Tweet
For those who asked, here is the RS180S-8 and RS28aS-4 MTM design that I did for a local recording studio. I was very pleased with how these sounded in my system. I could have lived with these speakers for quite a while. The owner of the studio still tells me how much he loves them everytime he see me.
The goal of the design was a speaker that had reasonbly decent sensitivity, good bass extension (there is a subwoofer in the room too), and very flat response, but was also capable of fairly high output levels when needed too.
I chose the MTM because of the extra cone area and the symmetrical radiation pattern. I kept the crossover point pretty low to accommodate the driver sizes and their spacing so that lobing would not become much of an issue. The crossover point is at about 1.7kHz. The acoustic crossover is a 4th order L-R type. I really prefer this with MTM's. The in-phase type crossovers have flat on-axis response but a 3dB dip in the power response. I find many MTM's to sound a little harsh or forward in the crossover region, probably due to lobing from the two midwoofers. I find the dip in the power response really takes the edge off of this and results in a very nice, highly listenable speaker. In my opinion, the preference many people have, including D'Appolito himself, for even order crossovers in MTMs is not due to the difference in lobing as much as due to the dip in the power response, but that's just my two cents worth.
Here's a pic of the speakers using the Parts Express 1.0 Cuft cabinet, ported in the back, and the tweeter offset about 1/2"
Here's the on-axis frequency response with the reverse null shown (for those that like that sort of thing...)
Here's the acoustic crossover of the drivers with a little different scale:
Here's the acoustic phase on the tweeter axis. I've seen the ads for the "best speaker on earth. Period." I think I can compete with it OK. Phase tracking is near perfect over a very wide bandwidth, with the textbook 180 degree wrap at the crossover point. Beat that :D
Here is the input impedance of the speaker. Pretty typical curve, but definitely should be treated as a 4 ohm load.
And finally, here is the crossover schematic. I believe in parallel crossovers and keeping them simple wherever possible. As you can see I achieve flat frequency response, a textbook acoustic crossover, and excellent phase tracking with only six parts. Again, in my opinion, this is the way a crossover should be done. Part of the trick here it let the drivers "talk to you" about the crossover point. There will tend to be a natural crossover point that allows the phase to come into alignment with a fairly simple circuit. I don't attempt to "force a crossover" to hit a predetermined frequency. Rather, I find the frequency where everything seems to come together at. I learned this a long time ago using CALSOD's optimizer. It always worked its way to a point like this and then tried to eliminate parts by making their value extremely small or large (depending on the component). After a while I just learned to do it myself without using any kind of optimizer.
As you can see the woofer circuit is simply a single 1.2mH inductor with a two element trap. My inductor was a fairly heavy gauge air core, but I don't remember the exact details. If you want to use a P-Core or laminate that would be fine too. I would recommend keeping the the DCR as low as possible here though. The little inductor in the trap is a Jantzen 18 ga., the PE part number is 255-198 (mine actually measured .023mH, which helped a little by tuning the trap a hair higher).
You will note the large values of the capacitors in the tweeter circuit. This is due to the fact that it is a 4 Ohm tweeter. If it had been an 8 ohm tweeter, or had some series resistance with it the caps would have been smaller. The 10uf cap is pretty easy to come by in a decent grade. However, I realize that 50uf poly's can be pretty salty. Just to let you know, I used an NPE in this location. I can't say it bothered me any. Like I said, they sounded very nice to me. If you build it, I'll let the choice be yours.
If you build it, as always, let me know what you think.
Jeff B.
The goal of the design was a speaker that had reasonbly decent sensitivity, good bass extension (there is a subwoofer in the room too), and very flat response, but was also capable of fairly high output levels when needed too.
I chose the MTM because of the extra cone area and the symmetrical radiation pattern. I kept the crossover point pretty low to accommodate the driver sizes and their spacing so that lobing would not become much of an issue. The crossover point is at about 1.7kHz. The acoustic crossover is a 4th order L-R type. I really prefer this with MTM's. The in-phase type crossovers have flat on-axis response but a 3dB dip in the power response. I find many MTM's to sound a little harsh or forward in the crossover region, probably due to lobing from the two midwoofers. I find the dip in the power response really takes the edge off of this and results in a very nice, highly listenable speaker. In my opinion, the preference many people have, including D'Appolito himself, for even order crossovers in MTMs is not due to the difference in lobing as much as due to the dip in the power response, but that's just my two cents worth.
Here's a pic of the speakers using the Parts Express 1.0 Cuft cabinet, ported in the back, and the tweeter offset about 1/2"
Here's the on-axis frequency response with the reverse null shown (for those that like that sort of thing...)
Here's the acoustic crossover of the drivers with a little different scale:
Here's the acoustic phase on the tweeter axis. I've seen the ads for the "best speaker on earth. Period." I think I can compete with it OK. Phase tracking is near perfect over a very wide bandwidth, with the textbook 180 degree wrap at the crossover point. Beat that :D
Here is the input impedance of the speaker. Pretty typical curve, but definitely should be treated as a 4 ohm load.
And finally, here is the crossover schematic. I believe in parallel crossovers and keeping them simple wherever possible. As you can see I achieve flat frequency response, a textbook acoustic crossover, and excellent phase tracking with only six parts. Again, in my opinion, this is the way a crossover should be done. Part of the trick here it let the drivers "talk to you" about the crossover point. There will tend to be a natural crossover point that allows the phase to come into alignment with a fairly simple circuit. I don't attempt to "force a crossover" to hit a predetermined frequency. Rather, I find the frequency where everything seems to come together at. I learned this a long time ago using CALSOD's optimizer. It always worked its way to a point like this and then tried to eliminate parts by making their value extremely small or large (depending on the component). After a while I just learned to do it myself without using any kind of optimizer.
As you can see the woofer circuit is simply a single 1.2mH inductor with a two element trap. My inductor was a fairly heavy gauge air core, but I don't remember the exact details. If you want to use a P-Core or laminate that would be fine too. I would recommend keeping the the DCR as low as possible here though. The little inductor in the trap is a Jantzen 18 ga., the PE part number is 255-198 (mine actually measured .023mH, which helped a little by tuning the trap a hair higher).
You will note the large values of the capacitors in the tweeter circuit. This is due to the fact that it is a 4 Ohm tweeter. If it had been an 8 ohm tweeter, or had some series resistance with it the caps would have been smaller. The 10uf cap is pretty easy to come by in a decent grade. However, I realize that 50uf poly's can be pretty salty. Just to let you know, I used an NPE in this location. I can't say it bothered me any. Like I said, they sounded very nice to me. If you build it, I'll let the choice be yours.
If you build it, as always, let me know what you think.
Jeff B.
Comment