Yet another query re TMM vs.MTM (using DEQX)

Having read many of the threads discussing the relative benefits of each, I must admit to some persistent confusion and would much appreciate some clarification as to potential issues I face with my upcoming build. :confused:

First I must confess that I am unequipped to do any serious cabinet building :eek: and am planning a drop-in approach using a pair of Usher 8948a 7's per side that go into a cubic foot enclosure measuring 22" tall. Were I using a conventional tweeter, I would probably just use an MTM approach and call it good. But I tend to like planar drivers and if I position the port in front, there is precious little real estate available. Even if I go out the rear, things are seriously cramped, particularly if I go with an AMT type device which is my current leaning--maybe Beyma or the Aurum Cantus aerostrictions. (This last preference BTW just is one of those bucket list items that dates back to my first hearing of the ESS Heils :D and vowing one day to build a speaker based on those and have owned a few magneto-planar based systems, most recently the big B&G 75's).

So I have decided just to build outboard tweeter modules and make those more or less ear level which drops the woofers to something like 2' and 1.5' above the floor. Now after having read up a bit I am stressing :headbash: about cancellation and floor bounce, whether it should be a 2.0 or 2.5 configuration, etc. Since I plan on using a DEQX for XO and bi or tri-amping the sysem, where I do have a potential advantage is very little overlap between the mids and tweeters and perhaps? the ability to lower the XO point to between 1.5 and 2 kHz. My back of the napkin calcs show a delay of about 0.1 ms worst case for direct arrival times but I know this is making some awfully simplistic assumptions. So looking for some enlightenment that will maybe improve the odds of getting a really good implementation vs using a pair of these::dice:

Thanks in advance for any help to this noob.
 

terry j

Member
2006-08-02 12:46 pm
just a tip, when designing your speaker never allow yourself to think 'I'll use the ability of the deqx' to overcome any sort of basic established speaker building principals.

It cannot make a silk purse out of a sows ear matter how much it might make it better.

So now I'll let the technical points be addressed by those that can help you.
 
just a tip, when designing your speaker never allow yourself to think 'I'll use the ability of the deqx' to overcome any sort of basic established speaker building principals.

It cannot make a silk purse out of a sows ear matter how much it might make it better.

So now I'll let the technical points be addressed by those that can help you.

So terry, appreciative of the advice and hope--really really hope that dropping a couple grand on some respectable drivers/cabinetry is not the proverbial sows ear: I been having to save coin for far too long. But I am also concerned that I am not engaged in wishful thinking here by supposing that at least some of the problems inherent to a 2.0 TMM design might be reduced by using steeper slopes and a lower XO to minimize the region of overlap and the phase errors in just the same way that a higher order passive XO might have an advantage over say a first order XO that has significant overlap among drivers for say a couple of octaves. It other words I hope there is nothing about the DEQX beyond the ability of any high order active XO I am counting on, and if so would want to be called out on it as I have been saving my coin for some time and want to get this right! The DEQX just happens to be one of the few toys from a much more prosperous time I still own, and have become quite attached to it as it is the one purchase that got me out of the vicious cycle of upgrade-itis and paying for the college educations of other peoples kids--namely my favorite audio dealer's, and showed me the potential once and forever of DIY.

As an example: Say I run the filters at 100 dB/octave with a XO=1500, if my math is OK, the f10 points will be like 1450 and 1575Hz. Now for grins compare this to a 4'th order network which many might consider steep, the 10dB down points will be 1225 and 1840--so in one case I have overlap of 100 Hz and in the other, over 600. Granted there are some philosophical tradeoffs (or engineering compromises) and potential irritating discontinuities associated with what might be called a lack of blending, just as there will always be a dyed in the wool camp that insists first order XO's are and have always been the most natural sounding. It all depends and what works for me is to just get it (the transition) over and done with. I listened to the Ushers for quite a while with and w/o DEQX eq before committing, and can only hope that the transition from moving coil to planar bugaboo doesn't bite me in the ****. If so I'll bail and retool my thinking. That's a big unknown as I have not had the opportunity to listen to the newer AMT's.

I guess my original question (and dumptruck I would love to hear your rationale), is why the 2.5 might be preferred given the tiny slice of shared bandwidth we are looking at. I truly want to understand as I can easily add an inductor to the 0.5 woofer which seems to have its set of +/-'s and BSC can be dialed in with the DEQX using one of 10 parametric EQ's. Is it something else besides the lower woofer-tweeter collision? Just trying to get a handle on all of this before creating sawdust.... The proof is always in the listening and if someone can suggest a XO point, I'll invest in a couple of coils to do some serious A-B listening.
 
Consider a modular (MT) W design, perhaps with the goal of designing one high'ish efficiency woofer with two different MT tops as time and money allows. AMT planars are typically expensive, and mated to higher quality ~95db/watt midbass like the Lambda TD10M or Beyma 8P300. There are several good DIY Usher 8948a + dome tweeter MT designs that can go above a woofer.

There are several well written, and well reviewed papers on M-T-M spacing with equations plus measurements for both the M-T and M-M spacings required to avoid lobing. These papers prove that an MTM using 7" midbass speakers with any Xover over ~700Hz will have noticable lobing.

Biro Technology

Vertically Symmetric Two-Way Loudspeaker Arrays Reconsidered
---------

If you are set on using just the four Ushers, then a 2.5 way (TM)M would be a good choice, where the low frequency M can have the Xover frequency and height adjusted for baffle step and floor bounce effects.

Don't forget rounded edges for diffraction control.
 

Attachments

  • max.jpg
    max.jpg
    173 KB · Views: 435

terry j

Member
2006-08-02 12:46 pm
maybe you took me the wrong way?

Don't fret, I use two deqxs in my system (four way plus subs), ALL I was 'latching' on to was Since I plan on using a DEQX for XO and bi or tri-amping the sysem, where I do have a potential advantage is very little overlap between the mids and tweeters and perhaps?

Sure, it was a question, yet also a question set against what seemed to be a recognition of possible 'poor' driver layout if you will.

My simple point being, no matter how wonderful a unit like the deqx may or may not be (and obv I think it is a good one) it is best used to maximise a good design rather than 'make amends' for a less optimum one.

I most certainly was not commenting on the quality or otherwise of your driver selections.

Was more a philosophical point if you will.

Sorry if I offended.
 
Consider a modular (MT) W design, perhaps with the goal of designing one high'ish efficiency woofer with two different MT tops as time and money allows. AMT planars are typically expensive, and mated to higher quality ~95db/watt midbass like the Lambda TD10M or Beyma 8P300. There are several good DIY Usher 8948a + dome tweeter MT designs that can go above a woofer.

There are several well written, and well reviewed papers on M-T-M spacing with equations plus measurements for both the M-T and M-M spacings required to avoid lobing. These papers prove that an MTM using 7" midbass speakers with any Xover over ~700Hz will have noticable lobing.

Biro Technology

Vertically Symmetric Two-Way Loudspeaker Arrays Reconsidered
---------

If you are set on using just the four Ushers, then a 2.5 way (TM)M would be a good choice, where the low frequency M can have the Xover frequency and height adjusted for baffle step and floor bounce effects.

Don't forget rounded edges for diffraction control.

Linesource--many thanks and will get back to you when I have a bit more time. Actually I started right where you suggest with bass cabs and a Mid/highs module on top. But then on the advice of nearly everyone while lurking decided that a 3 way might be a bit more of a bite than I am prepared to chew: in other words get a 2 way right (or 2.5) then tackle the 3 way. Nice thing about the ushers that the vented alignment and sealed (critically damped) box volumes are the same--so I just plug the vent when I go 3 way and am good to go. Part of using two/side was to push efficiency and max spl so as to kind of keep up with the AMT's should I go there.Any way morre later,appreciate the input.
 
maybe you took me the wrong way?

Don't fret, I use two deqxs in my system (four way plus subs), ALL I was 'latching' on to was Since I plan on using a DEQX for XO and bi or tri-amping the sysem, where I do have a potential advantage is very little overlap between the mids and tweeters and perhaps?

Sure, it was a question, yet also a question set against what seemed to be a recognition of possible 'poor' driver layout if you will.

My simple point being, no matter how wonderful a unit like the deqx may or may not be (and obv I think it is a good one) it is best used to maximise a good design rather than 'make amends' for a less optimum one.

I most certainly was not commenting on the quality or otherwise of your driver selections.

Was more a philosophical point if you will.

Sorry if I offended.
None taken whatsoever. What I was trying to get at is that I suspect many of the reasons that a TMM 2.0 is traditionally rejected in favor of a MTM or 2.5 implementation has to do with are 1) The overwhelming popularity of a D'Apppolito config whose engineering strengths may have more to do with minimizing the shortcomings of relatively shallow passive XO's than an absolute superiority given 2 midwoofers and a tweeter with no real constraints on XO topology. (I don't know the answer to this obviously but being the kind of guy that believes more often than not "conventional wisdom isn't," questioning whether it makes sense to at least reexamine the alternatives) and 2) that again most MTM configurations employ tweeters with relatively high XO's so that even small path length differences between the two woofers can result in nulls where the tweeter ain't present to fill in the gap. 3) Many audiophiles seem to be in love with visual symmetry even when they know it may be for the worse. ;)

So in short, the question I was trying to phrase and did a lousy job of is given some major technical differences such as a tweeter comfortable at 1500 Hz and ultra steep XO's are 2.5 or MTM systems necessarily superior? Which is sort of the inverse position of starting with a flawed design and counting on the DEQX to bail me out. So no offense taken--just trying to clarify the question, which after reading the link provided to Biro Audio seems more open than ever.
Cheers,
John S
 

terry j

Member
2006-08-02 12:46 pm
I'm glad then that I was not taken the wrong way.

I can't help you too much with any lobing technical questions, tho I wonder if it is actually true re the favored position of MTM. Maybe it is only 'lately' but I have the impression I am seeing quite a few TMM around. Perhaps the pendulum has swung a bit on that one?? (my gut feel is that ten-twenty years ago MTM were definitely the majority??)

Personally, I would question the 'superiority' of MTM, I really don't 'like' them at all. Not that that is saying TMM is superior, don't have much experience with them. I find TM (W I guess) to have a much larger soundstage.But then again my room has very high ceilings, and maybe one of the advantages of MTM is to reduce ceiling reflections (dunno if that is even true mind)

I think you could be right about the possible reasons you gave about MTM being so popular, 'they are everywhere' (becomes 'so that is what we all should do'..ie fashion) and visual symmetry.

I once read on the net a persons description of 'why I don't like D'appolito' or somesuch (am mentioning this because for the life of me I cannot find it again...this might jog someones memory) and as I read it I found myself nodding and saying 'yeah! that's right!' with every point.

So perhaps the upshot of all of that is, like everything, there are differing viewpoints and hence yes, you wondering if MTM is inherently superior is a valid one.

Any thoughts on adding subs eventually?
 
Terry,

Hi again. Subs? Definitely and would love to get your take on getting the most out of the DEQX as my former attempt not so good with music-used a pair of ridiculously long throw 18's in an IB enclosure with the back half directed into the crawl space. Enough output at 15 Hz to break out the windows but also good at exciting the entire flooring of the house-a brand new house at that. But a major whoopee for those into seat of the pants tactility, but I digress.

So will be looking for subs in the future once I get the mains dialed in. As I mentioned, at that time I will likely seal the enclosure which gives a critically damped alignment with an f3 of around 85 Hz. Likely run a stereo pair of subs and look for 10's or 12's that can be stuffed into a small enclosure. Tc Sound has something good IIRC.

And you may be absolutely right on regarding more recent trends. I'm just getting back into serious audio after a slumbering interest of a few years. What I did was just google MTM vs 2.5 or variations thereof and read all I could find--which really wasn't all that much... and which prompted this thread in the first place.

So at this point I'm feeling pretty good about the choice of running a TMM config. Will start with 2.0 listen, measure, tweak and if I don't like then get a couple of coils.

And after some soul searching, am gonna go with AMT's--just a matter of choice between Aurum Cantus (which I can find precious little about in DIY circles except they measure well) and the much more prevalent Beyma's which cost a bit more but aint Chinese. Of course I would love to get my hands on some of the ADAM's but don't see that as an option just as there are some damn fine ribbons that can't be had. :mad:
Cheers
 

terry j

Member
2006-08-02 12:46 pm
more than happy to help if I can with the subs. If I don't respond then PM me...hopefully I have 'email notice' as I might miss a post...find I am less and less checking forums.

guess that is the trouble with an IB...once they're in and you are less than impressed it is damned hard to do anything about it! At least placement wise.

If in your shoes I think I'd be pretty happy with whichever layout you went with. Tho personally not a real fan of MTM I find it hard to imagine that the diff between the two is so great that regret (if you magically went to the other) would be a word you'd use.

Still, if you are like me and curiosity bites real hard, with a bit of forethought building a box with interchangeable baffles (which can be permanently fixed later) is not a hard thing to do.

Then you can let us all know what you found for yourself! Finding out for yourself (rather than relying on posts like mine...ie someone else's opinion) is gold.

good luck
 
more than happy to help if I can with the subs. If I don't respond then PM me...hopefully I have 'email notice' as I might miss a post...find I am less and less checking forums.

guess that is the trouble with an IB...once they're in and you are less than impressed it is damned hard to do anything about it! At least placement wise.


good luck
Ha. Had to chuckle on that. Not only hard to move but hard to fix the nearly 2 x 2 hole you just put into the floor of you great room.:D:eek::D
 

Attachments

  • P1010028.JPG
    P1010028.JPG
    217.1 KB · Views: 204
  • P1010031.JPG
    P1010031.JPG
    206.3 KB · Views: 201
  • P1010034.JPG
    P1010034.JPG
    188.9 KB · Views: 193
more than happy to help if I can with the subs. If I don't respond then PM me...hopefully I have 'email notice' as I might miss a post...find I am less and less checking forums.

guess that is the trouble with an IB...once they're in and you are less than impressed it is damned hard to do anything about it! At least placement wise.


good luck
Ha. Had to chuckle on that. Not only hard to move but hard to fix the nearly 2 x 2 hole you just put into the floor of you great room.:D:eek::D
 
Consider a modular (MT) W design, perhaps with the goal of designing one high'ish efficiency woofer with two different MT tops as time and money allows. AMT planars are typically expensive, and mated to higher quality ~95db/watt midbass like the Lambda TD10M or Beyma 8P300. There are several good DIY Usher 8948a + dome tweeter MT designs that can go above a woofer.

There are several well written, and well reviewed papers on M-T-M spacing with equations plus measurements for both the M-T and M-M spacings required to avoid lobing. These papers prove that an MTM using 7" midbass speakers with any Xover over ~700Hz will have noticable lobing.

Biro Technology

Vertically Symmetric Two-Way Loudspeaker Arrays Reconsidered
---------

If you are set on usin effects.g just the four Ushers, then a 2.5 way (TM)M would be a good choice, where the low frequency M can have the Xover frequency and height adjusted for baffle step and floor bounce

Don't forget rounded edges for diffraction control.
LS,
I just wanted to thank you for taking the time to respond to my thread re MTM vs TMM. The article you pointed me to was very helpful in making a decision not to do an MTM. In fact I felt major league stupid for not seeing the problems inherent to a system that advances and retards arrival time simultaneously with movements along the y axis and provides no fill from the tweeter.

I also stumbled across a comment of yours on some Beyma thead which contained more wisdom in a paragraph than most books on the subject of audio re the pros of dipole, sealed box, horns etc which ought to be engraved somewhere. Likely could have saved a 100 bucks on a couple of almost useless books. And in retrospect I wish I had trusted my instincts a few months back when i wanted to get some pro audio midbass units like a PHP or Beyma--that sense of startling immediacy is what I crave. But inadvertently I got that right as the AMT will be outboard, the woofers underneath it all, and can swap out mids in a NY minute and have all the XOs available at the touch of the keyboard.

But really just wanted to say thanks for the outreach to the noobs.
JohnS
 

terry j

Member
2006-08-02 12:46 pm
well there ya go john, 'we are moving closer' albeit unplanned.

I use pro drivers everywhere 'cept tweet, PHL 6.5's, PHL 10's and PHL 18's (the subs are not pro...pro's don't worry about anything under thirty I guess) so use maelstrom 18's there.

Startling slam and jump factor, a real kick in the chest when it is in the music, ability to play loud without strain, minimum sensitivity about 96 db (none of this audiophile 90 db stuff)...all active (as yours will be)

The one common comment from all who hear it is 'well, THAT ain't your normal stereo'. Whether it is preferred or not is of course a personal thing, but all seem to recognise it is not a 'sound' you have heard before.
 
the AMT will be outboard, the woofers underneath it all, and can swap out mids in a NY minute

The MAX 3-way speaker picture that I posted earlier (TPL150H + 8"midbass above separate woofer) was designed by an engineer who understood how important it was to BUTT the TPL150H to the midbass with the smallest possible gap. No separate boxes with their >1" wood separation that risks lobing. Let's assume a TM 1600Hz Xover. 13550 / 1600Hz = 8.4" wavelength. Ideally you want your TM speakers separated by 1/4 this wavelength, but 1/2 is often accepted as the worst acceptable gap. Approaching even a 1/2 Lambda T-M separation is only possible when the M is butted next to the T, no extra box wood.

A free, easy to use program call xdir provides basic graphical lpolar data for M-T spacing vs Xover frequency. Life gets harder with the more you learn... Pizza quickly becomes a temptation of the Devil vs. a snack with College friends.

Tolvan Data
 
The MAX 3-way speaker picture that I posted earlier (TPL150H + 8"midbass above separate woofer) was designed by an engineer who understood how important it was to BUTT the TPL150H to the midbass with the smallest possible gap. No separate boxes with their >1" wood separation that risks lobing. Let's assume a TM 1600Hz Xover. 13550 / 1600Hz = 8.4" wavelength. Ideally you want your TM speakers separated by 1/4 this wavelength, but 1/2 is often accepted as the worst acceptable gap. Approaching even a 1/2 Lambda T-M separation is only possible when the M is butted next to the T, no extra box wood.

A free, easy to use program call xdir provides basic graphical lpolar data for M-T spacing vs Xover frequency. Life gets harder with the more you learn... Pizza quickly becomes a temptation of the Devil vs. a snack with College friends.

Tolvan Data
Linesource,
Very informative link and I see exactly what you mean re avoiding too much interdriver distance--makes perfect sense to me now why faceplates get carved up. Maybe you can help me with one question re XDir: The distance one specifies I assume is center to center from tweet to woofer. But nowhere does it ask what the distance between woofers is and yet displays TWW config data which in a few cases shows some strong lobing at +/- 60 degrees or so which I figure would result in nasty ceiling (and floor) reflections. I am a little worried at this point as I doubt I can get the distance to less than 4.6" to the lower edge of the Heil and another 3 to the center. But am I correct in thinking that the XDir assumes 2pi radiation and so maybe the result is not as bad, ie the 4.6" distance obtains, not the 7.6".
With heroic lengths I could contrive to knock almost an inch off the distance by doing a slightly stepped baffle where I bury the bottom part of the Heil faceplate underneath the flange of the woofer. But besides a PITA, have no idea what adverse effects on performance may have in addition to obvious detrimental impact on aesthetics. BTW leaning toward the Aurum cantus aerostriction 30130 at this point as it looks like 1100 to 1200Hz might be doable. But always that gocha-can't be used dipole--at least according to the PE guy. :mad:
 
well there ya go john, 'we are moving closer' albeit unplanned.

I use pro drivers everywhere 'cept tweet, PHL 6.5's, PHL 10's and PHL 18's (the subs are not pro...pro's don't worry about anything under thirty I guess) so use maelstrom 18's there.

Startling slam and jump factor, a real kick in the chest when it is in the music, ability to play loud without strain, minimum sensitivity about 96 db (none of this audiophile 90 db stuff)...all active (as yours will be)

The one common comment from all who hear it is 'well, THAT ain't your normal stereo'. Whether it is preferred or not is of course a personal thing, but all seem to recognise it is not a 'sound' you have heard before.
Sounds like a lot of fun. I think running pro drivers does makes a big diff, better yet if you like horns. Most of the DEQXed systems I have heard (where the measurements and implementation were done right) also have some of that snap and alacrity--just by getting the phases of the divers all in a ro--tho I have no proof and indeed have read umpteen references to the inaudibility to say nothing of the futility of a phase coherent full way speaker(s).
 
Last edited:
I don't think that the DEQX has that configuration as one of its capabilities. Usually a 2.5 way configuration is a passive crossover artifact that uses an inductor to achieve a 6 dB per octave slope on one of the two mid-range drivers that would otherwise cover the same frequency band. Essentially you can adapt this arrangement to achieve 6 dB baffle step compensation.

With its ability to calibrate amplitude, phase and time you have considerable capability built-in to the DEQX unit. This includes equalization and high order linear phase crossover slope capabilities, so the DEQX takes compensation to a new level. Typical slopes are 48 dB/octave or greater. I do not believe that the DEQX has the capability to separately calibrate over two overlapped frequency bands. I would just use a two-way DEQX configuration. You can set the amplitude EQ limits range to easily adjust the amplitude response so that you attain flat amplitude response so that a 2.5 way configuration is not needed. Bottom line is that you have race horse capable DSP so don't limit it to solving donkey era problems.
 
Last edited: