MTM

This MTM system was kicking around in my mind for almost two years. It was replacing an MMT system that I had in the room that used two 5-inch drivers that were no up to the task of reproducing the 100-200Hz region. The new design needed to fill a couple of goals: (1) a speaker for everyday listening at a variety of sound levels in a large room, (2) a medim sized speaker that could reproduce everything from about 100Hz on up and (3) I need to STOP BUYING MORE DRIVERS and instead use four 8" driver that I had left over from the Attic Subwoofer, which were originally in the 4-Active Tower. These excellent drivers are Seas L21 RNX/P (click here for data sheet) an aluminum cone design with phase plug. These had a nice response except for some nasty breakup peaks starting around 4kHz.

I knew that it would be pushing it to use these 8" woofers in an MTM system however if I could find a suitable tweeter that could crossover low (around 1.5kHz) then I would be able to push the breakup peaks far enough into the stop band to make them unnoticiable. I did a lot of searching for data on the web and heavily relied on the measurements of frequency response and distortion that can be found at Zaph Audio. In the end I decided that, at the price point, the Vifa DX-25 1-inch dome could not be beat and looked like it could be crossed over pretty low...

I summoned my trusty simulator, Unibox, and ventured into the unknown... Remember, I was recently using these drivers in a vented subwoofer down to 20Hz. Now I needed to shift gears entirely and try to fit them in to a small closed box. Here are some plots of the resulting design target for the system:

NOTE: while the graphs show 100W power input, it is actually 50W applied to a parallel pair of the L21's. I usually shoot for 110dB as my Xmax and power limited performance spec. This leaves lots of headroom for 'spirited listening".

After seeing these plots, I was happy to know that I could meet my low-end design goals with these drivers! With a 20L box, I would get a slightly high Q that I could control with a dense application of acoustic stuffing. With a final Q of around 0.85, this alignment has the advantage that the cone excusion is nicely controlled below resonance. Although the speaker passband extends only down to a little less than 100Hz so I was planning to cross over to an as-yet-to-be-built dedicated subwoofer in the 80-100Hz region anyway.

Since I couldn't quite shoehorn these drivers into a prefab Parts Express or Madisound cabinet I knew that a cabinet would have to be custom made for this project. I wish that I was one of those guys who could disappear in to my workshop and emerge with a beautiful veneered cabinet... but I'm not. I have little woodworking tools, experience or talents so I decided to look for someone that was up to the task make the cabinets for me. Luckily I found Brian King who was at least in state (CA). Click here to contact him via EMail if you are interested in having him build some cabinets for your project.

These aren't floating in air, I have them mounted to the wall using some "Smarter Speaker Mount" speaker brackets (part number 240-710) that I purchased from Parts Express. I highly recommend these-they are easy to work with and hold up to 50 Lbs (and more, says the manufacturer). Since my MTMs weight about 40+Lbs, there were not too many options but this gets the thumbs up from me for sure.

I set up the speakers and then did a couple of days experimenting. My setup includes a Behringer DCX2496, which I find invaluable for (relatively) quickly developing and fine tuning crossovers, even complicated multi-way ones, using test material, some measurements and extensive listening with well-known source matierial. After I went through this process, I ended up with something that was very close to what I had envisioned. The interesting part came when I had to choose the crossover frequency. I had originally wanted to crossover at about 1.3-1.5kHz, whcih is pretty low even for this tweeter but I thought it was maybe do-able. I tested several crossover frequency and slope combinations by shutting down the amp for the woofers and listening only to the tweeters. At higher playback levels, I could definitely discern some stress with sub 1.4kHz crossover points, even with an 8th order slope! I ended up using a 24dB/oct Linkwitz-Riley crossover at 1.9kHz. This was higher than I wanted by there were no problems with the cone breakup modes of the woofers because I was able to knock them down to inaudibility with a sharp notch filter.

The resulting frequency response is shown below, measured in the far field. I didn't use any fancy techniques so the two dips are probably some room resonances coming in to play. Not bad, eh?

This is probably one of the flattest frequency responses that I have achieved. It's about +/-2dB above 100Hz. The fall off at the high end is my microphone, not the tweeter.

This system should be around for a long time. It has definite spousal approval and a great soundstage to boot! Here is one more shot, looking up at the wall-mounted speaker.