only 2db worse but costs 64 USD. The Tang Band unit is an odd duck and worth looking at too.
 

HammerSandwich

Major Contributor
Forum Donor
Joined
Nov 22, 2018
Messages
1,137
Likes
1,497
I was looking at your summary charts in posts 123, 155 & 180. At least some of the drivers are already in Vcad's library, but I can pull datasheets.

So, shoot for max 40Hz & we'll see where Vb falls?
 
OP
617

617

Major Contributor
Forum Donor
Joined
Mar 18, 2019
Messages
2,404
Likes
5,296
Location
Somerville, MA
I was looking at your summary charts in posts 123, 155 & 180. At least some of the drivers are already in Vcad's library, but I can pull datasheets.

So, shoot for max 40Hz & we'll see where Vb falls?

Yes, and really the smaller the better. 30L is a good upper limit; it's actually hard to design a 3 way speaker with much less volume than that.

Try to figure out the assumptions VCad makes in the SPL chart - half space vs free field, 1M or 2M, etc.
 

dwkdnvr

Senior Member
Joined
Nov 2, 2018
Messages
418
Likes
698
I like the 30L range, if only because I have a set of the Parts Express 1 cu ft MTM cabs that I was planning to use as part of the prototyping. I've been using them as sealed woofer cabs with the RSS210, so they fit the direction of the project. The RS225 just barely fits on the baffle, and hopefully if a different 8" is brought into the mix it's not bigger than that. MT prototyping will initially be with some of the 0.25 cu ft cabs - these are narrower than the target design cabinet, but hopefully they'll work to get through the driver eval process.

I'm hoping to do a bit of evaluation of bass expectations with the RSS210s, and then switch to the RS225's and see whether there is really a difference between subwoofer and woofer. I suspect that at a 200-300 xover there won't be much difference, but it'll be interesting to see.

Complete fail on trying to print the waveguide today. The only sealed spool of filament I had doesn't properly fit my spool holder, and of course the filament wound around the holder and knotted itself, jamming the print head as the extruder pulled it over to the side of the printer. Hopefully no damage done.
 

HammerSandwich

Major Contributor
Forum Donor
Joined
Nov 22, 2018
Messages
1,137
Likes
1,497
Before comparing drivers, let's consider what box size affects in active speakers with DSP & a known amplifier.

Because we're plotting SPL at the amp's limit, we can only reduce level in DSP. For example, if we want to boost 40Hz, we need to reduce SPL everywhere else. That can give more relative 40Hz, but not increase actual SPL. We're already at 0dBFS, and this amp does NOT go to 11. We're simply defining the boundary of SPL versus frequency at this stage.

As you make a box smaller, the air trapped inside becomes stiffer. This requires more power for a given cone excursion, meaning bigger boxes can produce more SPL. In addition, as the driver moves, small boxes experience relatively more change in volume. This is non-linear. Vituixcad's default compression limit is 2%, so I'll highlight where that's a factor.

Example: the SB23MFCL45-4.

At Qtc 0.707, it models into a Vb of only 9.2L. But "max comp" must be increased to 3.9%, because the driver's 252cc Vd is a big portion of 9L. And the stiff air prevents the driver from reaching Xmax. Finally, note the big impedance peak at 60Hz, which indicates the box tuning, Fb. You can see that this has an inverse relationship with the power curve, because high impedance reduces the current drawn for our fixed input voltage. IOW, box tuning controls where the woofer operates most efficiently.

Increase Vb to 16.8L, and Qtc falls to 0.577. Tuning drops to 48Hz, and bass increases usefully by ~2dB below 50Hz. Because the area around 100Hz is now lower-impedance, it loses 0.5dB or so. (Higher up, SPL remains at 104dB, because it's too far from the resonance to be affected.) Compression drops below 2.1%, which is a big change.

Finally, let's jump up to 27.6L & down to Qtc 0.5. We gain 1dB below the new Fb of 41Hz, losing another small fraction around 100Hz. And we finally hit Xmax, at 23Hz. Compression falls below 1.3%.

I hope this highlights some of the tradeoffs involved. It's definitely a more nuanced situation than simply looking at F3, F6 & max SPL.

My plan for the driver comparisons is to provide an SPL graph along with brief notes on a few of the other parameters.

EDIT: SPL plots are 2pi @ 1M.

SB23MFCL versus box size.png
 
OP
617

617

Major Contributor
Forum Donor
Joined
Mar 18, 2019
Messages
2,404
Likes
5,296
Location
Somerville, MA
Before comparing drivers, let's consider what box size affects in active speakers with DSP & a known amplifier.

Because we're plotting SPL at the amp's limit, we can only reduce level in DSP. For example, if we want to boost 40Hz, we need to reduce SPL everywhere else. That can give more relative 40Hz, but not increase actual SPL. We're already at 0dBFS, and this amp does NOT go to 11. We're simply defining the boundary of SPL versus frequency at this stage.

As you make a box smaller, the air trapped inside becomes stiffer. This requires more power for a given cone excursion, meaning bigger boxes can produce more SPL. In addition, as the driver moves, small boxes experience relatively more change in volume. This is non-linear. Vituixcad's default compression limit is 2%, so I'll highlight where that's a factor.

Example: the SB23MFCL45-4.

At Qtc 0.707, it models into a Vb of only 9.2L. But "max comp" must be increased to 3.9%, because the driver's 252cc Vd is a big portion of 9L. And the stiff air prevents the driver from reaching Xmax. Finally, note the big impedance peak at 60Hz, which indicates the box tuning, Fb. You can see that this has an inverse relationship with the power curve, because high impedance reduces the current drawn for our fixed input voltage. IOW, box tuning controls where the woofer operates most efficiently.

Increase Vb to 16.8L, and Qtc falls to 0.577. Tuning drops to 48Hz, and bass increases usefully by ~2dB below 50Hz. Because the area around 100Hz is now lower-impedance, it loses 0.5dB or so. (Higher up, SPL remains at 104dB, because it's too far from the resonance to be affected.) Compression drops below 2.1%, which is a big change.

Finally, let's jump up to 27.6L & down to Qtc 0.5. We gain 1dB below the new Fb of 41Hz, losing another small fraction around 100Hz. And we finally hit Xmax, at 23Hz. Compression falls below 1.3%.

I hope this highlights some of the tradeoffs involved. It's definitely a more nuanced situation than simply looking at F3, F6 & max SPL.

My plan for the driver comparisons is to provide an SPL graph along with brief notes on a few of the other parameters.

EDIT: SPL plots are 2pi @ 1M.

View attachment 67938
This is a much more sophisticated treatment than mine. Thanks for doing this. I look forward to seeing what you come up with; I think we're on the same page as to overall size and cost restraints buy it will be interesting if certain woofers perform better after taking these parameters into consideration.
 

DeruDog

Member
Forum Donor
Joined
May 13, 2020
Messages
68
Likes
65
Location
Philadelphia, PA, USA
I'm not quite sure what you're advocating, but a fully passive design which just does the crossovers and makes heavy use of global EQ could be an interesting option. I'm designing this system under the assumption that the easiest, most appealing and cheapest way to create this speaker would be using these Dayton DSP amps (testing in progress), but of course there are a myriad of other options, such as outboard amps and dsp, minidsp units, and a whole host of amps. If a hybrid passive/active design is developed (active between woofer/mid, passive between mid/hf) even more options become available.

I have been looking at the Hypex DSP amps, both 3-way and 2-way. They seem to be a little cheaper, and the amplifier modules have a good reputation. Following this thread with great interest.
 

GelbeMusik

Senior Member
Joined
Apr 19, 2020
Messages
445
Likes
290
Incredibly impressive projects! I was blown away by his outdoor measurement rig that combined an in-ground infinite baffle with a rotating mic frame.

I found some points to be discussed in these reports. Anyway, he is after strong vertical directivity, which I'm not.

Thnaks to remind me of him.
 

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
3,518
Likes
7,030
Location
Stow, Ohio USA
It occurs to me that we have a small but diverse group of experts here, and it might be a fun way to 'give back' to the community as a whole if we pooled our resources and designed a really good speaker.

What I have in mind is a speaker (or several) which is small enough that a prototype can be shipped around at reasonable cost, which can be built with easy-to-source drivers, perhaps ultimately being measured and verified on the NFS if Amir is amenable.

The first decision to be made would be what price point we are targeting -
  • Very low pricepoint - a speaker using for example Dayton classic tier drivers - not great in terms of distortion, but very inexpensive. This would be an interesting exercise because as we have seen from recent reviews, cheap drivers can be used skillfully to great affect. This also has the advantage of making the project more attractive to builders.
  • Low pricepoint - think Dayton or Peerless level drivers. These are very high value and get you well into diminishing returns if used skillfully
  • Medium pricepoint - SB Acoustics, ScanSpeak Discovery sort of level. These are essentially high end drivers with great construction quality and terrific value.
  • High pricepoint - value starts dropping off dramatically here, but performance could justify it if low distortion, deep bass and a small box are desired.
The next decision would be topology, the basic premise of the speaker.
  • A simple two way with baffle mounted drivers can work surprisingly well with a small woofer and/or a robust tweeter, but output is limited. Smooth and wide directivity at the expense of low bass.
  • A larger two way (6.5"-8" woofer) works best with a waveguide, and here I think you would want to use an off-the-shelf solution to keep the speaker easy to build. This makes the speaker intrinsically more costly, as there are only a handful of good tweeters with significant waveguides, and none are really cheap. Still, this is the sort of speaker I would build for myself.
  • An interesting option would be to make a small stand mount 3 way, but here the problem is that crossover costs essentially double, and the size of the speaker increases. Still, the performance increase is significant, and people do seem to prefer the wide dispersion sound.

For me, I think the most interesting topology is actually the first - a simple two way with a small woofer. Why? Here at ASR we've had many discussions about subwoofers and how preference scores for full range speakers are almost invariably much higher if subwoofers are added. I think it would be really fun to design the 'ultimate satellite speaker', perhaps using a 5" woofer and a nice dome tweeter, focusing on robust output above 70hz or so. In terms of price point, I think this could be accomplished well using the "Low Pricepoint" category I have described, but I wouldn't argue against "medium pricepoint" if we could design something truly excellent.

The second most interesting combination is a cheap 3 way - I think people really do like wide dispersion speakers, and something using again, the "Low Pricepoint" would be appealing to a lot of people. We've had discussions about the excellent performance of inexpensive midrange drivers such as the BMRs and Vifa TG/TC9, coming in at around 15 dollars, and in using drivers such as these, you can use a much less expensive tweeter than a two way with excellent performance.

The third option would be a two way with a waveguide. I'm thinking the Seas DXT or the Wavecor unit would be the best two candidates - I'm not aware of any others. If you combine these with a sb17MFC or something similar, you get a small, high value full range speaker.

Fourth option would be something built for ultra high performance - we have the tools to do this. Perhaps target the Revel Gem2 or some other stand mount 3 way.

In terms of the community aspect, I think we can break down the tasks needed for this project:
  1. Specifying what we want the speaker to do in order to be successful - target SPL, bass extension and DI.
  2. Simulation of drivers from datasheets - finding interesting and viable drivers and nailing down the target performance
  3. Driver distortion measurement and initial characterization as needed
  4. Enclosure design, potentially for production if interested
  5. Enclosure prototype construction
  6. Crossover design
  7. Crossover construction, potentially PCB design
  8. Project management - keep the discussion focused, keep the budget on track
  9. Design verification using $100K measurement platform
I think most of us speaker DIYers can do most of these things, but we all excel or suck at some of them. Speaking for myself, my crossovers always look like hot garbage, and I don't really feel like making cabinets right now. On the other hand, some people have no interest in crossover design, but can make a nice mdf cabinet in an afternoon. Additionally, and most importantly, there are tasks which allow non speaker people to contribute, such as managing the project and tracking the parts cost and availability.

If you are interested in the project, please respond to this thread by telling me how you would like to contribute. We can then proceed to making a thread dedicated to specifying the topology, and nailing down a design target.

Willing to help, but have 2 speakers underway right now. Should finish one by end of week.

Only recently found this thread. So let me know if you still might need help doing something for this project. :cool:
 

dwkdnvr

Senior Member
Joined
Nov 2, 2018
Messages
418
Likes
698
I have been looking at the Hypex DSP amps, both 3-way and 2-way. They seem to be a little cheaper, and the amplifier modules have a good reputation. Following this thread with great interest.

If you are saying that the Hypex amps are cheaper than the Daytons, you're looking at the wrong Daytons. We're looking at the little TPA3116 amps - I'm guessing you may be looking at their high-end DSP plate amps. In that case, I tend to agree that the Daytons are a bit expensive for what they bring. The little TPA3116 amps are interesting though, and quite inexpensive. e.g. https://www.parts-express.com/dayto...ass-d-audio-amplifier-board-with-dsp--325-126

The Hypex amps are certainly interesting, and could certainly be used to execute the type of design we're working through here. They'd run about $750 for a pair of the cheapest 3-way units though, which would raise the barrier to entry for this project beyond what we're really trying to target. So, the reason we're looking at the Dayton amps as the 'default' recommendation is simply since they look to be in keeping with the desire to keep the overall project budget somewhere in the $500/pr range.

Having said that, the Hypex Fusion amps really don't seem to be any more expensive than the raw amp modules, so you almost get the DSP capabilty for free. Certainly an option that should be looked at for a 'high end' or 'final' build.
 

jae

Major Contributor
Joined
Dec 2, 2019
Messages
1,208
Likes
1,508
FWIW on the DSP amp front someone pointed these out the other day http://www.3e-audio.com/finished-amp/tpa32xx-all-in-one-amplifier/ . They're TPA3255/3251 amps with integrated ADAU1701. The TPA3251 in particular is a very good performer with the 3255 not far behind, meaning that the ADAU1701 itself is probably the limiting factor. They're just coming to market, but might provide an alternative to the Dayton amps for someone looking for an off-the-shelf option rather than a project. Not as attractive for building into a clean powered monitor style setup, though.

Looks like the same SigmaStudio programming workflow as the Dayton DSP amps, at least at first glance.

Looks like a very nice option for the price. Would be nice if the DSP profiles didn't need to be manually flashed.
 
OP
617

617

Major Contributor
Forum Donor
Joined
Mar 18, 2019
Messages
2,404
Likes
5,296
Location
Somerville, MA
Looks like a very nice option for the price. Would be nice if the DSP profiles didn't need to be manually flashed.

I've tried to buy amp modules from 3e before and nothing is ever available.
 

Lbstyling

Addicted to Fun and Learning
Forum Donor
Joined
Jan 2, 2019
Messages
500
Likes
461
Do we agree that the speaker should target reaching ref level (105db/1w/1m?) With a little headroom if it is achievable without significant costs?

If that's a criteria, and we are aiming for active amplification, 50w/4 ohm chip based plate amps are reasonably cheap, but would need 87db/1w/4ohm drivers minimum. 89db/1w/1m would be preferable

As more amplification is comparatively expensive Vs changing drivers to higher efficiency, especially in a cubic volume limited design, we need to be shooting for drivers that are this or higher.
 
Last edited:

Colonel7

Addicted to Fun and Learning
Joined
Feb 22, 2020
Messages
616
Likes
875
Location
Maryland, USA
I've tried to buy amp modules from 3e before and nothing is ever available.
Are they legit? Nothing about them seems so at least for a project like this. The other thread extolling them has a few people spouting the virtues of rolling amps that cost less than $75. Not plates but finished boxes
 
Last edited:

dwkdnvr

Senior Member
Joined
Nov 2, 2018
Messages
418
Likes
698
Are they legit? Nothing about them seems so at least for a project like this. The other thread extolling them has a few people extolling the virtues of rolling amps that cost less than $75.
They're legit as far as it goes - a small Chinese manufacturer with limited distribution outside China. I have one of their TPA3255 modules running my woofers, and it works perfectly well, and compared to other cheap Chinese amp boards I've seen the quality seems to be pretty good. There are a lot of folks on diyaudio using their boards, generally with success.

But, they're obviously not really a 'vendor' in the conventional sense - no support, no guarantee of availability etc. My intent in calling attention to them wasn't to suggest that this project semi-officially adopt them but just to make anyone considering similar projects aware of other options with different trade-offs.
 
Top Bottom