Maybe this 15" coax ? Expensice however. Another view of controlled directivity. Maybe Krivium could coment that design ?
Don't know if that helps.
Don't know if that helps.
Hi Diyiggy,
In my view, except for the driver size and open baffle it is spot on in fact! I just quickly looked at the pdf and it's a well thought design which seems nicely depicted.
The coax i've never heard. It's the 'big' B&C, Mark100 have one and used it with it's 'modular' approach ( a sub and top). I think he described it as having a more diffuse hf range than it's usual Synergy top and this was the most obvious difference.
Anyway if interested in it: Mark is very friendly and kind so asking him directly is the best thing to do imho. And he have a lots of reference to compare to.
Fredygump,
The 'modular' approach ( to have multiple box) has a lot of advantage if you want to compare different solutions for mid high with same low end. And once you sorted out issues related to baffle dimension ( edge diffraction, bsc needed) it is fast and easy to adapt to something new/different.
I don't think Adason as condescending tone, he is not 'wordy' and goes straigth to the point in his own way.
Not to forget he is a well known member with a lot of experience ( he probably tested most of what was possible within his preference and system- which at a point was modular too from the pictures i've seen).
Anyway the advice to select something already existing is not a bad advice for you to have a great result as outcome.
That said you already 'took the plunge' so now you are wet and in deep water... there is things to learn for sure ( we all still learn) but you seems motivated and willing, open minded... it should be ok imho.
About the 6" vs 12": i don't get why a 6" would be better. I'm serious and not trying to start a flamewar at all but would like to know your pov (@adason , Fredygump, or others).
The bigger sd of a 12" ( which for a 12"coax is close to a 10") will ensure much lower displacement and so less distortion for a given Spl.
The directivity behavior? Well maybe to have around 1khz under (directivity) control doesn't please everyone, but if you are looking for monitors which are able to perform quality assesment of played material it is an advantage in a domestic room as it lessens requirements about acoustic room treatments.
I think you'll better understand my point with this doc and why i think 12" for a coax is good compromise ( mainly about the directivity behavior and at which frequency it happens):
https://www.bbc.co.uk/rd/publications/rdreport_1995_04
In my view, except for the driver size and open baffle it is spot on in fact! I just quickly looked at the pdf and it's a well thought design which seems nicely depicted.
The coax i've never heard. It's the 'big' B&C, Mark100 have one and used it with it's 'modular' approach ( a sub and top). I think he described it as having a more diffuse hf range than it's usual Synergy top and this was the most obvious difference.
Anyway if interested in it: Mark is very friendly and kind so asking him directly is the best thing to do imho. And he have a lots of reference to compare to.
Fredygump,
The 'modular' approach ( to have multiple box) has a lot of advantage if you want to compare different solutions for mid high with same low end. And once you sorted out issues related to baffle dimension ( edge diffraction, bsc needed) it is fast and easy to adapt to something new/different.
I don't think Adason as condescending tone, he is not 'wordy' and goes straigth to the point in his own way.
Not to forget he is a well known member with a lot of experience ( he probably tested most of what was possible within his preference and system- which at a point was modular too from the pictures i've seen).
Anyway the advice to select something already existing is not a bad advice for you to have a great result as outcome.
That said you already 'took the plunge' so now you are wet and in deep water... there is things to learn for sure ( we all still learn) but you seems motivated and willing, open minded... it should be ok imho.
About the 6" vs 12": i don't get why a 6" would be better. I'm serious and not trying to start a flamewar at all but would like to know your pov (@adason , Fredygump, or others).
The bigger sd of a 12" ( which for a 12"coax is close to a 10") will ensure much lower displacement and so less distortion for a given Spl.
The directivity behavior? Well maybe to have around 1khz under (directivity) control doesn't please everyone, but if you are looking for monitors which are able to perform quality assesment of played material it is an advantage in a domestic room as it lessens requirements about acoustic room treatments.
I think you'll better understand my point with this doc and why i think 12" for a coax is good compromise ( mainly about the directivity behavior and at which frequency it happens):
https://www.bbc.co.uk/rd/publications/rdreport_1995_04
Hello Krivium,
Thanks for your feedback on the B&C coax.
Yep Mark100 is a nice guy. I have not direct interest in it but it was more to add gas to op thougths.
Faital has something nice in 12" coax...110° aperture. Maybe a little too much in vertical dispersion but surely nice in horizontal as the usual biradial that open till 120° works fine in home areas. The ferite Faital 12hx230 is less expensive that the neodynium one.
Cheers.
Thanks for your feedback on the B&C coax.
Yep Mark100 is a nice guy. I have not direct interest in it but it was more to add gas to op thougths.
Faital has something nice in 12" coax...110° aperture. Maybe a little too much in vertical dispersion but surely nice in horizontal as the usual biradial that open till 120° works fine in home areas. The ferite Faital 12hx230 is less expensive that the neodynium one.
Cheers.
In my view 120* is too large for monitoring purpose ( pro quality assessment).
The early reflections might add something for recreational listening ( however this is preference related), if for working there is too much coloration (it is misleading).
90* is a good compromise, 60* improve on accuracy but might lead to too narrow a sweetspot for my preference.
The paper you linked describe some very nice attribute. I agree with most of the point he describe, only open baffle i would be cautious ( for monitoring purpose). It depend on preference and if you'll have musician with you: too much different of typical box loudspeaker might be misleading to them with frustrations as outcome.
The 12" driver size come from the protobox Fredygump gave dimensions in previous post.
Another point why i talked about the Bms 12c362 is it have a nearly constant directivity behavior from 1khz to 5/6khz ( edit:4,5khz).
https://www.diyaudio.com/community/...using-bms-coaxial-12c362.318459/#post-5332697
It is way cheaper in there than any concurent too at TLHP( including Faitals). It might be different on the other side of Atlantic though.
I've never heard the ferrite variant (12c262) but it seems to share same chassis and components ( and is even cheaper at 260€) could be well be very good too.
Kimmosto used the 15" equivalent in the big Taipuu iirc.
For a bit more than the 362 there is the 382 which is the next coax i will invest in.
The early reflections might add something for recreational listening ( however this is preference related), if for working there is too much coloration (it is misleading).
90* is a good compromise, 60* improve on accuracy but might lead to too narrow a sweetspot for my preference.
The paper you linked describe some very nice attribute. I agree with most of the point he describe, only open baffle i would be cautious ( for monitoring purpose). It depend on preference and if you'll have musician with you: too much different of typical box loudspeaker might be misleading to them with frustrations as outcome.
The 12" driver size come from the protobox Fredygump gave dimensions in previous post.
Another point why i talked about the Bms 12c362 is it have a nearly constant directivity behavior from 1khz to 5/6khz ( edit:4,5khz).
https://www.diyaudio.com/community/...using-bms-coaxial-12c362.318459/#post-5332697
It is way cheaper in there than any concurent too at TLHP( including Faitals). It might be different on the other side of Atlantic though.
I've never heard the ferrite variant (12c262) but it seems to share same chassis and components ( and is even cheaper at 260€) could be well be very good too.
Kimmosto used the 15" equivalent in the big Taipuu iirc.
For a bit more than the 362 there is the 382 which is the next coax i will invest in.
Last edited:
I sugested kef, because they know what they are doing. LS50 is little over 5". That's ideal midrange to me. Its been discussed, i will find the link. Kef optimized tweeter to midbass transition to minimize ripples. Its well measuring great sounding coax. My hiking buddy bought it after i told him about it. He is happy with kef. With sub (or two) you get top notch sound.
https://www.diyaudio.com/community/threads/current-best-5-midrange-driver.378212/page-6
https://www.diyaudio.com/community/threads/current-best-5-midrange-driver.378212/page-6
Last edited:
I had big Tannoy, sold them. Had big B&C coax, gave it to musician friend for guitar practice. Had big open baffle hawthorne silver iris, sold. Had few korean coaxes. All gone to friends.
All had some issues. But i do not want to discourage anyone.
All had some issues. But i do not want to discourage anyone.
What is the off axis constant directivity that is good enough for home in stereo ? 45° should be more than enough if you are at the top of the triangle listening area ?
Well as said before it is not feasible with that big driver that ask big overlaps but perhaps low order slope for longer overlapp or without some help . Being horn or/and FIR.
The key point is many some liking big area drivers. Does the op needs pro mixing or home listening system (less near to the loudspeakers)? And a flat power response as one talk of voicing aka loudspeaker voicing ?
So what response curve target as already stated Krivium when talking of dome vs horn ?
Well as said before it is not feasible with that big driver that ask big overlaps but perhaps low order slope for longer overlapp or without some help . Being horn or/and FIR.
The key point is many some liking big area drivers. Does the op needs pro mixing or home listening system (less near to the loudspeakers)? And a flat power response as one talk of voicing aka loudspeaker voicing ?
So what response curve target as already stated Krivium when talking of dome vs horn ?
For my use, a 45 degree (actual) directivity would be enough for covering all listening positions. Most are rated for 80-90 or more, but I'm starting to notice that they are narrower than their rating implies, if you are expecting a deviation of only a few db.
I was looking at Faital's graphs vs their stated directivity, and their rating of 90 degree directivity looks more like 60 degrees to me. But I assume they are taking an average, and they are allowing a certain tolerance of spl variation to arrive at 90 degrees. I'm sure that this is true of every company, and Faital probably shares their data knowing it is better than other products?
The system is for home music listening. I don't do any mixing or professional work. But I have dabbled a little in pro audio, and I gravitate toward pro equipment rather than hifi.
I do like the idea of larger woofers. All the systems I have had in the past used small woofers, and for some time I have been gradually moving in the direction of wanting to try larger drivers. Maybe 10 years ago I had a really good quality 5.2 surround system (towers w/ 6" drivers), then I traded up for Genelec 8040As + 7060b. Then traded down for the M-Audio 3-8s with stereo subwoofers. I appreciate all of them, as different as they are. The Genelecs were technically impressive, but the M-Audio setup being a 4 way system is a completely different experience. If I can improve over the M-Audio system, getting more even (in room) bass response and better clarity and imaging, I'll be happy.
That said, I think 15" woofers are just too big. I started this project with the idea of using a sealed 15" driver plus ported 12", but the cabinet would have been too big.
I'm not entirely sure what the gain of a large coax in a 3-way is. In a PA system, it works as a point-source down to bass levels. A larger diaphragm cone has more output with less excursion, a higher sensitivity to match the tweeter/compression driver, and a higher maximum SPL. Subs to match, of course.
How are you controlling the midrange directivity? At 12" depending on the shape of the cone, mounting, etc. the woofer may beam as early as 800hz where the crossover may be as high as 1800. This is part of the "making your own crossover" that could break the whole system. You'll still end up with a narrow listening window. Smaller cones don't exhibit this and can be crossed higher before they break up and distort, too.
This design is hanging on a conbination of ideas. It is just 1 of 2 possible ideas I have going at the moment!'
As discussed in this thread, it is believed this driver configuration of the rear facing 10" woofer and the front facing 12" woofer can produce low frequency directivity with proper DSP settings. Then the 12" coax woofer will exert increasing directivity control as it approaches the crossover point. And the horn will exert directivity control above the crossover point.
I don't know if these 3 methods of control are capable of full range directivity control. I'm not even sure that the low frequency directivity part will work when the cabinet is placed near a wall! I have been reading about cardiod subwoofer configurations, and it seems pretty clear that they are not very effective when they are not in free space.
Theory 2, or my original theory, is just a lot of DSP, but in a clever way.
Initially I wasn't conscious of bass directivity, and I'm still not sure that this is what Genelec is doing. One product interview with a Genelec designer talked exclusively about how they can control what frequencies are played through each driver so that the result is a flat response. If one driver creates a null at the listening position at a specific frequency, they EQ it down and use the other driver to cover that frequency. It sounds a lot like DSP for multiple subwoofers used in home theater setups.
In their case, they have 5 (or 6?) different drivers that they can control independently. So I think it is possible they can use the overlap in frequency range between these different drivers and the different spacing between the drivers to EQ out peaks and nulls in frequency response, and to do so over a wide frequency range.
But it's all just theories right now. It's a little frustrating that the DSP unit i ordered was DOA, because I am stuck here waiting for a replacement! If I had it in hand, I could begin to test theories to see what works and what does not work.
The gain are the one you listed, as well as lower cut off freq and as it is relative too control of directivity lower in freq.
How is controled the directivity of midrange? First we have to define midrange. In my world mid area is considered to be from 250hz up to 6khz.
https://images.app.goo.gl/dZoREhwkE3iLzmk27
So for low mid directivity control you boot on the natural directivity behavior of your woofer part ( defined by sd rather than overall diameter. Even truer for a coax) to define a point where the waveguide that is going to take into control of the next freq range ( hi mid) have his own directivity matching. This define an 'optimum' xover point from a directivity pov. The waveguide will then do his job from then.
For the low mid freq range the natural 'beam' is not the only parameter at play in directivity behavior: the box play a role too: the shape have an influence.
Width of course ( it define BSC) but the depth of enclosure too. Height as well on vertical directivity.
The basic principle is what gave birth to the 'econowave' school of thoughts implementing mostly two ways with a woofer and waveguide ( usually constant directivity) as 'Zilch' made popular in DIY scene.
It was not something new, Dr.Earl Geddes (i linked to his whitepapers in a previous message in this thread) as well as W. Parham ( Pi Speakers) based their own different approach on this principles ( i suggest to read W.Parham philosophy too, very clever approach too: https://www.pispeakers.com/Pi_Speakers_Info.pdf ).
Now the cross over issue... yes most coax have what you describe D1sco (and i suppose it is one of the points which bothered Adason too). There is a possible weak link in a (limited) range overlaping xover freq. It is not easy to deal with this issue in passive ( but if -imo- correctly implemented it can solve the z offset at the same time... see the great plain audio series of articles i linked previously -i didn't say but the series start with the page linked and when you scroll at right for next article the author present it's way to handle it).
Going active analog filtering makes it a step easier to implement and an order of magnitude better for my own preference.
With dsp the issue is greatly minimised. And with FIR complementary high slope filters ( 48db/octave and up xover), eq and delay availlable i verified most nastys can be pushed beyond audibility ( break up). If you want to manipulate thrue inverting profile FIR it can be pushed even further...
I totally get it can not be the approach people wants to implement though ( multi amp/dsp).
I don't get your comment about a 'small listening window'. In my experience with coax, the issue you describe gives the opposite artefact: a 'waistbending' in the polar map ( as described in E.Geddes paper about loudspeakers directivity).
In fact when you face this 'mismatch' what happen is the woofer is already well 'beaming' but the waveguide doesn't already have control over wavelength, so it radiate more omnidirectionaly. As a result you'll have an higher intensity 'beam' centered on listening axys which add to a more wide spread one.
It makes the loudspeaker 'bloom' when it happen ( the Harbeth's afficionados usually like this effect which happen with them too). The bigger diameter coax makes this happen in a range which is much less an issue than smaller diameter's one.
With a 15" you can expect it in the 1khz range. For 12" around 1,2khz.
Some likes this effects some don't. I'm in the don't like it camp and if it can be minimised i happily use tools to do.
Anyway as a generalisation with coax, the (nearly) coincident location of the two way produce a stable and wide image.
If there is a narrowing feeling of listening spot imo, it comes from listening them on axys: as with all waveguide based system there is 'issues' on axys. But they are not supposed to be listened this way: 15/20* and almost all issues are gone ( and this this usually true for a wide angle if the loudspeaker box/driver is not plagued by diffraction).
The design goal must be adapted from on axis focused approach to power response/ directivity smoothness focus imho.
Last edited:
I read an article discussing this directivity in horns in a 2 way, and it used as an example a 60x40 degree horn. The article said what you are describing, that even though it is the horn playing, the directivity control does not immediately take effect. So in reality it is 3khz in one axis directivity fully takes affect, and 4khz in the other? (I don't remember what the actual numbers were.)
So the "bloom" you describe is the frequency below the range where the horn design produces directivity?
I'm just wondering if this effect is minimized or different in a coaxial with a symetrical pattern compared to a horn? The data I looked at for coaxial drivers usually starts around 2khz, and they appear to show a very consistent polar map above 2khz. But it also looks like the beam width may be beginning to widen below 2khz. But unfortunately they do not share this data.
The data for the Faital Pro 6" coaxial shows a dramatic widening of the beam width below 2.5khz, but this appears to be more controlled on the 12". Of course we don't know what happens below 2khz, so I'm curious what your experience is!
What you describe is known as 'pattern flip'. Its related to the fact the waveguide is wider as it's high (height is shorter). It happens because of the different dimension.
Most coax 'Tannoy' style doesn't have this effect because theyre waveguide are cone shaped ( constant dimensions). The price to pay is on axis abnomaly...
Altec style ( with protubering horn ) can exhibit this though.
You only see from 2khz because you only studyed Faital's datasheets. 😉
Beware the polar plots are presented in a way they are beautiful to look at ( in other words there is a lot of smoothing). That doesn't say they doesn't soundgood though. Look for independant measurements, they exist.
I've already told you about my experience which model i prefer and mostly why. 😉
I can't talk about the Faitals more than i did because i've not played with them ( i said probably heard them because we used to rent gear in a company who had some, and i probably used them as i don't open all box i meet! 🙂 ).
Most coax 'Tannoy' style doesn't have this effect because theyre waveguide are cone shaped ( constant dimensions). The price to pay is on axis abnomaly...
Altec style ( with protubering horn ) can exhibit this though.
You only see from 2khz because you only studyed Faital's datasheets. 😉
Beware the polar plots are presented in a way they are beautiful to look at ( in other words there is a lot of smoothing). That doesn't say they doesn't soundgood though. Look for independant measurements, they exist.
I've already told you about my experience which model i prefer and mostly why. 😉
I can't talk about the Faitals more than i did because i've not played with them ( i said probably heard them because we used to rent gear in a company who had some, and i probably used them as i don't open all box i meet! 🙂 ).
Last edited:
The 'bloom' is a compromise ( the term is not mine). Driver dependent in case of coax, it is not something unusual on other type of loudspeakers inclluding direct radiators ( my main threeway have this).
That said you might like the effect and if you do not then the paper on CID i linked before gives an alternate answer to absorbing panels on 1rst reflections points on ceilling and wall ( implementation of a kind of RFZ).
That said you might like the effect and if you do not then the paper on CID i linked before gives an alternate answer to absorbing panels on 1rst reflections points on ceilling and wall ( implementation of a kind of RFZ).
Again, I do appreciate all that you have shared! And I expect it will help others as well, as I have already seen this thread turn up in internet searches as I continue to search for more information.
What I meant in my previous comment was simply that the polar map provided by Faital is the most informative I have seen, but it is still incomplete since it starts at 2khz.
The data I find on directivity seems limited, or more precisely, I have not found the data that would answer the question asked by D1sco.
This is for anyone who would like to contribute: Pleae consider these polar maps as if they represent a typical coaxial design. One is a 6" coaxial and the other a 12" coaxial.
This image is the 6". The crossover is meant to be at 1.7khz, so the crossover is below the range of this map. But clearly there is a significant change in pattern width between 2-3khz. What is happening here? Is the 6" cone too small to create controlled HF under 3khz
Here is the second one. This is a 12" coaxial. It appears that the pattern might be widening at 2khz, but we don't know what happens next. Assuming a crossover of 1.2khz, what is the high frequency pattern doing at 1.2khz? Is it suddenly omnidirectional? Or does it gradually widen in proportion to frequency?
Think the cost of quality woofer, horn systems, then add for the lower quantity sales, higher machining costs.
IME you're 'screwed' 😉/🙁, though experimenting a bit with helping others mod their Avatar/Adire 12" coax kits ages ago, the driver/horn needed a lot of tweaking to get 'acceptable' performance (by the owner's standards, never got to audition one) and I use the term loosely since the best coax (or point source speaker system for that matter) I've auditioned is the Altec 604B w/1 kHz XO, huge multi-cell tweeter driven with a high output (~matching) impedance tube amp of the times with variable DF bass/treble tone controls.
Second best was a too briefly owned 604E (w/DIY foam horn extension) MLTL system driven with McIntosh 300B tube system with bass/treble tone controls.
In short, there's a certain synergy with tube driven horn systems, so regardless of coax, recommend driving the horn with a high output/matching impedance tube amp and a Class A SS amp for the woofer to get the best out of both.
I’m a big Tannoy lover. Im using the System 1000 now and have to build some cabinets for the 3139 drivers I have from a System 1200 that I will bi amp. The DMT drivers are great too but are very hard to find. Countless albums have been mixed on the Tannoys over the years. They are designed for mid field listening not for a huge room. They have fantastic imaging and are fairly flat in response. Coaxes have a magic when done right that no other speaker has, point source coincidental drivers. Tannoy was the first and best in my book. I run them with SET tube amps and they make me smile every time!
Thanks for the input. The pictures of the 604E are interesting in light of what Krivium said about "pattern flip". The directional horn is divided with square slats, which I expect are designed to prevent this pattern flip.
My concern is more about the fact that the available coax drivers are carefully optimized for something completely different than what I want to do. Making a good PA horn is not necessarily the same as making a good near field/ mid field horn. That's why I keep asking questions!
Do you know of pictures of the Avatar/ Adire 12" kits you mention? I didn't find anything immediately with an image search.
Last edited:
That's the spirit! I want to cover my bases so that the design "works" without major deficiencies, and then enjoy what it does well rather than measure how far away from perfection it is.
With this talk about Tannoys, I should keep my eyes open for a pair. I didn't know about coax studio monitors before I started with this idea.