JonMarsh

Notice pricing has increased for all the designs in this section of the forum. People should use the BOM's and check with suppliers for current prices

Modula MTM-CT Design and Construction- Introduction

Last updated 2-10-2005

This thread is going to detail the driver measurements, contruction details, crossover design, and final performance tuning for the MTM speaker using two Dayton RS180 and a ScanSpeak D2904/6000-01 tweeter. This is intended to be a high bang for the buck system, but to get the optimum level of performance from these drivers, a slightly more expensive crossover is usually necesssary. Still, the bottom line looks attractive compared with the performnace. This post will be edited and updated to reflect work in progress as steps are completed.

My personal versions of this design will be built in Parts Express 1 cu. ft enclosures designed for MTM's.

The long Xmax of the shielded Dayton RS180 (p/n 295-364) and the low non-linear distortion suggests it should provide compeitive performance, as long as a lower crososver point is used.

Measurement show that SS D2904/6000-01 offers low non-linear distortion and smooth wide frequency response; it should, since it's not exactly inexpensive. What it brings to the party is a very small footprint due to the use of nedymium magnets, and very low distortion due to the SD1 motor design from ScanSpeak.


Drivers - The Dayton Shielded RS180, (p/n 295-364)


The RS180 is a 7" aluminum cone driver using a phase plug, and featuring a relatively long Xmax for a driver of this type. Furthermore, as Mark K and MFK have shown, measured non-linear distortion and ETC tests are very good, competitive with well known drivers costing much more. However, like most metal cone drivers, they have upper range resonances which require more care in crossover design.







Here the Dayton RS180 (single driver) is measured in room in a Woodstyle test enclosure with the same baffle width as the planned enclousres (11"). Little smoothing was used, and the relatively long gating time shows the effects of room reflections.




Shown here is an impedance sweep for the Dayton RS180; especially the impedance curve above the midbass dip is critical information for the crossover design.





The SS D2904/6000-01 is a very compact tweeter, with linear repsonse extending to about 1 kHz in spite of it's dimunitive size.

This is the factory supplied FR data; later I'll post the measured FR in a comparable width cabinet to the actual design.

http://www.madisound.com/pdf/scansp...04_6000-01e.pdf

Cabinet Design and Construction

I chose to develop these projects for the PE (Parts Express) cabinets, because for the money, it's very hard to beat the quality and finish considering the time involved. The cabinets are well braced internally, also.




If you want to go deeper on the enclosure and use a larger port for more bottom end extension, then you may want to build something your self similar to the FP layout of the 1 cu ft enclosure.

Shown below is the front panel layout diagram for the center channel MTM. The tweeter is offset substantially because this allows moving the midwoofers in close to each other, and produces a flattter boundary and diffraction behavior in the 1-3 kHz reigon, according to Baffle Diffraction Simulator.






Crossover Design


Top optimize the performance, considering the large distance between the midwoofers and the cone resonances of the midwoofer above 5 kHz, a low, steep crossover point is preferred. The schematic below shows the topology for a cauer-elliptic filter two way crossover, and the component values for these drivers developed from the measured data. The crossover is designed to provide baffle step compensation and an acoustical 8th order Linkwitz-Riley transfer function for the first 48 dB of attenuation. The crossover frequency is nominally 1200 Hz.

These are the release candidate component values being ordered for testing; I spent some time and effort to try to have standard values so that unwinding coils, for example, wouldn't be necessary. The tweeter series caps will have to be made up from paralleled values; the values shown should be used with tolerances of a few percent, no more than 5% deviation.

If I did my work well, the final actual values "should" be pretty much as shown- this worked out pretty well with my last project, the Arvo Part. I really only expect that there might be some change in the tweeter LPAD resistances- I doubt any other component changes will occur. But who knows?

Note that rather than optimize primarily for on axis flat amplitude, I optimized the design more for total power response, especially off axis to 15 to 30 degrees. This results in some slight compromises in the on axis performance, but I believe that for the special case of a center channel speaker, this is a better choice, unless there's only one viewer/listener sitting dead on axis. In MOST home situtations, this won't be the case.




Here's the predicted SPL magnitude on axis:





Here's the predicted lateral (horizontal) polar response, based on baffle size, driver arrangement and size, and driver radiating diameters.




And the predicted off axis at 15 and 30 degrees, plus the net power response over that included angle - you can see that the power response into the included angle is how I've optimized the design:



Anyone wanting to build this speaker as an in-wall or on-wall unit should use the schematic seen in post #30 of this thread.

People should read this website regarding the placement of inductors prior to building their crossovers.
http://www.troelsgravesen.dk/coils.htm

__________________
the AudioWorx
_______________________________

Modula MT
M8ta
Modula Neo D CC
Ardent
Modula MT MkII

coming in 2010....

Modula MTWW
Modula TMWW Xtreme
Ardent CCX (Compact Center Xtreme)
Ardent Cxi (tower, updated design)

Steve Goff

Another great project, Jon! The off axis and net power responses look good. Am I correct in concluding that these curves are off the long axis of the box, since the speaker is meant to be wider than it is tall?

__________________
Steve Goff

JonMarsh

Yes, that's exactly right, Steve- though the dispersion in the other diection should be pretty decent also- it's actually easier that way, of course.

In a sense, trying to make a good MTM center is a little bit of trying to make a gold purse out of sow's ear- the "right" way to do it would be a WMTW, with the MT vertical between the W's; I've done that kind of design before for a couple of guys, but it's a lot more expensive to build. I'm hoping that this variation and optimization will do the trick for a moderate size system. I'm building one for myself and an old friend here locally, to start.

What will be interesting to see is how the perceived tonal balance compares with the other more conventional MTM- I have a suspicion that the power response optimization may be more important than the "on axis" optimiztaion, even if you don't use the speaker laterally. It will take a bit of time with music and HT material to suss this, though I suspect also that with pink noise some opinions might be developed pretty quickly- I've found listening to pink noise to be very helpful in the past in fine tuning crossovers quickly- more quickly than can be done with music.

~Jon

__________________
the AudioWorx
_______________________________

Modula MT
M8ta
Modula Neo D CC
Ardent
Modula MT MkII

coming in 2010....

Modula MTWW
Modula TMWW Xtreme
Ardent CCX (Compact Center Xtreme)
Ardent Cxi (tower, updated design)

PMazz

Go Jon, Go!!!! You've been busy lately.

Looks like it's time to make up an order for more GE caps. I'll be getting to build your Modula MTMs with the Seas tweet (for a friend) as well as this center channel (for myself). I already had 6 more 180s added to my backorder at PE, 20 in total. I'm looking forward to comparing the M8 MTMs to the Modulas. I wonder which designer will take the prize? Should be interesting....if I ever get them done, of course.

Many thanks,
Pete

TacoD

Interesting read, I don't seem to be able to simulate the power response in LspCAD 5 std edition. But the simulated of axis behaviour give a good enough approximation of the true 4Pi powerresponse ain't it?

JonMarsh

Yes, I think it does- you just have to do a little more work interpreting it.

LspCAD6 is nice in that you can specify the off axis angles you're interested it, weighting them, if you will, for a specific range, and it calculates the net power response averaged for those angles. It's a nice piece of software, though it's good that it has the full LspCAD 5.25 built in also, because there is not automatic update conversio- working concept, file structure, everything is totally different. It's a whole new program.

~Jon

__________________
the AudioWorx
_______________________________

Modula MT
M8ta
Modula Neo D CC
Ardent
Modula MT MkII

coming in 2010....

Modula MTWW
Modula TMWW Xtreme
Ardent CCX (Compact Center Xtreme)
Ardent Cxi (tower, updated design)

JonMarsh

That reminds me, I need to get a new PE order in! I need more RS180's and crossover parts, too!

~Jon

__________________
the AudioWorx
_______________________________

Modula MT
M8ta
Modula Neo D CC
Ardent
Modula MT MkII

coming in 2010....

Modula MTWW
Modula TMWW Xtreme
Ardent CCX (Compact Center Xtreme)
Ardent Cxi (tower, updated design)

cjd

Is that vertical or horizontal off-axis? Looks like it's "rotating" around the MTM line, not to one side of the MTM or the other...? My MTM is disappointing off-axis when it's in CC orientation, measured to the side.

I did have someone mention that an odd order crossover slope might work better in this instance. Anything worth looking into?

C

Steve Goff

cjd,

See my post and Jon's answer above: the off axis graphs show response along the MTM axis, since the speaker is meant to lie down, being wider than it is tall.

__________________
Steve Goff

cjd

deg nabbit. I read that, and didn't fully make sense. Went back over Jon's work and NOT the subsequent posts and then posted my Q. Ahh well. It's what I get I guess!

C

JonMarsh

Yeah, the "hard way" to do it is off the MTM axis, as for a laydown. You know the old line, "We can do this the easy way, or we can do this the hard way..."

This is the "hardway", and while I wish it could be even better, I think it's pretty dang respectable- maybe even good enough?

~Jon

__________________
the AudioWorx
_______________________________

Modula MT
M8ta
Modula Neo D CC
Ardent
Modula MT MkII

coming in 2010....

Modula MTWW
Modula TMWW Xtreme
Ardent CCX (Compact Center Xtreme)
Ardent Cxi (tower, updated design)

cjd

heh. I just tell everyone they all have to crowd into the same area. Though, I'm not actually using these MTMs in my HT, and the CC is for my sister. 20 degrees off-axis is fine, but more than that and it tanks. Pretty crazy. Very enlightening for me though.

Any thoughts on slope order though? I had figured it was almost entirely an issue of off-axis interaction of the mids, which couldn't really be fixed by crossover...

JonMarsh

Well, it is an off axis interaction of the mids, based on the wavelengths- the lower you can crossover, the less the interaction is before getting over to the tweeter.

So, there seems to be two factors at work- the higher order crossover allows crossing the tweeter over because you don't end up pushing it on the bottom end too hard (this network is down 10 dB at 1 kHz and -20 dB at 800 Hz, which is OK for the SS6000, and corresponds with a much higher crossover frequency with slower slopes); and the steep slope means you get off the midwoofer lower AND faster, so there's not as much residual interaction, because the -6 dB or -9 dB area is reached rather quickly.

This acoustic transfer function also has a wider crossover lobe, so you don't have the problems in the crossover region that a 4th order L-R has, plus I'm bumping up the on axis crossover region and just below it a little bit, so that the net power response is pretty smooth.

Well, I shouldn't sprain my arm patting myself on the back for this until the first one is finished and I'm listening and happy with it. As Willie the Shake said, there's many a slip twixt the cup and the lip. Cross fingers and order crossover parts and more drivers this weekend...

__________________
the AudioWorx
_______________________________

Modula MT
M8ta
Modula Neo D CC
Ardent
Modula MT MkII

coming in 2010....

Modula MTWW
Modula TMWW Xtreme
Ardent CCX (Compact Center Xtreme)
Ardent Cxi (tower, updated design)

Steve Goff

I hope you stole that from Willie: You know, "neither a borrower nor a lender be."

__________________
Steve Goff

cjd

It starts its big dip at 775Hz (-30dB). It averages about -12dB. Though, those measurements were taken when there was still some, uh... ironing out I had to do with the setup. Hope yours works a bit better! JeffB's PDC didn't project anything near this - it does vertical off-axis, so a traditional MTM should in theory reflect the off-axis response of a CC.

C

Captain Cojo

Greetings, Jon!
Any chance of getting more info on your WMTW center? Perhaps with the extremis? Thanks!

PMazz

FYI: Dayton RS180s are back in stock at PE! Get em while they're hot.

Pete

cjd

Captain Cojo: since this thread is about Jon's MTM CC I'll keep this short... but there are folks working on WTMW options (including me - doing a whole slew of options for HT use) using the RS drivers.

C

Captain Cojo

deleted by author

PMazz

My batch of RS180s came in today!

Jon, before I order xover parts, are these values the final design? Do these have BSC and/or should I worry about it? Or should I just wait a while longer?

Pete

JonMarsh

This design DOES have BSC; unless you mount the baffles flush with your wall or ceiling, you need some. Like vitamin C, you know.

I'm doing inventory and ordering parts this week; obviously, the safest thing is to wait for me to finish full testing with the actual crossover.

So, if I put on my Clint Eastwood serape (like Marty did in the last of the BTF trilogy), I have to ask, "Are you feeling lucky?" If so, you can order parts now.

It's unlikely any of the inductors will change in value; the caps might get trimmed a little one way or the other, but probably not the woofer crossover caps. Resistor adjustments for the tweeter level are quite possible... The last couple of projects I did pretty much worked as predicted, with only minor tweaks to the Lpad values. The ARvo's at ThomasW's were only LPAD tweaks- everything else worked as advertised.

So, as the crossover values go, non of the inductors are likely to change, the tweeter zobel networks won't change, R2/R3 and C5/C6 might get tweaked a bit; those caps are normally made from paralleled parts, so it's easy to add or take away; say, a 10 uF, 5 uF, and 3.3 uF for C5, for example. I'll be using film and foil for part of those values. C6 could be a 20 + 1 uf, but I'll probably use 5 uF film and foil there, also, so it will be a cluster of values.

Your choice... depends on your time table. I'm building baffles and ordering parts this weekend.

~Jon

__________________
the AudioWorx
_______________________________

Modula MT
M8ta
Modula Neo D CC
Ardent
Modula MT MkII

coming in 2010....

Modula MTWW
Modula TMWW Xtreme
Ardent CCX (Compact Center Xtreme)
Ardent Cxi (tower, updated design)

PMazz

As a matter of fact........




Pete

JonMarsh

OK, Pete, so I go back and "Do one for the Mazz Man", without BSC.

As my friends in Villach would say, "kine problema".


It'll be done by sometime this weekend....

__________________
the AudioWorx
_______________________________

Modula MT
M8ta
Modula Neo D CC
Ardent
Modula MT MkII

coming in 2010....

Modula MTWW
Modula TMWW Xtreme
Ardent CCX (Compact Center Xtreme)
Ardent Cxi (tower, updated design)

capslock

Vile dialect!

I hear their DRAM business is not doing too well, and they are selling off and closing fiber optics. I do hope power semiconductors are doing better.

By the way, ordered Infineon DDR400CL3 memory for my upgrade, because it was nearly the cheapest and the A64 is not that dependent on memory bandwidth. All the same, those overclock nicely.

PMazz

Oh, I thought it was something you just took out of the xover.....Sorry for creating yet more work. Maybe you can put the Evil Twin to work on some of this.

Thanks Jon.

Pete

Evil Twin

"Yeah, that's a good idea- get Evil Twin to do the work! Trouble is, he seems to just want to work on the "glamour" projects, like that overpriced, oversized "Saint-Saens" Death speaker or whatever it is... what we really need is a more complex, less expensive flux capacitor, but he just want's to develop a control interface betweeen a turbolaser energy accumulator and his modified Aragon monoblocks to drive the Ascendant audio dipole subs..."

This attitude is disturbing...

Do not be concerned, Pete. There aren't any components removed, it's just a matter of adjusting values in the low pass circuit, and resetting the tweeter Lpad. Plus, you'll pick up a fair amount of net sensitivity... this doesn't hurt, either.

Witness the power of the dark side- you'll have your crossover re-design today.

__________________
Your lack of confidence in the dark side is disturbing... do not forget that black is back...

PMazz

Many thanks Dark Lord.

Maybe I can wash and wax the old Ty Fighter?

Pete

JonMarsh

I have an updated design which looks pretty good- also tweaked the tweeter section, but also should discuss some points, I'm thinkin' - specically, what amplifier and load drive capability will you have for the CT? When one parallel's these Daytons, the impedance minima at 150 - 350 Hz is pretty low- and with a crossover without BSC compensation, it's even lower (smaller inductors). I've got a very nice response curve, I think, but over much of the frequency range the impedance is around 3 ohms below 1 kHz. A good solid amp with good 4 ohm ratings wouldn't have any problems, but an amp rated for 6-8 ohm loads only, like many HT units, wouldn't like it.

I could do a design variant with the woofers in series- this would be more amplifier friendly, but would knock down the net sensitivity for 2.83 VRMS (1 watt at 8 ohms) from ~ 92 dB to ~ 86 dB. The latter isn't all that bad compared to most commercial speakers. The former configuration would be optimum for an amp with lower output voltage, but relatively high current, like some HK's, NAD's, Anthem's, Bryston's, and of course, my MOR standby, Aragon.

Let me know what you think would work best for you.

I'll have the non-BSC design posted a little later this weekend- got chores to do now!

~Jon

__________________
the AudioWorx
_______________________________

Modula MT
M8ta
Modula Neo D CC
Ardent
Modula MT MkII

coming in 2010....

Modula MTWW
Modula TMWW Xtreme
Ardent CCX (Compact Center Xtreme)
Ardent Cxi (tower, updated design)

PMazz

My amp is the B&K Ref 7250. 200W@8ohm, 375W@4ohm.

Ref 7250 manual

My guess is it would be fine.

Now stop working and go get some work done!

Pete

JonMarsh

OK, here's the late weekend update, film at 11....


Schematic: note addition of tweeter contouring network (inductor and resistor in parallel).




On axis pressure amplitude response:




Off axis and net power response averaged over 0, 15 degrees, and 30 degrees off axis both directions on driver axis (gives somewhat greater weighting to off axis behavior, of course).




Baffle should be aimed somewhat downwards, if you can, and if it will help I can send you a piece of WhisperMat to help kill the ceiling reflection near the speaker (about 2 ft by 1 ft; acoustic foam with high densit vinyl).

__________________
the AudioWorx
_______________________________

Modula MT
M8ta
Modula Neo D CC
Ardent
Modula MT MkII

coming in 2010....

Modula MTWW
Modula TMWW Xtreme
Ardent CCX (Compact Center Xtreme)
Ardent Cxi (tower, updated design)

PMazz

Yep, first thing that jumped out at me. I said "Damn, that's a tweeter contouring network there." Jon, for all I know, those schematics could be the innards of a portable yak foreskin removal machine.

Many thanks!

Pete

Paul H

You'll want to start that rig up Very carefully Pete once it's built.

Paul

Ten 99

Careful now, you'll get Hank all excited.

JonMarsh

There are export restrictions on Yak foreskin removal equipment from California; we're allowed to ship to Texas, Alabama, Arkansas, Louisianna, and Mississippi, but states farther north or east are proscribed without a Federal Yak Export licensce, and those are a bear to get these days, since homeland security confused Yak foreskins with Russian air to ground attack aircraft.

Sorry, Pete.

__________________
the AudioWorx
_______________________________

Modula MT
M8ta
Modula Neo D CC
Ardent
Modula MT MkII

coming in 2010....

Modula MTWW
Modula TMWW Xtreme
Ardent CCX (Compact Center Xtreme)
Ardent Cxi (tower, updated design)