Don't get hung up on the "reference" wording as it's meant to be a reference point or datum one can compare other speakers to.
If I was doing this again I would probably use SB Acoustics drivers instead of SEAS as I've had great success with them. Something like the drivers in Jeff Bagby's Soprano with a SB29NRX woofer. For a el cheapo, the same tweeter with SB12PAC25-4 and 2 x SB15SFCR39-8.
If I was doing this again I would probably use SB Acoustics drivers instead of SEAS as I've had great success with them. Something like the drivers in Jeff Bagby's Soprano with a SB29NRX woofer. For a el cheapo, the same tweeter with SB12PAC25-4 and 2 x SB15SFCR39-8.
Reference is in the title of the thread but it has unfortunately come to have more than one common meaning. It can be used as a datum in the way an Auratone monitor is used as a reference for low quality sound in studios. However, it is also widely used by commercial companies with their reference lines or reference speakers where it is not intended to convey datum of variable quality. They intend "reference" to imply something like so good it is the standard others are judged against.
We also learnt during the "open source" group projects that those running DIYAudio do not want their name attached to a group project so we can perhaps keep "3-way" but both "reference" and "diyaudio" probably need to get the boot.
Whatever, is there much interest in a group project for a neutral high fidelity 3 way design using standard range drivers rather than premium range ones? This would put the budget in the region of $1000 for a pair perhaps a bit less but certainly above $500.
If there is some interest I would be tempted to put together with a bit of help from others a large signal model to help examine in a quantifiable manner the nonlinear performance of various configurations (i.e. what is required for high fidelity/reference performance). This would hopefully provide a detailed specification against which to seek appropriate standard range drivers and assess compromises in a quantitative manner.
SB Acoustics probably do offer the best value for money at the moment outside the US but it was different in the past and will almost certainly be different in the future. If the speaker is appropriately designed the drivers should operate within their designed "linear" operating region creating low levels of nonlinear distortion and with appropriate compensation for linear distortion it shouldn't matter much. What should matter most is appropriate speaker design.
WOW...11 post's yesterday, ^6 already today. That for an old topic that started many years ago.
Is it coming back to life?
.
The drivers that was agreed upon maybe need to reconsidered as time has changed.
Many seem to point out that SBaccoustics have good value/quality ratio.
Reference speaker? Maybe need a new definition on what a reference speakers is in this context.
For me it means that all aspects are considered for a proper speakers design. This is what makes it interesting for beginners...the learning process. What aspects need to be considered. Where you make trade-offs..etc.
Budget? I think around 600 US$. This is a good pricepoint for a beginner who takes the next step and dive into a 3-way design.
There is a lot of expertise on this site. If we could bundle all this expertise here would be great. I assume that for some their participation will be strictly academical as they are not able to build it. But that's OK..It can lead to some starting point where someone else takes over.
3-way designs. I like to go as broad as possible. Every design has it strengths and trade-offs. The interesting part is that the same set of drivers is used in the different designs. It is cost-saving if one like to explore those different designs and can re-use the drivers and probably some XO parts.
Far cheaper then starting from scratch. (Also if one use different drivers how can you compare the designs?)
- 3-way (vented/Closed/TL)
- Bookshelf / Tower
- Wide/narrow baffle
- OB ?
- low order XO / Higher order XO
- Serial/ parallel XO
- Time-aligment (design vs XO)
- ... (Im newby so maybe I overlook something here )
Drivers :
Some ideas of drivers that may fit
Tweeter : SS2608 ; CAT308 ; SEAS 27TFFC ; SB26STAC ; SB26ADC (some common options for not going all SBaccoustics)
MID : SB15MFC30 (are the different cone materials in this range exchangeable without XO impact? )
Woofer : SB23NBAC ; SB23NRX
In EU around (80+60+100) more or less 250 euro
Is it coming back to life?
.The drivers that was agreed upon maybe need to reconsidered as time has changed.
Many seem to point out that SBaccoustics have good value/quality ratio.
Reference speaker? Maybe need a new definition on what a reference speakers is in this context.
For me it means that all aspects are considered for a proper speakers design. This is what makes it interesting for beginners...the learning process. What aspects need to be considered. Where you make trade-offs..etc.
Budget? I think around 600 US$. This is a good pricepoint for a beginner who takes the next step and dive into a 3-way design.
There is a lot of expertise on this site. If we could bundle all this expertise here would be great. I assume that for some their participation will be strictly academical as they are not able to build it. But that's OK..It can lead to some starting point where someone else takes over.
3-way designs. I like to go as broad as possible. Every design has it strengths and trade-offs. The interesting part is that the same set of drivers is used in the different designs. It is cost-saving if one like to explore those different designs and can re-use the drivers and probably some XO parts.
Far cheaper then starting from scratch. (Also if one use different drivers how can you compare the designs?)
- 3-way (vented/Closed/TL)
- Bookshelf / Tower
- Wide/narrow baffle
- OB ?
- low order XO / Higher order XO
- Serial/ parallel XO
- Time-aligment (design vs XO)
- ... (Im newby so maybe I overlook something here
)Drivers :
Some ideas of drivers that may fit
Tweeter : SS2608 ; CAT308 ; SEAS 27TFFC ; SB26STAC ; SB26ADC (some common options for not going all SBaccoustics)
MID : SB15MFC30 (are the different cone materials in this range exchangeable without XO impact? )
Woofer : SB23NBAC ; SB23NRX
In EU around (80+60+100) more or less 250 euro
Is it coming back to life?
If there is enough interest but that has yet to be established and, perhaps more importantly, around what spec. In my case, just like last time, I am interested in getting involved with a group design of broadly the cheapest "proper" high fidelity ("reference"?) main speakers for the home. Setting aside for a later project requirements to control the low frequency room response. Just the mains for stereo to see if we can get things rolling.
If there is interest people will hopefully put forward what they are interested in, what they are not and what they would like to contribute. In my case I am fairly confident that an 8" woofer is insufficient for high fidelity for several reasons but I am more than happy to defer this sort of decision to a quantitative assessment based on required technical performance. I have little to no interest in brands until the performance, configuration and budget has been settled on.
Blockers for me would be too low a budget to get "proper" high fidelity performance, premium/prestige drivers pushing the budget too high, little interest in sorting out the design on paper, less than 2 or 3 people expressing genuine interest and possibly one or two other things.
Active and passive crossovers need discussion. If there is to be a passive crossover version it will introduce extra constraints on the choice of driver and cabinet configuration. I would only be interested in building active speakers for myself but I am not against passive crossovers and juggling the constraints they introduce into a design.
Yeah ground up design to a particular goal can be reference, for the same or similar goals. Luckily a 3 way seems to be enough ways to be able to have most basic issues sorted, otherwise the quest was doomed already You kind of already set the goal as " the cheapest "proper" high fidelity ("reference"?) main speakers for the home", try and refine that if needed and then keep the goal very clear in mind along the way as you conduct the process. Maybe include notion of the size or price target but the goal should be pretty low level to allow maximum design freedom. Maybe take the CTA 2034 standard as somekind of reference or metric so it is kind of easy to find the path through and evaluate success in the end?
Choosing drivers is the last step and limiting oneself for certain stuff like active or passive xo before even starting to go through the design process seems backwards to me as well. This often leads to a dead end limiting getting more performance for the goal, what ever that is.
Good luck!
You kind of already set the goal as " the cheapest "proper" high fidelity ("reference"?) main speakers for the home", try and refine that if needed and then keep the goal very clear in mind along the way as you conduct the process. Maybe include notion of the size or price target but the goal should be pretty low level to allow maximum design freedom. Maybe take the CTA 2034 standard as somekind of reference or metric so it is kind of easy to find the path through and evaluate success in the end?Choosing drivers is the last step and limiting oneself for certain stuff like active or passive xo before even starting to go through the design process seems backwards to me as well. This often leads to a dead end limiting getting more performance for the goal, what ever that is.
Good luck!
Last edited:
Which currently available smallish affordable driver gives the best performance in the range 300 to 3000Hz and works well with the minimum of crossover components but will work [ perhaps better?] with steeper slopes etc.
I like the idea of the woofer section being able to be switched closed box or ported to enable the future use of a subwoofer.
A parochial approach to drivers won't work IMO, the selected components have to be readily available almost everywhere and the XO has to be passive as most people won't be able to utilise a system with as many amplifiers etc as some of us
I like the idea of the woofer section being able to be switched closed box or ported to enable the future use of a subwoofer.
A parochial approach to drivers won't work IMO, the selected components have to be readily available almost everywhere and the XO has to be passive as most people won't be able to utilise a system with as many amplifiers etc as some of us
^ Moondog, Yeah it would be paramount to have drivers that are available for anyone. I'm also with you that a passive crossover would make it easier and perhaps cheaper to approach by anyone.
The driver requirement you specify reads like you already have a design that reguires such driver? What would be the SPL capability target? and what kind of system it is otherwise, the other drivers and what dispersion would be nice?
Also, why does crossover need as few components as possible? I think it is going to have minimum amount for the particular system anyway, and if that is what is required for a reference sound then it is also the cheapest. I mean the crossover is rather trivial to optimize when the system fits together nicely. To achieve this one should perhaps think any crossover is possible until it is time to consider it. And unless the thing ends up very exotic passive crossover should be possible, just select suitable drivers when it is time to select the drivers.
As disclaimer I have to say I haven't designed or built any passive crossover ever. I've played a lot with VituixCAD though (digital filters) and the system ends up as good as the measurements are, in other words the system without xo, the transducers and the construct. There are many ways to come up with a system, but I think a reference system is something where every compromise is taken to achieve the goal and not for any other reason.
In the end it is just about meeting a goal very well, well enough worthy the title reference Define the goal, meet it in few different designs and decide what is better, the is going to be some that cost too much, or are too big, or just dont sound good enough with high enough SPL. Then minimize the cost.
The driver requirement you specify reads like you already have a design that reguires such driver? What would be the SPL capability target? and what kind of system it is otherwise, the other drivers and what dispersion would be nice?
Also, why does crossover need as few components as possible? I think it is going to have minimum amount for the particular system anyway, and if that is what is required for a reference sound then it is also the cheapest. I mean the crossover is rather trivial to optimize when the system fits together nicely. To achieve this one should perhaps think any crossover is possible until it is time to consider it. And unless the thing ends up very exotic passive crossover should be possible, just select suitable drivers when it is time to select the drivers.
As disclaimer I have to say I haven't designed or built any passive crossover ever. I've played a lot with VituixCAD though (digital filters) and the system ends up as good as the measurements are, in other words the system without xo, the transducers and the construct. There are many ways to come up with a system, but I think a reference system is something where every compromise is taken to achieve the goal and not for any other reason.
In the end it is just about meeting a goal very well, well enough worthy the title reference
Define the goal, meet it in few different designs and decide what is better, the is going to be some that cost too much, or are too big, or just dont sound good enough with high enough SPL. Then minimize the cost.
Last edited:
Yes I have a set of Vifa drivers but that is now of little relevance as they haven't been available for a long while
But I was of the opinion that the basic set should be 8" woofer and either a 4" or 5" mid and a decent low Fs tweeter.
I did make a minimal XO speaker using the old Vifa P-13 when the 2-Way project was active.
But that was a simple 2.5 using one capacitor and one cored inductor and the Vifa M-22 and the D-27-45 tweeter. With good drivers such speakers work well ; or well enough to use as comparisons to see where improvements can be made.
For my own use I went active decades ago so my XO skills are not what they should be
But I was of the opinion that the basic set should be 8" woofer and either a 4" or 5" mid and a decent low Fs tweeter.
I did make a minimal XO speaker using the old Vifa P-13 when the 2-Way project was active.
But that was a simple 2.5 using one capacitor and one cored inductor and the Vifa M-22 and the D-27-45 tweeter. With good drivers such speakers work well ; or well enough to use as comparisons to see where improvements can be made.
For my own use I went active decades ago so my XO skills are not what they should be
Basic setup would be 3 drivers. 8" 4-5"- Tweet.
Low fs Tweeter isn't needed I asume. Target XO 300 & 3k Hz but easy drivers to work with.
Adding an additional woofer would have small impact on the design. This can be tackled out of scope of this project.
Easiest XO to lower complexity for beginners. Maybe a basic XO can be developed. Mention the trade-offs and some remedies can be proposed. This way a beginner can explore and learn.
Many mention SBacousitcs have good quality/price ratio and are widely available.
Maybe proposing drivers isn't the right way to start, but for me it's a condition to work with. This will form the base for all designs possible.
People who have access to active crossover and give input but the goal should be to develop different passive XO's. That's the learning school for beginners. Possibilities with active are unlimited not the case with passive. Apprentice sailers also start with a sailboat before stepping into a motorised vessel.
Low fs Tweeter isn't needed I asume. Target XO 300 & 3k Hz but easy drivers to work with.
Adding an additional woofer would have small impact on the design. This can be tackled out of scope of this project.
Easiest XO to lower complexity for beginners. Maybe a basic XO can be developed. Mention the trade-offs and some remedies can be proposed. This way a beginner can explore and learn.
Many mention SBacousitcs have good quality/price ratio and are widely available.
Maybe proposing drivers isn't the right way to start, but for me it's a condition to work with. This will form the base for all designs possible.
People who have access to active crossover and give input but the goal should be to develop different passive XO's. That's the learning school for beginners. Possibilities with active are unlimited not the case with passive. Apprentice sailers also start with a sailboat before stepping into a motorised vessel.
Last edited:
To be blunt, for a passive crossover, this is not correct. Raising the low frequency output by 6 dB or so (depending on the details of how it is done) substantially changes how baffle step correction is handled, raises the sensitivity requirements of the midrange and, if competently done, raises the clean overall performance of the whole speaker by that 6 dB or so.
The overall performance of a 3 way speaker with an 8" woofer is going to be limited by the performance of the woofer because it is too small relative to a 4-5" midrange to handle the requirements of increasing area due to lower frequency plus the sound increasingly radiating in all directions rather than mainly forward. With a passive crossover and a reasonable amount of baffle step compensation such a speaker is not going to be much different in performance to an equivalent 2 way.
Of course, one could look at a 3" midrange for a possibly more balanced (but low output) design in the manner of pro "nearfield" monitors, subwoofers for more output but unless stood under the mains with a raised crossover frequency there will be an SPL hole above 80 Hz, etc...
Of course the whole point of DIY is to build what one wants rather than what the laws of physics and economics dictates but it can be wise to be aware when one's designs are drifting away from the latter.
^ I guess the term reference is then mostly for best possible sound for given application and to a cost?! Throwing infinite amount of money in would make better sound but beating a reference with less money would be challenge to anyone, to beat the reference in its own game and become new reference?
And for 8" bass driver, there should be SPL goal to reach and then calculate if 8" driver is enough. Aint the official reference listening level 85dB plus peaks, where the ear is linear and the music is mixed at? Then that is the target, over the whole bandwidth. What is the bandwidth of reference system? 20Hz-20kHz or a bit relaxed? Is there a way to do this within a budget or does it define the minimum budget?
It is all about setting the goal for a reference speaker.
And for 8" bass driver, there should be SPL goal to reach and then calculate if 8" driver is enough. Aint the official reference listening level 85dB plus peaks, where the ear is linear and the music is mixed at? Then that is the target, over the whole bandwidth. What is the bandwidth of reference system? 20Hz-20kHz or a bit relaxed? Is there a way to do this within a budget or does it define the minimum budget?
It is all about setting the goal for a reference speaker.
Last edited:
No. Given enough time, it seems to work out that it is limited by the knowledge of the builder.
What do you think about showing only a set of polar measurements and letting the builder find their own way to achieve it?MrHifiTunes said:
Agreed but defining it in a flexible quantitative manner which enables alternatives to be assessed is not something that beginners are able to do at present. It comes with study and experience but this tends to be coloured somewhat by understandable hobby enthusiasms for a wide variety of things. I have mine. The apparent significant status/presence given to a number of subjective audiophile personalities mixing in a range of audiophile foo with useful advice also doesn't help the beginner recognise what raises technical performance and what might look/feel good but does little in this respect.
As mentioned earlier, I am pondering putting together a large signal model for a speaker to enable much of this to be addressed. Hopefully to the extent of being able to show the performance expected from a range of driver configurations and hence where the practical minimum lies for a high fidelity/reference speaker. The increasing amount of "Kilppel-type" data knocking around should enable a large part of such a model to be setup and checked although there are one or two areas like the thermal side where there is likely to be little useful data and the empiricism will be subject to somewhat wider tolerances. May require some supporting 3D simulations to determine reasonable bounds for estimates.
I had hoped for some interest in a modestly priced "proper"/full high fidelity 3 way group speaker project to prod me into activity but it doesn't look to be happening. I still might give it a go but interesting non-essential projects like this have had a tendency in the past to drift down the to-do list and off the bottom without a deliverable to make them a requirement.
I would be interested in collaborating on a 3-way project. I will not be able to build prototypes, as I have too much else going on right now (2 speaker projects in work)... but I can help with simulations and other aspects.
Reading through the modern posts, I am not sure there is a defined set of guidelines or requirements... I have heard some ideas, but I don't think we have reached a consensus yet. Here are some of the ideas I have heard:
1) A TMW 3 way tower with an 8" woofer. This is reasonable, but with such a system we need to have realistic expectations about max SPL and sensitivity. Twin woofers would result in greater SPL and higher sensitivity.
2) A budget of $600. Is this for the drivers, or for everything? I would think that $600 for drivers is reasonable, but a total budget for everything of $600 starts to really limit our choices of drivers and crossovers. It would mean a driver budget of about $450, and that is rather stingy for 6 drivers.
3) A choice of drivers which are readily available in most nations. From what I have gathered, Dayton is somewhat over-priced in Europe, and Seas is hard to get (expensive) in Australia. Are there any other driver OEMs that should be excluded?
Any discussion on the above three requirements? Any other requirements? This is the brainstorming phase of this process, so throw out anything that comes to mind... we can distill it down later.
j.
Reading through the modern posts, I am not sure there is a defined set of guidelines or requirements... I have heard some ideas, but I don't think we have reached a consensus yet. Here are some of the ideas I have heard:
1) A TMW 3 way tower with an 8" woofer. This is reasonable, but with such a system we need to have realistic expectations about max SPL and sensitivity. Twin woofers would result in greater SPL and higher sensitivity.
2) A budget of $600. Is this for the drivers, or for everything? I would think that $600 for drivers is reasonable, but a total budget for everything of $600 starts to really limit our choices of drivers and crossovers. It would mean a driver budget of about $450, and that is rather stingy for 6 drivers.
3) A choice of drivers which are readily available in most nations. From what I have gathered, Dayton is somewhat over-priced in Europe, and Seas is hard to get (expensive) in Australia. Are there any other driver OEMs that should be excluded?
Any discussion on the above three requirements? Any other requirements? This is the brainstorming phase of this process, so throw out anything that comes to mind... we can distill it down later.
j.
This post is a very practical way to start the conceptualizing.
In my opinion, using either the ScanSpeak 10F or the Satori MR13 would be a big step in the direction of success... these drivers will consume a good chunk of the budget, but the midrange is a very important driver to get right...
Agree about the importance of the mid driver.
Though the satri MR13 will be around 60% of the available budget.
SS 10F around 30%.
Sd from Satori =70cm2 SS 10F only 36cm2
double surface -> double price
.I would go paper or poly for mid's. with a preference for poly.