You're welcome.
Depends on the ultimate design really. In an ideal world, you don't want an XO in the telephone band or an octave either side of it at all. But we don't live in an ideal world.
Will you hear the transition? Maybe. Maybe not. More pertinently -would you care? That depends on what you happen to focus upon when listening. We all basically hear the same thing (physiological losses aside) but we tend to listen differently & value specific things. Those who value total midband transparency & tie themselves up in knots at the suggestion of anything at all other than 1 driver even near that region are unlikely to be thrilled. Although I suspect it's more the idea of having another driver in that region than the sonics they'd object to. OTOH, people who place value on low-distortion, high SPLs & massive LF transient swings at ~live levels won't be happy with one driver, however it's loaded, period. They'll want that extra capacity in the LF from a big woofer.
FWIW, (& assuming the design allows for it!) 500Hz is probably one of the least evil places within the midband to cross, for reasons I've noted above. View it as Credit / Debt sheet. Which side the balance falls upon is down to individual preference. Minimal interference is how I'd tend to categorise it, with more advantages in some situations (assuming you aren't obsessed with total transparancy across the entire midband irrespective of the cost) than disadvantages. If it worries you & you can sacrifice some power-handling, you can always cross lower, below 200Hz, should you wish. Both work well, they just have different compromises.
Looking at your rough requirements, the 10in or 12in suitably loaded should do you for mains in a room that size. I'd be inclined to go with the former, for simple size reasons as the 12in really would prefer a larger box than most people can stomach to start giving of its best, but YMMV as always.
Depends on the ultimate design really. In an ideal world, you don't want an XO in the telephone band or an octave either side of it at all. But we don't live in an ideal world.
Will you hear the transition? Maybe. Maybe not. More pertinently -would you care? That depends on what you happen to focus upon when listening. We all basically hear the same thing (physiological losses aside) but we tend to listen differently & value specific things. Those who value total midband transparency & tie themselves up in knots at the suggestion of anything at all other than 1 driver even near that region are unlikely to be thrilled. Although I suspect it's more the idea of having another driver in that region than the sonics they'd object to. OTOH, people who place value on low-distortion, high SPLs & massive LF transient swings at ~live levels won't be happy with one driver, however it's loaded, period. They'll want that extra capacity in the LF from a big woofer.
FWIW, (& assuming the design allows for it!) 500Hz is probably one of the least evil places within the midband to cross, for reasons I've noted above. View it as Credit / Debt sheet. Which side the balance falls upon is down to individual preference. Minimal interference is how I'd tend to categorise it, with more advantages in some situations (assuming you aren't obsessed with total transparancy across the entire midband irrespective of the cost) than disadvantages. If it worries you & you can sacrifice some power-handling, you can always cross lower, below 200Hz, should you wish. Both work well, they just have different compromises.
Looking at your rough requirements, the 10in or 12in suitably loaded should do you for mains in a room that size. I'd be inclined to go with the former, for simple size reasons as the 12in really would prefer a larger box than most people can stomach to start giving of its best, but YMMV as always.
Thanks for the assurance, Scottmoose.
And a concise reality overflight.
Roger least evil 500Hz.
Big time research tonight as this opens a legitimate front along with a new struggle with my single driver demons 
But it would save me hours if you gave a few sentences to expand on typically expected compromises.
What I can imagine a little better than crossing at 500Hz is the effect of crossing lower: such as for every 100Hz I go below 500 means a lot of acoustic energy the smaller driver would need to use up in the same range, causing it to be what - ? - Not just poorer in transients but possibly impacting what kind of enclosure can be utilized? HF extension... Seems in this case like a good job for the AN Cast 10. It already wants to roll off in the 90's sealed and has a surplus in the HF it can stand to lose. Might there be a way to force it to roll off a little higher without too much of a real crossover? That would be a plus.
vs
Staying at 500Hz and aside from some minor(?) coherency issues and the demand for a more robust and proper crossover - what ? - it will only be more wonderful? ...Tell me a little more?
And..
So actually, the ~3.3ft^3 is the "absolute" smallest... That begs for a drop of exposition. What is gained is a reasonable FR. Could you offer a rough Qtc equivalency or what is lost?
Perhaps I should ask, at what number of cu. ft. does the 12 become well clear of the term "absolutely" and enter into the realm of "pathetically" smallest. Or better
-jerry
And a concise reality overflight.
Roger least evil 500Hz.
Big time research tonight as this opens a legitimate front along with a new struggle with my single driver demons But it would save me hours if you gave a few sentences to expand on typically expected compromises.
What I can imagine a little better than crossing at 500Hz is the effect of crossing lower: such as for every 100Hz I go below 500 means a lot of acoustic energy the smaller driver would need to use up in the same range, causing it to be what - ? - Not just poorer in transients but possibly impacting what kind of enclosure can be utilized? HF extension... Seems in this case like a good job for the AN Cast 10. It already wants to roll off in the 90's sealed and has a surplus in the HF it can stand to lose. Might there be a way to force it to roll off a little higher without too much of a real crossover? That would be a plus.
vs
Staying at 500Hz and aside from some minor(?) coherency issues and the demand for a more robust and proper crossover - what ? - it will only be more wonderful? ...Tell me a little more?
And..
So actually, the ~3.3ft^3 is the "absolute" smallest... That begs for a drop of exposition. What is gained is a reasonable FR. Could you offer a rough Qtc equivalency or what is lost?
Perhaps I should ask, at what number of cu. ft. does the 12 become well clear of the term "absolutely" and enter into the realm of "pathetically" smallest. Or better
-jerry
Fostex FF225K has low Qts and roll off "early" at 175hz with -6db around 130hz in 4liter closed, but it will need a supertweeter above 10khz ... appears to be a pretty smooth middriver with very high sensitivity
Wonder if used in surround whether only main speakers need the additional woofer and supertweet and the rest could do with only the FF225K, and no xo on any of them ? ... well, then theres the BSC issue
btw, AN 8" CF seems to do exactly the same thing
Wonder if used in surround whether only main speakers need the additional woofer and supertweet and the rest could do with only the FF225K, and no xo on any of them ?
... well, then theres the BSC issue btw, AN 8" CF seems to do exactly the same thing
Hi tinitus,
Thank you for the variations on Scottmoose's theme.
What I like is the idea that if a crossover is used that it is in, to use his language which I like, "a least evil place." However, I haven't by any means given up on a totally single driver solution. What is certain though, is that now as I judge any single driver approach I want to be sure at the very least there's room in the design for augmenting woofers. (If anyone's frustrated and said, "It's about bloody time," I try to address that in the next post below.)
I understand a crossover above 6K or 7K or 10K would also be out of the way of critical listening but I feel if I were to use a crossover at all I'd only want to use one. Of the 2 options: high and low, I feel the HT and low level requirements make this choice for me - down low.
I have to look into the advantages and disadvantages of these things. But, gee, a separate amp for augmenting woofers? I wouldn't mind for later, but do you really think I can't use the 120W I have in SS now to drive everything?
-jerry
Thank you for the variations on Scottmoose's theme.
What I like is the idea that if a crossover is used that it is in, to use his language which I like, "a least evil place." However, I haven't by any means given up on a totally single driver solution. What is certain though, is that now as I judge any single driver approach I want to be sure at the very least there's room in the design for augmenting woofers. (If anyone's frustrated and said, "It's about bloody time," I try to address that in the next post below.)
I understand a crossover above 6K or 7K or 10K would also be out of the way of critical listening but I feel if I were to use a crossover at all I'd only want to use one. Of the 2 options: high and low, I feel the HT and low level requirements make this choice for me - down low.
I have to look into the advantages and disadvantages of these things. But, gee, a separate amp for augmenting woofers? I wouldn't mind for later, but do you really think I can't use the 120W I have in SS now to drive everything?
-jerry
To all,
I just also want to add that understanding what will and what will not so much take away from the single driver experience is clearly my fulcrum point.
That's why I starting a parallel thread on the matter: How much can you add to a single driver design before risking its unique qualities?
I hope any surprises and revelations there will allow me to make the best use of the efforts of all the people here. It's something I should have done very early however I didn't know if it would be appropriate forum behavior to do so. But weighing that against the risk of frustrating or otherwise exhausting your generosity I've decided to do what I can to protect it. Plus they are two legitimately different subjects.
Besides, the double research I was doing all along was getting to me anyway.
I just also want to add that understanding what will and what will not so much take away from the single driver experience is clearly my fulcrum point.
That's why I starting a parallel thread on the matter: How much can you add to a single driver design before risking its unique qualities?
I hope any surprises and revelations there will allow me to make the best use of the efforts of all the people here. It's something I should have done very early however I didn't know if it would be appropriate forum behavior to do so. But weighing that against the risk of frustrating or otherwise exhausting your generosity I've decided to do what I can to protect it. Plus they are two legitimately different subjects.
Besides, the double research I was doing all along was getting to me anyway.
This is what happens when things get taken out of context.
'One of the least evil places in the telephone BW.' Not 'least evil.' And that, as I've kept harping on about, is design dependant. Nothing is a panacea. You select the compromise that will work best for you & design a system based upon that; ergo if any XO is going to be done, where & how will be part of the design from the start.
Re your several Qtc queries, it's not a sealed box so the Qtc is somewhat moot. Probably around 0.4 in terms of its roll-off contour, but as I say, it's not a sealed box. WRT 'absolutely' to 'pathetically', that is so subjective as to make any answer I could give worthless.
As for crossing lower -lower you cross the less potentially audible the XO point (although you don't want it in the middle of a couple of instruments BW) but the harder the wide-band unit will have to work, potentially increasing distortion if it needs to do more than it's capable of.
'One of the least evil places in the telephone BW.' Not 'least evil.' And that, as I've kept harping on about, is design dependant. Nothing is a panacea. You select the compromise that will work best for you & design a system based upon that; ergo if any XO is going to be done, where & how will be part of the design from the start.
Re your several Qtc queries, it's not a sealed box so the Qtc is somewhat moot. Probably around 0.4 in terms of its roll-off contour, but as I say, it's not a sealed box. WRT 'absolutely' to 'pathetically', that is so subjective as to make any answer I could give worthless.
As for crossing lower -lower you cross the less potentially audible the XO point (although you don't want it in the middle of a couple of instruments BW) but the harder the wide-band unit will have to work, potentially increasing distortion if it needs to do more than it's capable of.
What do you call something only 30% bigger than "absolute" minimum? I thought it self-deprecating to call it "pathetic." But the humor was obtuse and wasted your time and I apologize for my poor editorial oversight. Got all-nighter punchy.
I've been probably misusing Qtc's to get a comparative sense of transient quality - a way of comparing apples to oranges. It would have been better to ask:
Would the resultant transient quality of this quasi aperiodic be better or worse than what you might expect from a sealed box of Qtc=0.6? (This being Dave's suggested target for this application.)
And if not at this minimal size, at roughly how much larger?
On the crossover at 500: An area of least interference, but I appreciate the delicacy of the situation when checking an instrument bandwidth chart. Still in the thick of things logarithmically speaking.
It won't help relieve the full range driver but is there a name to having a full time woofer run off a separate amp with a simple first order rolloff up top? No crossovers again. Both drivers operating. Would the interference be that much worse to 500Hz with both drivers operating rather than one fully crossed over? That is, avoiding a crossover to get some on-demand HT weight in exchange for sacrificing some quality on the main driver.
Thank you
-jerry
Hi VanJerry, compare Scottmoose's quasi-aperiodic MLTL design he made for you:
http://www.diyaudio.com/forums/showthread.php?postid=1566378#post1566378
With the crude 300-liter WinISD sim (yellow line) with Qtc of .66:
http://www.diyaudio.com/forums/showthread.php?postid=1548700#post1548700
The tradeoffs are in those graphs. Scott's design has the same (arguably better actually) performance than the gigantic sealed box, but in a vastly better-sized enclosure.
In terms of where to cross over from FR to helper woofer, pick any frequency, really. If you listen at low volumes, choose as low an xover point as you can tolerate. If you want high volumes, you'll be forced to use a higher xover point. You might end up adjusting it a bunch of times.
I realize you're just trying to understand the trade-offs. Ultimately, though, this subject has some mysteries which can only be unlocked with your saw!
http://www.diyaudio.com/forums/showthread.php?postid=1566378#post1566378
With the crude 300-liter WinISD sim (yellow line) with Qtc of .66:
http://www.diyaudio.com/forums/showthread.php?postid=1548700#post1548700
The tradeoffs are in those graphs. Scott's design has the same (arguably better actually) performance than the gigantic sealed box, but in a vastly better-sized enclosure.
In terms of where to cross over from FR to helper woofer, pick any frequency, really. If you listen at low volumes, choose as low an xover point as you can tolerate. If you want high volumes, you'll be forced to use a higher xover point. You might end up adjusting it a bunch of times.
I realize you're just trying to understand the trade-offs. Ultimately, though, this subject has some mysteries which can only be unlocked with your saw!
??? Absolute minimum is a driver suspended in free air from the ceiling. Minimum box size technically is one that seals the basket holes as done on compression drivers, sealed back mids, tweeters, i.e. minimum box size is the one that closest meets the needs of the app WRT its F3, F6, F10, etc. both on and off-axis.
Acoustic bass or similar players and folks who believe they can tell the difference between small variances in Qtc in the LF BW tend to prefer a 0.5 Qtc, which is the largest sealed box size and as it shrinks, its F3, Fb rises. Ditto TLs (AKA ~ 'aperiodic'). Not many 'free lunches' in audio system design.
A max flat impedance TL is considered the best trade-off between BW, impulse (transient) response and box size, providing a benign load to the amp, though due to the vanishingly low output impedance of an SS amp the latter is moot.
WRT where to put an XO point......... the pioneers of audio were only interested in reproducing the human voice BW as accurately as practical, so the majority of early audio was based on Bell Labs' human hearing research, not a musical instrument's BW chart, and when the available movie soundtrack BW had increased enough to require an XO point approaching our acute hearing BW, 300 Hz/2nd order was chosen and as more research was done it was bumped to 500 Hz as an acceptable trade-off to ensure that all of the most dynamic portion of the music spectrum (amp power wise) be covered by the woofers and later still, upped to 1.2 kHz in acoustically small venues. In the near-field of recording, broadcasting studios, 1.5-1.6 kHz became the norm for decades until electronic advances allowed any point to be chosen since the signal can be manipulated in the digital domain.
Point source (cone/dome) drivers can roll off its HF at ~6 dB/octave if it has zero inductance, so an impedance compensation (AKA zobel) filter is required, but in a subwoofer's BW the components are large, expensive and raises its Qts, so EQing its HF response is the usual choice.
Anyway, using the AN12's published specs and assuming a zobel is added, a max flat impedance TL that except for size meets all the criteria of reasonably extended LF BW to allow a low acoustic subwoofer XO point, ~critically damped roll off slope for excellent impulse response, benign impedance if using a high output impedance amp and when properly located in-room may not even need a sub system for the majority of movie soundtracks:
L = 85.468"
CSA = 335.121"^2
zdriver = 32.197"
All dims approximate and sim assumes 1.091 lbs/ft^3 of polyfil. Compared to a sealed version it has a much lower Q Fs impedance, ergo flatter phase and slightly better impulse response, faster roll off below Fs and slightly more gain BW above it. If SS driven I doubt you can hear these subtle differences in-room though since it dominates down low, but you can always seal up the terminus (port) to test it.
Note that if you end load the driver its impulse response further improves a bit with the trade-off of its response mirroring the sealed's down to Fs.
GM
Acoustic bass or similar players and folks who believe they can tell the difference between small variances in Qtc in the LF BW tend to prefer a 0.5 Qtc, which is the largest sealed box size and as it shrinks, its F3, Fb rises. Ditto TLs (AKA ~ 'aperiodic'). Not many 'free lunches' in audio system design.
A max flat impedance TL is considered the best trade-off between BW, impulse (transient) response and box size, providing a benign load to the amp, though due to the vanishingly low output impedance of an SS amp the latter is moot.
WRT where to put an XO point......... the pioneers of audio were only interested in reproducing the human voice BW as accurately as practical, so the majority of early audio was based on Bell Labs' human hearing research, not a musical instrument's BW chart, and when the available movie soundtrack BW had increased enough to require an XO point approaching our acute hearing BW, 300 Hz/2nd order was chosen and as more research was done it was bumped to 500 Hz as an acceptable trade-off to ensure that all of the most dynamic portion of the music spectrum (amp power wise) be covered by the woofers and later still, upped to 1.2 kHz in acoustically small venues. In the near-field of recording, broadcasting studios, 1.5-1.6 kHz became the norm for decades until electronic advances allowed any point to be chosen since the signal can be manipulated in the digital domain.
Point source (cone/dome) drivers can roll off its HF at ~6 dB/octave if it has zero inductance, so an impedance compensation (AKA zobel) filter is required, but in a subwoofer's BW the components are large, expensive and raises its Qts, so EQing its HF response is the usual choice.
Anyway, using the AN12's published specs and assuming a zobel is added, a max flat impedance TL that except for size meets all the criteria of reasonably extended LF BW to allow a low acoustic subwoofer XO point, ~critically damped roll off slope for excellent impulse response, benign impedance if using a high output impedance amp and when properly located in-room may not even need a sub system for the majority of movie soundtracks:
L = 85.468"
CSA = 335.121"^2
zdriver = 32.197"
All dims approximate and sim assumes 1.091 lbs/ft^3 of polyfil. Compared to a sealed version it has a much lower Q Fs impedance, ergo flatter phase and slightly better impulse response, faster roll off below Fs and slightly more gain BW above it. If SS driven I doubt you can hear these subtle differences in-room though since it dominates down low, but you can always seal up the terminus (port) to test it.
Note that if you end load the driver its impulse response further improves a bit with the trade-off of its response mirroring the sealed's down to Fs.
GM
Attachments
rjbond3rd said:
Yeah, where box size is the primary goal, 'forcing' an under-damped vented alignment into a well damped one is the way to go IMO, but then you're usually forced into adding a (sub) woofer system, so it's a 'pay me now or pay me later' scenario in total box bulk and system budget with the benefit of increased potential performance and shape, placement flexibility.
GM
If the acoustical output from the drivers does not sum properly on-axis, you will always notice it, no matter what crossover frequencies are chosen. There is no way to hide the effects of bad crossover design. On the other hand, a good crossover is hardly noticed at all, but it may have to include compensation like steps, notches, etc...
True, in an ideal world, but we're talking about some usefully effective points to place them, reducing the audibility of the transition & which also in some cases may permit simpler topologies.
One would hope that all XOs are well designed for the task in hand. Which sadly, is rarely the case. Active / digital is the way forward IMHO.
One would hope that all XOs are well designed for the task in hand. Which sadly, is rarely the case. Active / digital is the way forward IMHO.
I think the AN10" CF ought to work with a woofer driven by Reckhorn A-410 as it has both high and low pass adjustable up to 300hz along with sub Eq as well as lower midbass Eq
Fore the surround channels the AN10 may work ok on its own
It doesnt need a very big box
A notch from 2khz and up may work wonders
Its not perfect and if its possibel at all to make the connections with a surround receiver I do not know, but doing a passive design takes some experience, a crossover in the midbass is not as simple as it sounds and can be very tricky
Fore the surround channels the AN10 may work ok on its own
It doesnt need a very big box
A notch from 2khz and up may work wonders
Its not perfect and if its possibel at all to make the connections with a surround receiver I do not know, but doing a passive design takes some experience, a crossover in the midbass is not as simple as it sounds and can be very tricky
Just got in recently.
Hi rjbond3rd,
Interesting idea to use these two sims to compare. Thank you.
I'm not sure if I'm using your sim properly though to I compare GM's and Scottmoose's sims. Anybody slap my wrists here...
What I observe is that Scottmooses's curve seems to have little or no rise around 100Hz and a subsequent linear roll-off slope similar to something between the green and yellow lines in your sim. This would correspond to a Qtc somewhere in the middle between 0.66 and 0.92.
What I observe with GM's curve is a more rounded roll-off slope but I'm not sure if it has a slight rise around 80 or 90Hz. Still, I'd call it a bit closer to the yellow line corresponding to a Qtc of 0.66 than that of Scottmoose's.
Now, being that these are near aperiodic TL's something tells me I'd be wrong to simply assign Scottmoose's curve, say a... (0.66 + 0.92) / 2 = 0.79. I'm not attempting to do that.
But would it be permissible to hazard that GM's curve would represent a tighter sound, while Scottmoose's would be easier to mate with a sub?
Hm... Low xover at low volumes, high at high. Up to now I thought low volumes could use more weight and so it would be useful to turn the sub's crossover up to cover more of the low band. But obviously you know something I'm not aware of yet.
I'd expect it'd be important to keep this sub strategy in mind as any design particulars are worked out. But I have to admit after thinking abut it I can't grasp the basis for this - so I guess it represents a critical hole in my understanding. Can you expand on this?
-jerry
rjbond3rd said:
Hi rjbond3rd,
Interesting idea to use these two sims to compare. Thank you.
I'm not sure if I'm using your sim properly though to I compare GM's and Scottmoose's sims. Anybody slap my wrists here...
What I observe is that Scottmooses's curve seems to have little or no rise around 100Hz and a subsequent linear roll-off slope similar to something between the green and yellow lines in your sim. This would correspond to a Qtc somewhere in the middle between 0.66 and 0.92.
What I observe with GM's curve is a more rounded roll-off slope but I'm not sure if it has a slight rise around 80 or 90Hz. Still, I'd call it a bit closer to the yellow line corresponding to a Qtc of 0.66 than that of Scottmoose's.
Now, being that these are near aperiodic TL's something tells me I'd be wrong to simply assign Scottmoose's curve, say a... (0.66 + 0.92) / 2 = 0.79. I'm not attempting to do that.
But would it be permissible to hazard that GM's curve would represent a tighter sound, while Scottmoose's would be easier to mate with a sub?
Hm... Low xover at low volumes, high at high. Up to now I thought low volumes could use more weight and so it would be useful to turn the sub's crossover up to cover more of the low band. But obviously you know something I'm not aware of yet.
I'd expect it'd be important to keep this sub strategy in mind as any design particulars are worked out. But I have to admit after thinking abut it I can't grasp the basis for this - so I guess it represents a critical hole in my understanding. Can you expand on this?
-jerry
GM said:
Hello GM,
Thank you for responding. There is a depth of information in your post that will have me working on it a while.
I wanted to ask you some things but your post stimulated in me a need to come a bit clean on TL's.
I believe in honoring the generosity and effort of everyone who posts to me, but I must confess that despite my best efforts I've not been able to catch up to this recent subject as much as I'd like. Of the reading I've done on enclosures since this thread began much of it was on BR, then sealed, the much simpler "leaky sealed" aperiodic and a bit on horns/BVR's mixed in the middle. Of my original attempts to grasp TL's what impressed me most was the advice warning that it would require some competence to pull off successfully so I thought it wise to invest the reading time I had available on simpler boxes. Thinking back, I'm a little concerned when I mentioned the Metronome earlier that I may have given a different sense of my abilities. Then on top of this I wasted precious time the past few days to finish off my reading of the simpler aperiodic - the use of a controlled leak in an otherwise plain sealed box - to make sure I had no other choice but to dive into the TL world.
So, suddenly I find myself unprepared as I find myself in a relative sweep of suggested TL-based aperiodic solutions. Yet of course, with my size and performance concerns, and that curious AN driver... what else could it have been? But I will make it my priority to get a handle on this over the next few days.
Questions:
I sense what you've suggested to me is inspired. But if I really did have faith in my sealed servo Rythmik sub and wanted to make the fullest use of it, would there be a variation of this enclosure that sacrifices a bit off the LF? ie. If not needed (and I again apologize if I seem an ingrate for your achievement but) could that extra LF capacity go towards more damping etc?
I am ignorant perhaps, whether this would have the effect of also losing, rather than helping the
And secondly, may I ask what HxWxD these measurements would roughly correspond to? (L = 85.468" ; CSA = 335.121"^2 ; zdriver = 32.197")
Thank you,
-jerry
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.