Tweet
This design came together quick once I got to the point of selecting drivers. I spent several months reading and learning how to design xovers, here and over at PE Tech-Talk.
I had thought about using many drivers and many configurations but couldn't resolve (in my own mind) various issues. And then Zaph released the ZMV5. I thought, "Maybe I could do a ZMVM5." But the MCM 5.25" driver requires a lot of volume and I wanted small. Then Zaph noted the MCM 4" cast frame driver, 55-3853, would drop into the Bargin Mini (aka. ZBM4). Ding. The MCM 55-3853 works very well in a small vented enclosure (F3 is equivalent to the 5.25" driver in a third the size).
Zaph says about the ZBM4, "This system represents a very good value, and performs excellently regardless of price. Small, smooth and clean sum it up nicely. On the negative side, there are really only two issues: output is severely limited when running full range and driver mounting is somewhat difficult." This was stated about the stamped frame MCM driver. The cast frame exhibits slightly better distortion. An MTM of the 55-3853 would solve the first problem and using Nik Brewer (CNC operator) solved the second.
The tweeter was simple, the Vifa DQ25SC-16 as used in the ZMV5. The performance for the price could not be questioned and Zaph proved it's compatibility with the MCM drivers.
So, the result:
The tweeter high-pass circuit should be familiar to those who've looked at the ZMV5, as it's very similar: 2nd order, 8.2uf cap in series and 0.3mH coil (0.26 ohms) with an l-pad of 3ohms in series and 6 ohms in parallel. My design uses less baffle step compensation than the ZMV5 justifying the lower coil inductance and the twin 4" drivers are more efficient resulting in the different padding. The impedance of the tweeter and circuit is quite respectable and shouldn't cause any problems. As Zaph did with the ZMV5, I left the top octave alone. And I haven't heard any issues in doing so.
Although the MCM 55-3853 is capable of being crossed higher, I noticed in Zaph's measurements that the 3rd, 4th and 5th order distortions start rising above 2kHz. So, I followed the ZMV5 lead. In the end, the xover point is at 2100 Hz. This also helps with appropriate C2C spacing.
The pair of 55-3853 drivers summed quite nicely into 87db sensitivity. Zaph used a 2nd order low pass in the ZBM4 and a 3rd order in the ZMV5 (the 55-3870 is also a cast frame, metal cone driver that is essentially the big brother to the 55-3853). I tried both and was having some issues with impedance dropping too low. In the end, I used a 3rd order electrical with a Zobel and as you can see above, there is no issue. Minimum impedance is at 3.6 ohms, a classic 4-ohm number. This is a little more "complex" than I would have liked but there is no real fault with it other than parts count. The reverse null is ~40db down. The components for the M's are (using PCD nomenclature):
L2 => 0.75 mH (0.42 ohms)
C2 => 15 uF
L3 => 0.15 mH (0.16 ohms)
C5 => 24 uF (Zobel RC)
R7 => 5.1 ohms (Zobel RC)
For my build, I happened to have a couple of 0.1 uF high quality caps that I used to bypass the 15 uF electrolytic. I don't know if it make a difference in the sound but I had'em at hand, so why not. For the 24 uF, I used 2 parallel 12uF caps. These numbers represent about a -4db baffle step diffraction (-6 to -7db is normal). I did this be my speakers could not be place away from other larger objects. This turned out to be a lucky guess as I wouldn't change anything.
Cost wise, it was $55 for the drivers and $25 for the xover, per speaker. Shipping was free but nothing else was on sale. So this is an inexpensive speaker.
The enclosure is 7.5 liters net volume vented with a 2.5"x1" slot port that's 6.75" long. I lined the walls and added a little Acousta-Stuff in the corners and behind the tweeter. The baffle is 13"x6". All drivers are off-center of the vertical axis. The port is on the front, next to the tweeter. The 55-3853's are above and below. This may be difficult to envision but pictures will be posted soon. The cool thing is that the other 5 pieces each have an 11.5" dimension, meaning one cut and easy fitting: #3 11.5"x6" and #2 11.5"x10.75". And as 0.75" of the port is in the front face, the port can be cut to 6". This is a simple, table-saw speaker. Since I had my baffles machined by Nik, it was a very simple build.
I will state now, and this will be obvious in the pics, that I misquoted the outer dimension of the DQ25. It should be 2.56" (and not 2.625"). If you order the baffles from him, and I will say the pincushion cutout was fantastic, make sure he's incorporated the change.
So, how do they sound? As Zaph said, smooth and clean. They're very articulate with the DQ25 and the MCM drivers are clear. Every instrument sounds like itself with secondary reverberations sounding quite natural. There doesn't seem to be any coloration from the speaker itself. The bass extension is very good and has been improving as the drivers break-in. I was able to greatly decrease the xover point and level of my sub (more so an indication of the poor quality of the other speakers). Integration between the mains and the sub is superb (although the sub is getting replaced in a couple of months). I listened with my AVR set to both large and small. The upper midrange did suffer a bit when set to "large" but with more extension and still very listenable. If I didn't have a sub, I'd use them full range. BTW, Fb is ~57Hz with an F3=~48Hz. Xmax occurs with ~20W.
The MTM is not obviously loud. I was listening to some music and wandered into another room to fetch something and realized the output was much greater than my other speakers. But sitting in front of them, at that level, there was no strain from the speakers, or fatigue in my ears. By the same token, when turned down, there was no change in the frequency response. Everything came across quite even. Nothing extra in the upper end and no decrease in the midbass. It also seems to have a decent off-axis response (determined by ear). Vertical lobing is there but wasn't noticeable (i.e. listening for drops in level) until standing within 1.5 meters (closer than the listening positions in my room).
The MTM does have the "air" of some high-end speakers I've heard but that's not a fault, just a reflection of what went into it. This is a great sounding speaker built from inexpensive components.
This isn't a monitor. Inside of a meter, the multiple drivers becomes obvious. This is not for a large room. My room is 11x14 and it's perfect. IMO, it would work well in a room up to 15x20 but I think it would "get lost" in anything bigger.
Note that this was designed entirely with simulation. I don't have the capability to measure, either the drivers or the speakers. And considering how happy I am with the results, I know I was lucky. Could this speaker be improved? I think there are a few tweaks that would improve the "weight" of the midrange and it may be better without the Zobel if the response could be aligned without it. But for now, I'm just going to enjoy them for a while. I have a sub to design and build, and then I have some other speakers to look at (my computer speakers seem too dynamic now.)
So, there you have most of the story. Pics and some graphics will be forthcoming. Feel free to ask any questions.
I had thought about using many drivers and many configurations but couldn't resolve (in my own mind) various issues. And then Zaph released the ZMV5. I thought, "Maybe I could do a ZMVM5." But the MCM 5.25" driver requires a lot of volume and I wanted small. Then Zaph noted the MCM 4" cast frame driver, 55-3853, would drop into the Bargin Mini (aka. ZBM4). Ding. The MCM 55-3853 works very well in a small vented enclosure (F3 is equivalent to the 5.25" driver in a third the size).
Zaph says about the ZBM4, "This system represents a very good value, and performs excellently regardless of price. Small, smooth and clean sum it up nicely. On the negative side, there are really only two issues: output is severely limited when running full range and driver mounting is somewhat difficult." This was stated about the stamped frame MCM driver. The cast frame exhibits slightly better distortion. An MTM of the 55-3853 would solve the first problem and using Nik Brewer (CNC operator) solved the second.
The tweeter was simple, the Vifa DQ25SC-16 as used in the ZMV5. The performance for the price could not be questioned and Zaph proved it's compatibility with the MCM drivers.
So, the result:
The tweeter high-pass circuit should be familiar to those who've looked at the ZMV5, as it's very similar: 2nd order, 8.2uf cap in series and 0.3mH coil (0.26 ohms) with an l-pad of 3ohms in series and 6 ohms in parallel. My design uses less baffle step compensation than the ZMV5 justifying the lower coil inductance and the twin 4" drivers are more efficient resulting in the different padding. The impedance of the tweeter and circuit is quite respectable and shouldn't cause any problems. As Zaph did with the ZMV5, I left the top octave alone. And I haven't heard any issues in doing so.
Although the MCM 55-3853 is capable of being crossed higher, I noticed in Zaph's measurements that the 3rd, 4th and 5th order distortions start rising above 2kHz. So, I followed the ZMV5 lead. In the end, the xover point is at 2100 Hz. This also helps with appropriate C2C spacing.
The pair of 55-3853 drivers summed quite nicely into 87db sensitivity. Zaph used a 2nd order low pass in the ZBM4 and a 3rd order in the ZMV5 (the 55-3870 is also a cast frame, metal cone driver that is essentially the big brother to the 55-3853). I tried both and was having some issues with impedance dropping too low. In the end, I used a 3rd order electrical with a Zobel and as you can see above, there is no issue. Minimum impedance is at 3.6 ohms, a classic 4-ohm number. This is a little more "complex" than I would have liked but there is no real fault with it other than parts count. The reverse null is ~40db down. The components for the M's are (using PCD nomenclature):
L2 => 0.75 mH (0.42 ohms)
C2 => 15 uF
L3 => 0.15 mH (0.16 ohms)
C5 => 24 uF (Zobel RC)
R7 => 5.1 ohms (Zobel RC)
For my build, I happened to have a couple of 0.1 uF high quality caps that I used to bypass the 15 uF electrolytic. I don't know if it make a difference in the sound but I had'em at hand, so why not. For the 24 uF, I used 2 parallel 12uF caps. These numbers represent about a -4db baffle step diffraction (-6 to -7db is normal). I did this be my speakers could not be place away from other larger objects. This turned out to be a lucky guess as I wouldn't change anything.
Cost wise, it was $55 for the drivers and $25 for the xover, per speaker. Shipping was free but nothing else was on sale. So this is an inexpensive speaker.
The enclosure is 7.5 liters net volume vented with a 2.5"x1" slot port that's 6.75" long. I lined the walls and added a little Acousta-Stuff in the corners and behind the tweeter. The baffle is 13"x6". All drivers are off-center of the vertical axis. The port is on the front, next to the tweeter. The 55-3853's are above and below. This may be difficult to envision but pictures will be posted soon. The cool thing is that the other 5 pieces each have an 11.5" dimension, meaning one cut and easy fitting: #3 11.5"x6" and #2 11.5"x10.75". And as 0.75" of the port is in the front face, the port can be cut to 6". This is a simple, table-saw speaker. Since I had my baffles machined by Nik, it was a very simple build.
I will state now, and this will be obvious in the pics, that I misquoted the outer dimension of the DQ25. It should be 2.56" (and not 2.625"). If you order the baffles from him, and I will say the pincushion cutout was fantastic, make sure he's incorporated the change.
So, how do they sound? As Zaph said, smooth and clean. They're very articulate with the DQ25 and the MCM drivers are clear. Every instrument sounds like itself with secondary reverberations sounding quite natural. There doesn't seem to be any coloration from the speaker itself. The bass extension is very good and has been improving as the drivers break-in. I was able to greatly decrease the xover point and level of my sub (more so an indication of the poor quality of the other speakers). Integration between the mains and the sub is superb (although the sub is getting replaced in a couple of months). I listened with my AVR set to both large and small. The upper midrange did suffer a bit when set to "large" but with more extension and still very listenable. If I didn't have a sub, I'd use them full range. BTW, Fb is ~57Hz with an F3=~48Hz. Xmax occurs with ~20W.
The MTM is not obviously loud. I was listening to some music and wandered into another room to fetch something and realized the output was much greater than my other speakers. But sitting in front of them, at that level, there was no strain from the speakers, or fatigue in my ears. By the same token, when turned down, there was no change in the frequency response. Everything came across quite even. Nothing extra in the upper end and no decrease in the midbass. It also seems to have a decent off-axis response (determined by ear). Vertical lobing is there but wasn't noticeable (i.e. listening for drops in level) until standing within 1.5 meters (closer than the listening positions in my room).
The MTM does have the "air" of some high-end speakers I've heard but that's not a fault, just a reflection of what went into it. This is a great sounding speaker built from inexpensive components.
This isn't a monitor. Inside of a meter, the multiple drivers becomes obvious. This is not for a large room. My room is 11x14 and it's perfect. IMO, it would work well in a room up to 15x20 but I think it would "get lost" in anything bigger.
Note that this was designed entirely with simulation. I don't have the capability to measure, either the drivers or the speakers. And considering how happy I am with the results, I know I was lucky. Could this speaker be improved? I think there are a few tweaks that would improve the "weight" of the midrange and it may be better without the Zobel if the response could be aligned without it. But for now, I'm just going to enjoy them for a while. I have a sub to design and build, and then I have some other speakers to look at (my computer speakers seem too dynamic now.)
So, there you have most of the story. Pics and some graphics will be forthcoming. Feel free to ask any questions.
Comment