Oh, one slightly possible and at the same time rule braking configuration would be running 2x7" mids in resistance cabinet down to 80-100Hz range and finish top with Esotar or some other tweeter. This would definitely need 2x7" or even larger mids to achieve. Below 80-100Hz could be handeled with single dipole or cardioid subwoofer, bit like many commercial 2-way + sub systems, only in directive mode.
Jussi
Jussi
Per Skaaning also hinted about 5" cone Flex (137cm2, 7" overall) unit with 92dB efficiency, 94dB with 4 ohm coil. +-3mm linear, suspension freely selectable...
I thought of very low loss suspension. Efficiency would be nice. But would it still give that much more compared to standard C-Quenze 7"?
Jussi
I thought of very low loss suspension. Efficiency would be nice. But would it still give that much more compared to standard C-Quenze 7"?
Jussi
Jussi said:
I'd ask Per (or a rep) about 1 vs. the other.. just from the website it says the only two significant factors are:
1. The frame is better on the Flex
2. No User T/S params adjustments
I doubt the frame will make much difference for this type of driver UNLESS you decide to go dipole (and even then there might not be much difference).
The adjustable T/S however may well be worth it.. Consider that you could adjust fs and Qts to a relativly high value that may be more appropiate for either an acoustic resistance box or dipole. Furthermore because of the design VAS shouldn't be much of a consideration and inductance won't be serious issue either (so you could increase force while be unconcerned with extended bandwidth past a nominal 2 kHz). I think that you could effectivly tailor the driver to achieve baffle-step compensation naturally, have increased eff., and limit lower freq. response naturally which should also limit distortion due to excursion. (..and there are other reasons why decreasing excursion is a good idea.) Additionally, you might be able to spec. a base-line for acceptable THD and IMD (freq. dependent) vs. power handling.
Note that none of their standard units (either the flex or the C-quenze) really meet the above paramaters.
Check out the post ScottMoose provided for Martin King's new modeling spreadsheet:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=69760
In particular note the freq. responses with respect to baffle step (corrected vs. uncorrected). Now I don't know if Martin will have modeling for Acoustic Resistance designs or not, but if sp - loss is anything like it is for normal bafflestep then you have a lot of "padding" down to achieve a moderatly flat response in a normal room. Of course with an active configuration it will be much less detrimental to "pad", but I can't help but think that a more "natural" soloution would be better.
Some other things to think about..
I'd tell Per the amount of stuffing/resistane near the driver, and see what he has to say abouit this for your design when used with the esotar (especially considering tonal character). (..and tell him your plan in general. including midrange bandwidth, active crossover and slope, amp output impeadance, desired power handling, etc.)
Though I don't know, I'd suspect that suspension and cone material changes would be considerably more expensive. (.. something I'd probably "steer clear of" unless the cost was nominal.)
(and though its obvious.. I'd also try to spec. matching within almost +/-.5 db.)
I don't think I'll start customizing drivers that far. It just makes them unusable in different application and very difficult to sell if necessary in the future.
AT seems to be pretty popular choise with Esotars. I guess they are as close to Dynaudios own midwoofers as possible, or possibly even better.
8" AT units start to be very expensive so I'd have to settle for a 7" model from their selection.
Overall I think I'll be experimenting some with my present MTM prototype and then return to MT if it doesn't work. Cabinet has flush mounted positions for Seas 18cm basket, shame they don't have that many options for that one. W18EX is still the strongest candidate.
Jussi
AT seems to be pretty popular choise with Esotars. I guess they are as close to Dynaudios own midwoofers as possible, or possibly even better.
8" AT units start to be very expensive so I'd have to settle for a 7" model from their selection.
Overall I think I'll be experimenting some with my present MTM prototype and then return to MT if it doesn't work. Cabinet has flush mounted positions for Seas 18cm basket, shame they don't have that many options for that one. W18EX is still the strongest candidate.
Jussi
Hey, whats wrong with Vifa XG18? Zaph measured very good midrange distorsion from them and they cost a lot less than AT or Seas Excel stuff. Do we have a black horse in this race?
Strangely John mensioned in some thread that L18 has more midrange detail and transparency while L18 had a bit more distorsion in that area.
Single XG is pretty inefficient but a pair could do it. Lower some THD as well.
Jussi
Strangely John mensioned in some thread that L18 has more midrange detail and transparency while L18 had a bit more distorsion in that area.
Single XG is pretty inefficient but a pair could do it. Lower some THD as well.
Jussi
Jussi said:
I thought about them as well.. MUCH better match with esotar than the seas mag. drivers (for the diaphram material). Like you said eff. is was put me off them, in addition less expensive drivers often sound less expensive (aka worse) despite their measuring better (..don't know about it in this instance though). For instance Zaph plainly states that the cheaper Seas tweets not only measure better but sound better than the millenium tweeter, whereas any number of people who have heard both would disagree. So you never know whats going to work out in the end.
If your thinking about it I'd say try one driver first and then purchase the other 3 if it works for you (or purchase 2 to start with and sell them on Ebay if it doesn't work out).
Considering your design, and all the stock drivers I've gone through, my pick would be the Fostex (though I'd probably take a black magic marker to the diaphram). Additionally I'd only use one driver and take the eff. hit with the active crossover.
http://www.madisound.com/pdf/fostexdrivers/fw168hpx.pdf
Here is a review of the lower "q" version in a speaker:
http://www.6moons.com/audioreviews/tonian/monitor.html
I've also been considering a WWMT dipole/cardioid system. I have four 25W/8565-01 Scan Speaks and while they don't that well deep bass in dipole they might be good in upperbass-low mid area. In this app I thought of those custom Flex drivers 300-3000Hz or so. I'll attach here one of the final scetches of this idea. Midrange could be in totally open space there like drawn or use resistance cabinet. Overall appearance would support ribbon as a tweeter but I don't know any reasonably prices HQ ribbon for the task. Naturally this can be replaced with Esotar in the top.
This solution has some very deeply considered technical details in the baffle and overall structure. I've spend pretty considerable amount of time thinking this thing through.
Naturally that kind of system needs at least a subwoofer to assist below 40-60Hz. 25W can propably handle 40-60Hz up to 250-350Hz, midrange (for example custom Flex) 250-350Hz up to 2500-3500Hz and some tweeter up there to finish.
What do you think?
I've been doubting this design while I need a high efficiency midrange for that single driver full open dipole midrange and my primary choises like Excels just don't have the sensitivity for it. PHL and some other PA brands have a few but still I don't want to make compromisses with the quality. AT 5" cone Flex is the strongest here but I still don't know that much about them.. Should I take the chance.. Expensive misfire at least if they don't perform well.
Jussi
This solution has some very deeply considered technical details in the baffle and overall structure. I've spend pretty considerable amount of time thinking this thing through.
Naturally that kind of system needs at least a subwoofer to assist below 40-60Hz. 25W can propably handle 40-60Hz up to 250-350Hz, midrange (for example custom Flex) 250-350Hz up to 2500-3500Hz and some tweeter up there to finish.
What do you think?
I've been doubting this design while I need a high efficiency midrange for that single driver full open dipole midrange and my primary choises like Excels just don't have the sensitivity for it. PHL and some other PA brands have a few but still I don't want to make compromisses with the quality. AT 5" cone Flex is the strongest here but I still don't know that much about them.. Should I take the chance.. Expensive misfire at least if they don't perform well.
Jussi
Attachments
That's a really impressive design Jussi.
I don't think you need to be concerned about the 5" AT drivers.
IMHO, world class performance in all respects (I use the Cquenze 5").
You can get some further opinions and test results from looking up the Peak Consult, Sonus Faber, & Rockport designs at the stereophile website. The Esotar or Morel Supreme 110 is a good match.
I don't think you need to be concerned about the 5" AT drivers.
IMHO, world class performance in all respects (I use the Cquenze 5").
You can get some further opinions and test results from looking up the Peak Consult, Sonus Faber, & Rockport designs at the stereophile website. The Esotar or Morel Supreme 110 is a good match.
Thanks David!
Personally I'd like to have more sophisticated equipment to study things but DIY builder just can't have it all. So most details in this concept are designed while they feel good and right. All drivers have separated parts, I thought to place midrange on top of wooferbaffle with three spikes so it just floats there. And tweeter on top of mid with same method. Woofer panel has floating panel inside it and woofers are attached to it. So minimum potentially radiating surface to the outside and maximum absorbing area inside. That inside baffle is placed in 8mm+ bitumen layer all around it. Spine also has bitumen core so it should absorb woofer vibrations pretty well. 25W is also pretty low mms woofer so it shouldn't shake things that much.
But all these are just assumptions. Would need a gear with accelerometer, time to make enough prototypes and test all this.
The custom Flex I asked would have very low loss suspension, 4 ohm coil, 94dB efficiency, 137cm2 cone, +-3mm linear. But I'm not so sure about that suspension. There are some drivers in the market that have severe problems with cone-suspension cooperation. W21EX001 papercone Excel is one of them, those distorsion graphs sure are ugly. So matching specific cone with specific basket with variable suspension, are all of the different suspension options stabile together... Somehow very low loss suspension combined with low Q efficient motor suits my idea of precision. Propably just a perversion with no real benefits. 😉
Jussi
Personally I'd like to have more sophisticated equipment to study things but DIY builder just can't have it all. So most details in this concept are designed while they feel good and right. All drivers have separated parts, I thought to place midrange on top of wooferbaffle with three spikes so it just floats there. And tweeter on top of mid with same method. Woofer panel has floating panel inside it and woofers are attached to it. So minimum potentially radiating surface to the outside and maximum absorbing area inside. That inside baffle is placed in 8mm+ bitumen layer all around it. Spine also has bitumen core so it should absorb woofer vibrations pretty well. 25W is also pretty low mms woofer so it shouldn't shake things that much.
But all these are just assumptions. Would need a gear with accelerometer, time to make enough prototypes and test all this.
The custom Flex I asked would have very low loss suspension, 4 ohm coil, 94dB efficiency, 137cm2 cone, +-3mm linear. But I'm not so sure about that suspension. There are some drivers in the market that have severe problems with cone-suspension cooperation. W21EX001 papercone Excel is one of them, those distorsion graphs sure are ugly. So matching specific cone with specific basket with variable suspension, are all of the different suspension options stabile together... Somehow very low loss suspension combined with low Q efficient motor suits my idea of precision. Propably just a perversion with no real benefits. 😉
Jussi
Member
Joined 2003
Jussi,
That is truly inspired styling...all I can say is WOW!
I imagine construction would make a very big mess in your workshop!
Regards,
Paul
That is truly inspired styling...all I can say is WOW!
I imagine construction would make a very big mess in your workshop!
Regards,
Paul
Paul,
I have a friend in local carpentershop with CNC goodies. But still I had to find out what he can do and what he can't and then hit my ideas to paper.
Personally I don't know does that look exceptionally good with all its exotics or is it just exceptionally ugly. Still balancing between... 😉
Tried to pack some crazy electro-acoustic visions with slightly different overall appearance. Unfortunately I don't have the means to prototype all this and many things have to be made guessing but hopefully I can analyze the final product detailed enough and see if I achieved at least some of the ideas I searched for.
Scott,
Earlier you mensioned normal dynamic drivers are overdamped and need low damping amp to drive it. Would this balance be different if driver is equipped with very low loss suspension?
Shame I'm not so sure about it's benefits. Common knowledge serves the idea of low loss with strong motor but if it causes cone edge resonances, unmatched suspension-cone problems or uncontrolled standing waves in the cone it surely isn't worth it.
Jussi
I have a friend in local carpentershop with CNC goodies. But still I had to find out what he can do and what he can't and then hit my ideas to paper.
Personally I don't know does that look exceptionally good with all its exotics or is it just exceptionally ugly. Still balancing between... 😉
Tried to pack some crazy electro-acoustic visions with slightly different overall appearance. Unfortunately I don't have the means to prototype all this and many things have to be made guessing but hopefully I can analyze the final product detailed enough and see if I achieved at least some of the ideas I searched for.
Scott,
Earlier you mensioned normal dynamic drivers are overdamped and need low damping amp to drive it. Would this balance be different if driver is equipped with very low loss suspension?
Shame I'm not so sure about it's benefits. Common knowledge serves the idea of low loss with strong motor but if it causes cone edge resonances, unmatched suspension-cone problems or uncontrolled standing waves in the cone it surely isn't worth it.
Jussi
Jussi said:
Looks real good!
I think that limiting the output of the mid (electrically or acoustically) is a good idea, and 250 Hz is certainly better than 150 or less. (..to me its not an issue with distortion.. but rather excursion vs. that of the tweeter.. if they don't "stroke" relativly the same distance during normal use then timing/intellegibility sounds "off" when spl's are increased. In fact this is the reason I generally prefer 1st order crossover - not because of phase, but because they have more excursion which usually mimicks the behaviour/stroke of the midrange.)
Before creating the mid's "shell" I'd try it with a mock-up vs. the dipole pole configuration.. This isn't a concern with radiation/directivity, but rather with acoustic resistance near the driver which tends to "suck the life out of the driver" (..as many others have said). I think Martin King said he was looking into this.. trying to discover the reasons related to the detriment in sound (..no doubt something to do with slow release stored energy). Of couse If your not using "stuffing" near the driver then it probably isn't as much of a concern.
40 Hz sounds like a good cut-off to me for the midbass section, but considering the amount of loss due to dipole configuration you'll likely need more than 25 watts for compensation with headroom.
I'd only look at the flex if your sure your going to use the esotar OR another tweeter like it (from seas or scan speak). IF you decide to save the esotar for another project (or sell for a premium on Ebay or Audiogon) then I'd probably go back to Accuton for both the midrange AND the tweeter - specifically the C90-T6 mid and the C24-6 tweeter.. Expensive, but no doubt worth it. You won't have tonal problems with this pair (at least none related to diaprham material), nor has it been my experience that you would have tonal problems with dissimilar diaphram material between the mids and a dipole midbass driver (when crossed over near the baffle step region). As to my preference one way or the other.. 🙄 it would have to be the Accutons.
Jussi said:
This is freq. dependent. It COULD be different near a driver's fs (better or worse), but it likely WILL be worse (overdampened) well above the driver's fs.
Furthermore, as you have pointed out, just what is that "low loss suspension" really doing? What's the positive attributes, and the negative? Generally when a manufacturer states this "buzz" phrase he is saying that the surround is compliant (non-stiff).. and possibly the spider as well. It also virtually gurantees that the surround will be made of rubber - which is great for longevity, but has some problems - not the least of which is being "dense" and NON-COMPLIANT with very short excursions (which is why eff. (and fs) usually drops with the use of a rubber surround). It also beg's the question: how is the VC being kept centered (..though with AT's reputation this is something they have probably "mastered" and its something I wouldn't worry about from them).
To your design specifically.. all it really means is a driver with a higher (than normal) qts (..which should be better in an acoustic resistance box).
Bottom line:
It PROBABLY is just marketing, and shouldn't be "bought" into generally.
Hans L said:
Effectivly, NO. (no AES article)
this is a subjective impression I've formed over time.. it could of course be flawed. (..but I wouldn't have stated it if I thought that was the case.)
While not exactly my favorite artist (hardly ever listen to him), Rod Stewart's: You're in My Heart. is absolutley fantastic for determining this problem (along with actually looking at the drivers while the song is playing). The song essentially adds instrument after instrument as it progresses.. becomming more and more complex. With loudspeakers that have an otherwise orderly soundstage and image ability at low volumes - this song maintains discreet images. Increase the volume though to the point where the tweeter isn't visibly stroking but the midrang/midwoofer is and *I* have always found that the sound starts becoming "homogenized" and less "orderly".
Its a fun easy test.. if you have the track and an otherwise good speaker that has this trait, try it out, you might find that you agree.
Scott,
Unfortunately C2-90 are out of my range. And I don't know any more of them that I know of AT drivers. Manufacturers own measurements sure but at least response is smoothed. And the price is that high so I can have a pair of W18EX on each side and still get away cheaper and they can play 102dB with below 0,3% THD, don't think single C90 can match that.
I think both crossoverpoints can be made with a 2nd order slope. 1st order would be nice but makes response to different angles way too unstable. And using a dipolemid without electrical highpass (just 6dB/oct dipole roll off) makes it move pretty dramatically since it's excursion is still rising 6dB/oct towards bass. But this is one thing I have to examine if this think comes operational.
40Hz is pretty much the bottom I ever expect 2x25W to play in dipolebaffle. Power shouldn't be that big issue since usable SPL range isn't that high. Propably more than 25W in action but still a very reasonable amount. Naturally continuous power is very small.
So you think I should ditch my MTM idea and go for this futuristic monument?
Jussi
Unfortunately C2-90 are out of my range. And I don't know any more of them that I know of AT drivers. Manufacturers own measurements sure but at least response is smoothed. And the price is that high so I can have a pair of W18EX on each side and still get away cheaper and they can play 102dB with below 0,3% THD, don't think single C90 can match that.
I think both crossoverpoints can be made with a 2nd order slope. 1st order would be nice but makes response to different angles way too unstable. And using a dipolemid without electrical highpass (just 6dB/oct dipole roll off) makes it move pretty dramatically since it's excursion is still rising 6dB/oct towards bass. But this is one thing I have to examine if this think comes operational.
40Hz is pretty much the bottom I ever expect 2x25W to play in dipolebaffle. Power shouldn't be that big issue since usable SPL range isn't that high. Propably more than 25W in action but still a very reasonable amount. Naturally continuous power is very small.
So you think I should ditch my MTM idea and go for this futuristic monument?
Jussi
I'm intruiged by your idea, but have trouble accepting it with my current understanding of speakerdesign. If anything, low excursion is a good thing. If you stress a tweeter to obtain a higher excursion, more in the line of the mid, you will get more distortion. I could a agree with the idea that an equally distributed distortion signature accross the mid and tweet may be benificial compared to one driver performing very cleanly and one distorting more. That is also a wide guess that has no technical backing. Your idea would mean that a second order acousitc roll off would be 'perfect', because it has constant excursion below the xo point assuming a flat input signal. But it would also depend on the xo point and the HP on the mid. I'll try to find the tracks you mentioned and see what I hear... who knows 😉ScottG said:
Jussi said:
Monument (yea or nea).. thats up to you, I will say that I think it offers much more potential IF the midrange is a dipole (with an appropriate bandpass and eq. - active of course).
As to the original design.. guess what? I managed to find a low distortion 8 inch (directivity challenged) that is poly based, has very low distortion (perhaps comperable to a seas unit when factoring eff.. possibly better), an eff. of 90 db across your prospective midrange (100-2kHz), quite a flat freq. reponse from 4-1.6 kHz, high qts. AND a relativly high excursion, and finally - not horribly expensive (though not cheap either).
It has one "quirk" though - its a coaxial unit (and as far as I know its only sold with the central tweeter), of course there would be no need or desire to use the tweeter. Other than this "feature" it seems almost ideal for what you were looking for originally.
Its the Beyma 8BX
http://profesional.beyma.com/ENGLISH/producto.php
(under the studio coaxial section, you'll see the pdf on it.)
frankly if you want to go back to the original design - IMO this driver is a "No-Brainer".
Hans L said:
Note the freq. range where I've found problems is the 2-4 kHz (primarily). I.E. a planar tweeter with crosssover at 2.5 kHz and a visibly "pumping" midrange should let you hear what I'm describing without difficulty. If not.. well then I'll need to go back and re-think this..😉
- Status
- Not open for further replies.