I'm not sure if it was Kanye or some other high profile rapper, but I heard an anecdote about so much bass in the studio that it made people physically ill. It sounds scary, so yeah, I'll definitely do whatever they say!
It is interesting about the wide range of speaker designs that are possible. Not long ago I was listening to an ATC fanboy (PresentDayProduction on youtube) raving about the ATC 3" dome tweeter. So using a 10" or 12" coaxial driver to cover midrange is the absolute opposite end of the spectrum.
After I figured out how much it would cost to get an ATC type 3" dome midrange (Volt 3" dome), I started looking at the Morel 6" drivers. They look good to me, and they'd match my other drivers aesthetically.
And of course all the home theater and hifi speaker designs today use progressively smaller drivers for higher frequencies.
I will have to investigate this passive cardiod idea. I don't think my simulation software knows how to handle it, so it sounds like a lot of experimenting!
Yes, I do plan on using DSP. I am thinking of using a passive crossover between the coax woofer and the compression driver, and to DSP the signal being fed into the passive crossover.
Atc design philosophy is at the other spectrum than what we talk yes. Results are differents i agree but not because of drivers or propriatary technology, but about the acoustic design: the properties behavior of drivers in their intended range.
I lived with Scm110A as mains for two years in the past. I understand fan of the mid dome but as i was involved in the whole control room build/design i know what are the trade off to this omni radiating pattern: it shift the requirements to acoustic treatments... needing large amount of damping and early reflection treatments.
To have directivity control relief this needs.
It's all about preferences in rendering.
There is no absolute, only set of trade off and what please you.
I lived with Scm110A as mains for two years in the past. I understand fan of the mid dome but as i was involved in the whole control room build/design i know what are the trade off to this omni radiating pattern: it shift the requirements to acoustic treatments... needing large amount of damping and early reflection treatments.
To have directivity control relief this needs.
It's all about preferences in rendering.
There is no absolute, only set of trade off and what please you.
It is much better to linearize each driver first to beyond the cover frequency. This will first of all show you if the crossover point you have in mind is realistic. And second it will make it a breaze to apply the crossover points without surprises. Any dip will be the result of a timing issue and this can be easily addressed by providing one of the drivers a delay. Of course this will require two stereo amps, but an amp is cheaper to than an analog crossover anyways.
I'm with vacuphile, if you plan to insert fir then do it by driver rather than overall imo. A couple of toping amp and you are good to go... if you plan on small coax, check your delay accept low value (0,xxx ms range, or better sample range ( on the tannoy i have something like 6/7 sample delay at 44,1khz to time align iirc...).
That said i don't use fir 'manipulation' ( 'room treatment' or whatever trying to 'correct' acoustical/electrical deficiencies) but complementary filters ( cross over).
That is not to say it might not be benificial.
That said i don't use fir 'manipulation' ( 'room treatment' or whatever trying to 'correct' acoustical/electrical deficiencies) but complementary filters ( cross over).
That is not to say it might not be benificial.
That makes sense. I purchased a 4x8 channel DSP, which has enough channels if I keep everything simple. (It was defective and is being returned, but I should have the new one soon.)
I had an idea for a more complicated arrangement that would use extra DSP outputs. It may be unnecessary or it may be a misunderstanding on my part, but...
Based on the idea that drivers in different locations in the room create peaks and nulls in different places, I wanted to send a combined signal to the subwoofers, which would be a summed signal below 60hz and a stereo signal above 60hz.
The idea is that both subwoofers would be playing the same low frequency content sub 60hz, and therefore I can EQ them individually so that they produce a more even bass response throughout the room. This is sometimes done in home theater with 2 or more subwoofers.
I took some measurements and found that when comparing the in room frequency response of my front facing woofer and the rear facing woofer of my cabinet. Their frequency response in room was different between 60hz and 200hz, so I can EQ them against each other in that range. But below 60hz they were identical. So I realized can't control frequencies below 60hz by EQing between the 2 drivers in the same cabinet. But maybe I can Eq between subwoofers in the two separate cabinets to control the lower frequency room nodes? It should work.
But there is one thing I do not know. I don't know if there is enough difference in the low frequency content between left and right channels to make it necessary to sum the two? If I can assume that the sub 60hz content is identical or nearly identical on the left and right channel, then I can forget about it and free up a DSP channel.
I do know that vinyl records used a summed low frequency, because otherwise the needle would jump too much. But it stands to reason that there won't be much of a difference in the low frequency between the left and right channels in a modern recording.
I've done a few sims and the WCW orientation has better directivity than the CWW format if you're crossing between 200-300hz. This depends on what you're willing to accept in terms of speaker height, but otherwise you get these ugly floor bounces and woofer combing.
One of the woofers in my design is actually a passive radiator, so I assumed it is not counted as a woofer? I assume that would change the simulation you are describing.
The best info I can find about PRs seems to agree that the PR doesn't create any significant output above the tuning frequency. At the tuning frequency, the PR is making most of the output, not the driver. So the PR should not matter in terms of it interacting with the coaxial?
I am planning on crossing the coaxial to the woofer at 250hz or lower,and the PR shouldn't be doing anything at that frequency. So the only thing that should matter is the distance between the woofer and the coaxial.
The PR shouldn't be doing much of anything until somewhere around 60hz, and there won't be any 60hz content coming from the coaxial driver.
But I am by no means an expert at any of this!
I missed that! You're right, it will act like a 3.5-way with the PR working out of phase to produce extreme low frequency. I'm used to seeing them on the back! This tends to help disperse out of phase noise in the correct direction-- backwards. Because it's more or less in phase on the reverse side, it creates a bump instead of a null. This can be EQ'd out easily with a high Q PEQ filter, whereas filling that void with an equalization filter would be much more difficult. This all, of course, depends on specific drivers, boxes, and arrangements. None of these are hard and fast rules, but it's consistent with data collected from Erin's Audio Corner. Worth looking into.
Is that really enough passive radiator? Most of the time, with similar suspensions and surrounds to the woofer, you need approximately double the radiator surface area. Usually this translates to one "size up", like a 15" PR for a 12" woofer.
This is a rad project. I really hope it works out.
Is that really enough passive radiator? Most of the time, with similar suspensions and surrounds to the woofer, you need approximately double the radiator surface area. Usually this translates to one "size up", like a 15" PR for a 12" woofer.
This is a rad project. I really hope it works out.
According to the simulation the PR is enough. The woofer has 5.5mm Xmax, and the PR has 13mm. So the PR has slightly more than 2x the displacement of the driver.
I'm not sure about the phase of a PR. The common understanding is that the PR is out of phase, but I've read several forum threads where people are discussing how PRs are out of phase, and then someone who seems very knowledgeable shows up and explains that actually the PR is always in phase. I am leaning toward the latter point of view, but I don't actually know.
I think the simulation looks fine. Below is the data in a 70L enclosure. My previous drawing showed a 55L volume, which results in an F3 of 39hz and a +1db peak between 50 and 60hz.
I have to thank you for helping to develop this idea. I had a vague idea, and you provided the "how" and "why" to do it. I wouldn't have chosen a 12" coax on my own, but the 12" makes perfect sense when considering the directivity.
I'm still stuck on choosing the particular driver. I feel that the 12" coax and compression driver are the most critical to getting a good result, but I'm not sure how much I need to spend or how much I'm willing to spend to get ideal results.
Parts Express has a few 12" coaxials ranging from $124 to $765. They don't offer the FaitalPro 12" at all, but I found it on Ebay for $445. Another store sells it for ~$700, so ebay is a pretty good deal. But It's still $900 for the pair! And I still need to buy another amp!
Maybe it's a silly question, but how much should I expect to pay for a new, good quality, coax driver? Would a cheap ~$250 driver be a regrettable decision? Or is this a case where the more expensive drivers are just adding higher sensitivity/ higher output for PA applications?
And second, if I got a woofer with a screw in compression driver, would it make sense to focus more on the woofer or on the compression driver?
I have time to decide, but I don't have much knowledge or experience of my own to use to make a decision!
Even if you select something decent, final result will depend on proper crossover. With so many questions, it seems like you are far better of with following published well accepted design.
Hi,
Happy my rumblings inspired you.
May i suggest you rather focus on 'acoustic design' than choosing drivers (for the moments, nothing stop you to look at them meanwhile 😉 ).
About baffle edge diffraction : https://heissmann-acoustics.de/en/kantendiffraktion-sekundaerschallquellen-treiberanordnun/
A coax can be seen as a waveguided tweeter. I repeat again you should download all Kef's whitepaper too (if you have not already done it).
Lots of good design practice and science based approach in there.
I suggest those readings too:
http://www.gedlee.com/Papers/papers.aspx
This should help you refine the 'form factor' of your final box, location of drivers, radius of roundover ( if any), etc,etc,...
Now drivers. Well it is very difficult to guide anyone about this: we all listen to different things, have different preferences.
Not taking into account the implementation it's even more difficult... i agree with Adason about xover, it can really destroy the potential outcome if not 'correctly'* implemented.
Eg the 6,5" Tannoy from passive System600:
https://www.audiosciencereview.com/forum/index.php?threads/tannoy-system-600-speaker-review.11919/
This 1,5khz dip is the result of the offset in z axys i talked about previously. In practice it isn't as problematic as it can seems ( they are not unlistenable as Amir suggest and an EQ won't solve the issue either... but as he didn't even listened to them and don't have considered Tannoy philosophy about passive filtering... he couldn't bother investigate about it. 🙄 this is one of the bad side of being too much in the 'objectivist' camp imho, but i disgress...).
Anyway, there is way to solve this issue and push the driver to it's full potential going active filtering. 😉
Fir are an answer but not the only one and IIR ( or passive) strategy can be implemented too.**
So all that won't help you choosing a reference but can put things into perspective.
I would suggest you to try to contact some dedicated to PA company around you and ask if they own wedge using coax, which brand they use and go for a listen or loan to listen by yourself. Explain your plan too, it could help ( if you contacted me about this while i was still active i would have helped you for free as it have the potential to bring info/knowledge).
This is mostly what i do or have done to make my own opinion ( easier for me as a former pro with still connections in the field).
Here is one example of a Bms 12c362 used with dsp:
http://westlab-audio.com/sites/default/files/downloads/2017-05/Datenblatt LR12.pdf
As you can see the polar map is not bad at all... and as i've heard this driver in another design ( box was close though) i can tell i liked their sound. At 300euro/piece i find them a bargain imho.
* in the passive tannoy s600 example, the xover can be seen as underpar... until you focus on Tannoy preference for low part count on passive xover design and use of the 'natural' behavior of transducers ( shallow slope using 'natural' transducer behavior)... as a results they have choosen to compromise on the 1,5khz suckout. This open to another model in the range 'improving' on this and open to other sales...
** please read on the series of article on how to handle the issue with iir filters:
https://greatplainsaudio.com/gpa-vintage-altec-ezine/altec-lansing-duplex-speakers/
Happy my rumblings inspired you.
May i suggest you rather focus on 'acoustic design' than choosing drivers (for the moments, nothing stop you to look at them meanwhile 😉 ).
About baffle edge diffraction : https://heissmann-acoustics.de/en/kantendiffraktion-sekundaerschallquellen-treiberanordnun/
A coax can be seen as a waveguided tweeter. I repeat again you should download all Kef's whitepaper too (if you have not already done it).
Lots of good design practice and science based approach in there.
I suggest those readings too:
http://www.gedlee.com/Papers/papers.aspx
This should help you refine the 'form factor' of your final box, location of drivers, radius of roundover ( if any), etc,etc,...
Now drivers. Well it is very difficult to guide anyone about this: we all listen to different things, have different preferences.
Not taking into account the implementation it's even more difficult... i agree with Adason about xover, it can really destroy the potential outcome if not 'correctly'* implemented.
Eg the 6,5" Tannoy from passive System600:
https://www.audiosciencereview.com/forum/index.php?threads/tannoy-system-600-speaker-review.11919/
This 1,5khz dip is the result of the offset in z axys i talked about previously. In practice it isn't as problematic as it can seems ( they are not unlistenable as Amir suggest and an EQ won't solve the issue either... but as he didn't even listened to them and don't have considered Tannoy philosophy about passive filtering... he couldn't bother investigate about it. 🙄 this is one of the bad side of being too much in the 'objectivist' camp imho, but i disgress...).
Anyway, there is way to solve this issue and push the driver to it's full potential going active filtering. 😉
Fir are an answer but not the only one and IIR ( or passive) strategy can be implemented too.**
So all that won't help you choosing a reference but can put things into perspective.
I would suggest you to try to contact some dedicated to PA company around you and ask if they own wedge using coax, which brand they use and go for a listen or loan to listen by yourself. Explain your plan too, it could help ( if you contacted me about this while i was still active i would have helped you for free as it have the potential to bring info/knowledge).
This is mostly what i do or have done to make my own opinion ( easier for me as a former pro with still connections in the field).
Here is one example of a Bms 12c362 used with dsp:
http://westlab-audio.com/sites/default/files/downloads/2017-05/Datenblatt LR12.pdf
As you can see the polar map is not bad at all... and as i've heard this driver in another design ( box was close though) i can tell i liked their sound. At 300euro/piece i find them a bargain imho.
* in the passive tannoy s600 example, the xover can be seen as underpar... until you focus on Tannoy preference for low part count on passive xover design and use of the 'natural' behavior of transducers ( shallow slope using 'natural' transducer behavior)... as a results they have choosen to compromise on the 1,5khz suckout. This open to another model in the range 'improving' on this and open to other sales...
** please read on the series of article on how to handle the issue with iir filters:
https://greatplainsaudio.com/gpa-vintage-altec-ezine/altec-lansing-duplex-speakers/
Last edited:
What's the point of a forum, if asking questions is going to be held against me? That's just silly!
At the end of the day we're all playing a game. In this case the game is designing speakers. I won't credit or blame you for the result!
Weight isn't an issue for this project. But you remind me; I have a certain memory of trying to put a large and heavy active speaker on a speaker stand by myself. It was not smart, but both the speaker and I survived!
You are clearly missunderstanding my advice. Good luck!
Wanted to point out advantages of smaller midrange, but you do not need that, obviously.
When you talk about the criticisms other people have about the Tannoy monitors, I can't help but think of the classic NS10. Nobody thinks the NS10s sound good or that that they are good speakers in objective tests, but everyone has to have a pair! If everyone knows the NS10s measure poorly, why is it particularly bad if the Tannoy doesn't measure perfectly? Certainly it has plenty of advantages over other speakers of it's day. I was previously unaware of the Tannoy coax design, but I am a little intrigued now!
I know that many "classic" hifi speakers don't measure very well based on modern standards. They break all the basic design rules of modern speaker designs, but they're still desirable and enjoyed by many.
I ask about drivers because I don't want to buy the cheapest driver only to find that it is unusuable and unfixable. I've been disappointed a few times! In most cases I have learned to actively avoid buying the cheapest version of something, since often the product is cheaper because of a deficiency that becomes glaringly obvious after the first use!
Of course I have a lot to do before I need to buy drivers. I will continue to do research. The links are appreciated!
Before I do anything else I need to test the cabinet I already put together. Is the actual performance of the speakers matching the simulation? Is the bass directivity control working? Over what frequency range? Does this directivity control improve in room response? And there are more questions. Of course I don't expect to build a complete set of speakers over a weekend and expect everything to be perfect without testing first!
If you had said I should use a 6" midrange plus a traditional tweeter, then I would say that is my original idea! I had a 6" driver picked out that matches my 12" woofers quite nicely! But instead you said I should go with a proven design, which comes across as condescending and unclear at best.
This discussion about achieving controlled directivity in the upper midrange by using a 12" coax for midrange and HF has caught my imagination, and I want to try it out. It's a new idea to me, so of course I'm asking questions. I won't dive into this completely blind; I can test the principles and get a good idea of effectiveness before I buy anything or commit to anything beyond the original idea.