Upscale MTM?

[TLS Guy]

You raise a very good point. An MTM is not an MTM is not an MTM. Very few, if any, current MTM's conform to the design parameters that Joe specified in his 1983 paper. My takes on the MTM configuration certainly don't meet the 1983 specs. What I find interesting is that Joe himself abandoned that design. For example, his Thor TL MTM used a 4th order acoustic slope. So I'm not sure where that leaves us in the current discussion. As promised, I did compare a couple of my monitor designs with an MTM that used the same drivers as one of the MT's, using a volume-compensated A-B switching preamp. The two speakers with the same drivers sounded almost identical. The only real difference was a slightly heavier bass on the MTM due to a cabinet that wasn't twice as large as the MT--the cabinet tuning was a little higher and a bit humped up. The other MT, which used more expensive drivers, including a ribbon tweeter and a larger woofer, sounded more spacious and cleaner over all. So I remain unconvinced that simply mounting drivers in an MTM configuration and optimizing a crossover for flat response will yield any particular sonic benefits other than increased sensitivity and lower harmonic distortion in the bass.

I don't want to suggest the MTM is vastly superior to MT. The results are subtle.

The Thor crossover is very unusual. It is not fourth order electrically. It is electrically third order with a second order zero. The combined acoustic and electrical crossover is fourth order. The crossover does provide the required tilt.

Now Gilbert Briggs pointed out back in the 50s that as you extend bass response the psycho acoustic effect is increased spaciousness and increased sense of space. So it does not surprise me that your speaker with the larger driver sounded more spacious.

In my speakers, I have an MTM extended with two 10" drivers in each speaker, handling the BSC and extending the response very close to flat at 20 Hz. These speakers produce by far the most realistic sound stage and spaciousness I have ever encountered. I think the MTM gives it an edge, but I can't be certain. One thing I do know is that having the two drivers gives it awesome power and reserve. I was watching Edward Elgar's the Dream of Gerontious under Barenboim on the BPO site a couple of nights ago. The shear power and sweep from the softest sounds to the huge orchestral, organ and choral crescendos was astonishing. Never was the slightest strain or dynamic restriction felt. You were just transported on the huge wash of sound. So this arrangement allowing for the increased power and reduction of thermal compression in the vital midrange, is justification enough for me.

 

[TheWarrior]

The Thor crossover is very unusual. It is not fourth order electrically. It is electrically third order with a second order zero. The combined acoustic and electrical crossover is fourth order. The crossover does provide the required tilt.

Is this a result of flanking each capacitor with a resistor that terminates on the negative?

You were just transported on the huge wash of sound. So this arrangement allowing for the increased power and reduction of thermal compression in the vital midrange, is justification enough for me.

Do you use any additional subs in your room? It doesn't look too terribly large as to need it, I am just curious. In my stereo configuration, I greatly enjoy having the second sub behind me to provide an even more immersive experience. But I do not have full range towers.

 

[TLS Guy]

Is this a result of flanking each capacitor with a resistor that terminates on the negative?





Do you use any additional subs in your room? It doesn't look too terribly large as to need it, I am just curious. In my stereo configuration, I greatly enjoy having the second sub behind me to provide an even more immersive experience. But I do not have full range towers.

There are no resistors flanking capacitors. The circuit is very economical. The high pass is classic third order 18 db per octave below crossover at 2.5 KHz. It transitions to second order where the tweeters roll off becomes becomes steeper to give a combined acoustic and electrical roll off fourth order. There is also a network padding the tweeter to match the woofer output and slightly equalize tweeter response. The tilt occurs because this is an odd order electrical crossover.

The low pass is basically second order electrically. The series inductor is higher than usual calculation and attenuates response starting at 400 Hz to provide BSC compensation. Its higher value also compensates for the woofers rise in impedance and saves the expense of components required for a Zobel. The parallel inductor not only provides faster roll off at crossover but there is also a tuned circuit realized by the resistor and capacitor in series with L2, this forms the notch filter suppressing the woofers 4.4 KHz break up mode peak by 36 db. The combined acoustic and electrical response is fourth order.

This is a very ingenious design, with components multitasking!

The circuit looks very simple, but it carries out a multiplicity of tasks with a low component count.

As far as subs, no I don't have any in the room. The large lines have more than enough output, and since the lower frequency crossovers are all active, the LFE channel is mixed in.

This is a low Q design, and the pipes are highly efficient acoustic couplers to the room. They generate enough low frequency energy to be structurally damaging to the house if not handled with sense.

The rear backs are also very capable in the LF department, with large output well down into the twenties. So when called upon they can pack a punch in sub range.

There are no discrete subs and they are definitely not required.

 

[TheWarrior]

This is a very ingenious design, with components multitasking!

The circuit looks very simple, but it carries out a multiplicity of tasks with a low component count.

Exactly what I needed, now to study further so that I get my verbage right and can truly appreciate what this design accomplishes and offers. Thank you again, sir!

 

[Dennis Murphy]

I don't want to suggest the MTM is vastly superior to MT. The results are subtle.

The Thor crossover is very unusual. It is not fourth order electrically. It is electrically third order with a second order zero. The combined acoustic and electrical crossover is fourth order. The crossover does provide the required tilt.

.

Actually, that was a very conventional crossover, and turned out to be identical to the one I came up with for the same complement of drivers (and pretty much used in the Salk Excel speakers). You always end up with a trap circuit to ground and either one or two series inductors with the Seas Excel woofers. When I reference a crossover slope, it's always acoustic, not electrical. The electrical is pretty much irrelevant--it's what comes out of the drivers that counts. If that happened to provide the requisite lobing tilt, it was accidental for or subordinate to my design, because I was just going for reasonably flat response over the listening window. Maybe you only get that with correct tilt, in which case the Salk HT2 conforms to Joe's original goals. But apparently you don't need odd-order acoustic slopes to accomplish that.

The Thor never did very well because Joe was less skilled in transmission line tuning than he was in crossover design.

 

[TLS Guy]

Actually, that was a very conventional crossover, and turned out to be identical to the one I came up with for the same complement of drivers (and pretty much used in the Salk Excel speakers). You always end up with a trap circuit to ground and either one or two series inductors with the Seas Excel woofers. When I reference a crossover slope, it's always acoustic, not electrical. The electrical is pretty much irrelevant--it's what comes out of the drivers that counts. If that happened to provide the requisite lobing tilt, it was accidental for or subordinate to my design, because I was just going for reasonably flat response over the listening window. Maybe you only get that with correct tilt, in which case the Salk HT2 conforms to Joe's original goals. But apparently you don't need odd-order acoustic slopes to accomplish that.

The Thor never did very well because Joe was less skilled in transmission line tuning than he was in crossover design.

I'm going to come to Joe's defense about that line.

It is a traditional well damped reverse tapered line.

Joe admits that because of WAF the line is undersized in terms of volume but not length or tuning. For those drivers, LP and FP are just where they should be. VP is undersized based on the VAS of the two drivers.

I don't think you could get lower output from that pair, but you likely could get you more output. But where would that get you? The frequency response is very smooth with and an F3 of 44 Hz. Because the line was, in my view, correctly damped to leave just one peak of impedance and not over damped, as Joe shows in his graphs, roll off is second order. This results in a measured 12 db point of 25 Hz.

From what I can tell this side of the water people like their LF somewhat resonant.

Lines like I have been using for years are not, and sound different. However I maintain they sound correct. In a live concert the bass touches lightly with good detail. I have come to know one thing about this design approach is that listeners who live with them never want to go back to anything else.

The other issue about the Thor acceptance is that if you are going to use a sub, it absolutely needs to be a TL designed using the same principles. You know how rare they are.

In my case I took a leaf out of John Wrights book and used two lines in one enclosure. To do this well does require an active design. This results in seamless driver support over at least two and a half octaves.

 

[TheWarrior]

Since Dennis and Dr. Carter are both analyzing this, I am going to take the opportunity to learn!

1) I see this as 3rd Order Electrical. What makes this 4th order acoustic? The sum of parallel and series circuits?

2) Is the purpose of L3 to balance the voltage between high pass and low pass resulting from the tweeters lower impedance?

3) What is the purpose of 2 resistors going to ground? Is that the 'trap' for left over current to be converted to heat?

4) I see the notch filter at L2 and the purpose of the big inductor at L1. But referencing TLS' comment about 'parts multitasking', what else is multi-tasking? Does this tie into my (hopefully accurate) question #2?

I hope these questions make sense. Thank you for any assistance in this!

 

[Dennis Murphy]

I'm going to come to Joe's defense about that line.

It is a traditional well damped reverse tapered line.

Joe admits that because of WAF the line is undersized in terms of volume but not length or tuning. For those drivers, LP and FP are just where they should be. VP is undersized based on the VAS of the two drivers.

I don't think you could get lower output from that pair, but you likely could get you more output. But where would that get you? The frequency response is very smooth with and an F3 of 44 Hz. Because the line was, in my view, correctly damped to leave just one peak of impedance and not over damped, as Joe shows in his graphs, roll off is second order. This results in a measured 12 db point of 25 Hz.

From what I can tell this side of the water people like their LF somewhat resonant.

Lines like I have been using for years are not, and sound different. However I maintain they sound correct. In a live concert the bass touches lightly with good detail. I have come to know one thing about this design approach is that listeners who live with them never want to go back to anything else.

The other issue about the Thor acceptance is that if you are going to use a sub, it absolutely needs to be a TL designed using the same principles. You know how rare they are.

In my case I took a leaf out of John Wrights book and used two lines in one enclosure. To do this well does require an active design. This results in seamless driver support over at least two and a half octaves.

Looks like we'll have to agree to disagree on this one. I can't tell you how many requests I've received from Thor owners to have me rework the crossover. The complaint is always the same--the sound is thin and fatiguing. I always tell them the same thing--the crossover is fine. It's the TL that's lending the thin quality, along with the sonic character of the Excel drivers in the midbass. All the TL's I've worked with have been designed using Martin King's software, and I'm pretty sure they're optimized correctly. None of them sounds thin. And the bass I hear coming from my orchestra isn't thin either. It's a little out of tune, but not thin. Actually, I'm always surprised at how resonant it sounds compared with my speakers. But I think we can agree that transmission lines are the way to go if the woofer(s) and Significant Others lend themselves to that approach.

 

[TheWarrior]

\


Back to question #1, is it the steepness of the impedance spike corresponding to the roll off at the crossover frequency?

 

[TheWarrior]

Looks like we'll have to agree to disagree on this one. I can't tell you how many requests I've received from Thor owners to have me rework the crossover. The complaint is always the same--the sound is thin and fatiguing. I always tell them the same thing--the crossover is fine. It's the TL that's lending the thin quality, along with the sonic character of the Excel drivers in the midbass. All the TL's I've worked with have been designed using Martin King's software, and I'm pretty sure they're optimized correctly. None of them sounds thin. And the bass I hear coming from my orchestra isn't thin either. It's a little out of tune, but not thin. Actually, I'm always surprised at how resonant it sounds compared with my speakers. But I think we can agree that transmission lines are the way to go if the woofer(s) and Significant Others lend themselves to that approach.

As I will be making 1.25" thick cabinets, would it behoove me to modify the cabinet dimensions? I'll be using these with subwoofers, but would like to think for most music they will be fine by themselves.

 

[TLS Guy]

Since Dennis and Dr. Carter are both analyzing this, I am going to take the opportunity to learn!

1) I see this as 3rd Order Electrical. What makes this 4th order acoustic? The sum of parallel and series circuits?

2) Is the purpose of L3 to balance the voltage between high pass and low pass resulting from the tweeters lower impedance?

3) What is the purpose of 2 resistors going to ground? Is that the 'trap' for left over current to be converted to heat?

4) I see the notch filter at L2 and the purpose of the big inductor at L1. But referencing TLS' comment about 'parts multitasking', what else is multi-tasking? Does this tie into my (hopefully accurate) question #2?

I hope these questions make sense. Thank you for any assistance in this!

The overall slope is the sum of the electrical and acoustic slopes. There is 6db roll off per order of slope. So a third order electrical slope plus a first order acoustic response of a driver, makes for a total fourth order response, which is 24 db per octave.

Now L1 and C1 alone would be standard second order low pass topology, the addition of L1 creates a resonant acceptor circuit with C1 to deal with the break up peak at 4.4 K Hz. R1 gets the Q of the filter in the correct range.

C2 L3 and C3 is standard third order high pass topology.

R2 and R3 is a standard L-pad to make the tweeter output match that of the woofers.

The addition of C4 shapes the tweeter response slightly.

Now L1 starts to cut off the woofers at 400 Hz to provide BSC. Remember a passive circuit can never boost only cut. So with a passive circuit to boost you have to cut above where you want to boost. Its inductance is high enough the it cuts out fast enough to compensate for the rise in woofer impedance that would other wise prevent the crossover cutting off the woofers properly. Often this function is carried out by adding a zobel network.

 

[TLS Guy]

As I will be making 1.25" thick cabinets, would it behoove me to modify the cabinet dimensions? I'll be using these with subwoofers, but would like to think for most music they will be fine by themselves.

Dennis Murphy said: - Looks like we'll have to agree to disagree on this one. I can't tell you how many requests I've received from Thor owners to have me rework the crossover. The complaint is always the same--the sound is thin and fatiguing. I always tell them the same thing--the crossover is fine. It's the TL that's lending the thin quality, along with the sonic character of the Excel drivers in the midbass. All the TL's I've worked with have been designed using Martin King's software, and I'm pretty sure they're optimized correctly. None of them sounds thin. And the bass I hear coming from my orchestra isn't thin either. It's a little out of tune, but not thin. Actually, I'm always surprised at how resonant it sounds compared with my speakers. But I think we can agree that transmission lines are the way to go if the woofer(s) and Significant Others lend themselves to that approach.

I think the problem is actually that the builders did not pay nearly enough attention to stuffing the line.
I built a pair for my father, and used fishing net, (The Medway Towns are among other things a fishing port), to correctly hold the damping in place and get the correct and uniform stuffing density. This is really important. The lines that I built did not sound thin at all. They sounded just right. I would bet many constructors over stuffed, or even blocked the pipe in places.

When I built my lines, I went to garden center that also sold Christmas trees, and used the netting used to hold them compressed for transfer. You can see this here.

Unfortunately many home constructors pay scant attention to the stuffing process.

Many people also bought pre made cabinets, that they had to stuff.

This can not work. You can NOT stuff a line after it is built, but only during the building process. If you stuff a built line, it will NEVER perform properly.

 

[TheWarrior]

R1 gets the Q of the filter in the correct range.

I need to find more information about this. Never do I recall reading about resistors HELPING electrical Q. I had forgotten the information about impedance matching in regards to R2/R3.

The addition of C4 shapes the tweeter response slightly.

Can you elaborate on this shape? What is the advantage over using, say, a 20uF capacitor? Or is does this still apply to impedance matching per the L-pad?

Remember a passive circuit can never boost only cut. So with a passive circuit to boost you have to cut above where you want to boost. Its inductance is high enough the it cuts out fast enough to compensate for the rise in woofer impedance that would other wise prevent the crossover cutting off the woofers properly. Often this function is carried out by adding a zobel network.

So again, its L1 that really makes this crossover work so well?

 

[TheWarrior]

Further, TLS, in regards to my question about cabinet size, is this simply a matter of opinion about what is 'adequate' bass? I definitely prefer tight, low Q bass. So I should just build the design as it was intended and quit over analyzing..... everything?

 

[TLS Guy]

Looks like we'll have to agree to disagree on this one. I can't tell you how many requests I've received from Thor owners to have me rework the crossover. The complaint is always the same--the sound is thin and fatiguing. I always tell them the same thing--the crossover is fine. It's the TL that's lending the thin quality, along with the sonic character of the Excel drivers in the midbass. All the TL's I've worked with have been designed using Martin King's software, and I'm pretty sure they're optimized correctly. None of them sounds thin. And the bass I hear coming from my orchestra isn't thin either. It's a little out of tune, but not thin. Actually, I'm always surprised at how resonant it sounds compared with my speakers. But I think we can agree that transmission lines are the way to go if the woofer(s) and Significant Others lend themselves to that approach.

I doubt the lines were built properly.

I think the problem is actually that the builders did not pay nearly enough attention to stuffing the line.
I built a pair for my father, and used fishing net, (The Medway Towns are among other things a fishing port), to correctly hold the damping in place and get the correct and uniform stuffing density. This is really important. The lines that I built did not sound thin at all. They sounded just right. I would bet many constructors over stuffed, or even blocked the pipe in places.

When I built my lines, I went to garden center that also sold Christmas trees, and used the netting used to hold them compressed for transfer. You can see this here.

Unfortunately many home constructors pay scant attention to the stuffing process.

Many people also bought pre made cabinets, that they had to stuff.

This can not work. You can NOT stuff a line after it is built, but only during the building process. If you stuff a built line, it will NEVER perform properly.

 

[TLS Guy]

Further, TLS, in regards to my question about cabinet size, is this simply a matter of opinion about what is 'adequate' bass? I definitely prefer tight, low Q bass. So I should just build the design as it was intended and quit over analyzing..... everything?

All I can tell you is that the pair I built sounded fantastic. You could make the cabinet bigger, but I would worry the bass output might be excessive. See my comments about stuffing.

You need 24.1 Oz of Polyfill per speaker. Do not use glue to keep the fill in place, it soaks in and hardens it. Keep the fill in place evenly throughout the line with net, and stop the fill about 3 to 4 inches short of the port.

 

[TLS Guy]

I need to find more information about this. Never do I recall reading about resistors HELPING electrical Q. I had forgotten the information about impedance matching in regards to R2/R3.



Can you elaborate on this shape? What is the advantage over using, say, a 20uF capacitor? Or is does this still apply to impedance matching per the L-pad?



So again, its L1 that really makes this crossover work so well?

I think you need to do some reading on your own, on the function of basic components.

It is the total design that works and not one individual component.

The L-pad is purely voltage control and obviously C4 has to shape tweeter response as a capacitor is a high pass component.

Any resistor in series with a tuned circuit, will change it parameters, and alter the width of the notch, which is Q.

 

[Dennis Murphy]

I think you need to do some reading on your own, on the function of basic components.

It is the total design that works and not one individual component.

The L-pad is purely voltage control and obviously C4 has to shape tweeter response as a capacitor is a high pass component.

Any resistor in series with a tuned circuit, will change it parameters, and alter the width of the notch, which is Q.

All I can tell you is that the pair I built sounded fantastic. You could make the cabinet bigger, but I would worry the bass output might be excessive. See my comments about stuffing.

You need 24.1 Oz of Polyfill per speaker. Do not use glue to keep the fill in place, it soaks in and hardens it. Keep the fill in place evenly throughout the line with net, and stop the fill about 3 to 4 inches short of the port.

I guess that's possible, although I think you're overstating the importance of stuffing during construction vs. post-construction. If you know how much fill to put in, and have ready access to the relevant chamber(s), I think you can get it right, or right enough. But I don't doubt that a lot of Thor builders crammed in a whole lotta stuffing. As for my comment about the Thor crossover being "conventional," that's because almost no one uses the old text book construct of separate impedance compensation or text book values for components. Optimization programs let you do all that in one integrated circuit, and for the magnesium Excel woofers with a single symmetrical breakup mode, you always get where you need to go with one, and at times 2, series inductors and a trap that flattens the breakup and hits the 4th order slope function. Sometimes you need a resistor in the trap, sometimes you don't.

 

[TLS Guy]

I guess that's possible, although I think you're overstating the importance of stuffing during construction vs. post-construction. If you know how much fill to put in, and have ready access to the relevant chamber(s), I think you can get it right, or right enough. But I don't doubt that a lot of Thor builders crammed in a whole lotta stuffing. As for my comment about the Thor crossover being "conventional," that's because almost no one uses the old text book construct of separate impedance compensation or text book values for components. Optimization programs let you do all that in one integrated circuit, and for the magnesium Excel woofers with a single symmetrical breakup mode, you always get where you need to go with one, and at times 2, series inductors and a trap that flattens the breakup and hits the 4th order slope function. Sometimes you need a resistor in the trap, sometimes you don't.

I agree completely with your comments about the driver. I use the same type of topology with my SEAS units and it works fine.

As regards, fill I think you do have to be somewhat obsessional about it.

With the narrow port on the rear, and the turn, I don't see how you would have adequate access to the line once the cabinet is complete. I could be wrong about that. I have been suspicious about this all along. All I can tell you is that the ones I built for my father sounded excellent. After his death I set them up for my youngest sister. She is a semi-pro Mezzo and loves these speakers.

All I can say is, if they sound bad, then someone did not build them properly.

 

[TheWarrior]

What do you mean a guy with no hands on experience can't master speaker design after reading a single book and never actually applying what was gained!? Lol

I'd just like to thank Dennis Murphy and TLS Guy for your commentary. I'll be able to refer back to this after I read, study and apply. Probably will present a theoretical crossover as my 'test' in applying the knowledge. Sure it can be scrutinized as there may be multiple approaches in certain situations, but I enjoy this stuff enough to really challenge myself to know it better!