Hi All,
I've got a couple of post up throughout the forum as i'm working on a few different projects so keeping things seperate for clarity.
I'm going to be building a set of 3 way active Sb acoustic floor standers with beryllium tweeter, 6.5" textreme mid and 12" bass.
I'm now looking at building a center speaker to compliment these for movies.
From my reading I know that MTM isn't a great setup due to lobing issues. So the ideal setup seems to be
WTMW with the tweeter on top of the mid.
As for drivers, i'm thinking the TW29RN-B/MR13-P/MW19P (Dual).
This speaker will be passive.
I haven't looked at crossovers or anything yet, i'm just trying to work out what drivers so I can get them all ordered.
Any thoughts on the combo?
There doesn't seem to be too much info on the forum about center speakers, so any advice would be great.
Thanks heaps
I've got a couple of post up throughout the forum as i'm working on a few different projects so keeping things seperate for clarity.
I'm going to be building a set of 3 way active Sb acoustic floor standers with beryllium tweeter, 6.5" textreme mid and 12" bass.
I'm now looking at building a center speaker to compliment these for movies.
From my reading I know that MTM isn't a great setup due to lobing issues. So the ideal setup seems to be
WTMW with the tweeter on top of the mid.
As for drivers, i'm thinking the TW29RN-B/MR13-P/MW19P (Dual).
This speaker will be passive.
I haven't looked at crossovers or anything yet, i'm just trying to work out what drivers so I can get them all ordered.
Any thoughts on the combo?
There doesn't seem to be too much info on the forum about center speakers, so any advice would be great.
Thanks heaps
I had my first proper experiment using VituixCad.
I traced the plots from the driver brochures.
I've used all 8 ohm drivers, then the 2 x MW19P wired in parallel to give 4 ohms.
Found i had to invert the mid to make the graphs look right.
I added an Lpad to the tweeter to bring it down a few db.
Not 100% sure on working out crossover points correctly, but i went with 300hz & 2khz. 2nd order Linkwitz as it looked like a good midpoint on the graphs where it was relatively smooth either side.
How did I do?
A couple of questions...
Thanks all.
I traced the plots from the driver brochures.
I've used all 8 ohm drivers, then the 2 x MW19P wired in parallel to give 4 ohms.
Found i had to invert the mid to make the graphs look right.
I added an Lpad to the tweeter to bring it down a few db.
Not 100% sure on working out crossover points correctly, but i went with 300hz & 2khz. 2nd order Linkwitz as it looked like a good midpoint on the graphs where it was relatively smooth either side.
How did I do?
A couple of questions...
- I've got the lowpass on the woofers at 300hz. Those drivers have a fair bump at about 38hz. Am i suppose to do something to tame that down? (notch?)
- I think I will plan on having the speaker sealed, as it will be mounted pretty close up against a wall, so i dont really want it ported back into it.
- When i look at the total frequency response, it is a bit up and down, but it's only within a few db. What the accepted norm? Eg. I often see a speaker advertised as for example +-3db 50hz-20k. Does that sort of mean you will always have the lumps and bumps when its not a smoothed response, +-3bd is fairly normal?
- Also one weird thing i did discover with inverting drivers, there is a very different phase response when I have the mid inverted, vs having the tweet and woofer both inverted. Only discovered this as I don't really know what im doing lol. Im assuming you invert only one driver, not the other two.
Thanks all.
Attachments
Jim,
The drivers should be as close as possible.
If you want an MTM like this.
LF-MF XO art or below the centre to centre of the 2 woofers.
dave
The drivers should be as close as possible.
If you want an MTM like this.
LF-MF XO art or below the centre to centre of the 2 woofers.
dave
Lots to consider……even question If you actually need the center. The answers lie in the original surround specs for commercial installations redone for the home environment.
Left and Right mains with wide horizontal dispersion and close to the display designed with on axis listening will not need a center and my own personal listening experience for decades has proven the system actually sounds better without one if the mains are truly well designed.
That being said, our summer cottage just had a pair of Magnepan LRS added to a 60” LCD and those are NOT suitable mains for the task. I approached the issue with practicality in mind and simply added a single KEF Q150 speaker under the display tilted up slightly. Impeccable dialogue intelligently due to the nature of the concentric driver. Because these are Maggie’s with poor low end response, 2 powered subs were of course added. The arrangement is close to crossing the KEF at 100hz works extremely well. The arrangement is this Left-sub-Center-Sub-Right
I first tried the vertical TM with a simple bookshelf as i didn’t need the bass response…..the sub placement makes this possible, even preferable. The issue with the TM bookshelf/section was the angle needed for the midrange driver to cover the vertical dispersion……..it simply couldn’t do that as well as I thought it should as it needs to be on the horizontal listening axis.…….the lobe between the TM drivers was obvious………..if slumped down in the couch, the dialogue intelligibility collapsed…..same thing if I stood up( not ideal or recommended for viewing but hey….worth mentioning) But I noticed the issue was the same on the horizontal 15 degrees left and right of center. At that point, the left and right content dominated and the dialogue intelligibility suffered again. The concentric KEF doesn’t suffer from any of this. But what really turned out to be the main advantage was the specific voicing the KEF has……a few db suckout in the 500hz range to eliminate the boxy sound found in male voices and another at 2k which smoothes out female voices. From there, the KEF response tilts up slightly to give you the 6.3khz range where Ssss and Ttttts are so intelligibility is highlighted and 8k which gives you the vertical location we humans need……..it’s right below the screen so content is locked to the viewing angle out to 30 degrees each way……it’s kinda eerie that it worked out that way…..and the best $200 I’ve spent on a bandaid solution ever.
My observational conclusion from the above experiment is that for domestic home theater?…..the coax/coincident driver is by far the best center channel solution.
Designing a passive 3 way speaker is no small effort…..not saying you’re not up to the task at all…..just emphasizing that the reward may not justify the effort and cost. If you’re going to be using/adding subs to your HT, placement options will assist.
Either way, good luck with your build…….the SB active mains sound like a fun project and I bet those Textreme drivers will be incredible!
Left and Right mains with wide horizontal dispersion and close to the display designed with on axis listening will not need a center and my own personal listening experience for decades has proven the system actually sounds better without one if the mains are truly well designed.
That being said, our summer cottage just had a pair of Magnepan LRS added to a 60” LCD and those are NOT suitable mains for the task. I approached the issue with practicality in mind and simply added a single KEF Q150 speaker under the display tilted up slightly. Impeccable dialogue intelligently due to the nature of the concentric driver. Because these are Maggie’s with poor low end response, 2 powered subs were of course added. The arrangement is close to crossing the KEF at 100hz works extremely well. The arrangement is this Left-sub-Center-Sub-Right
I first tried the vertical TM with a simple bookshelf as i didn’t need the bass response…..the sub placement makes this possible, even preferable. The issue with the TM bookshelf/section was the angle needed for the midrange driver to cover the vertical dispersion……..it simply couldn’t do that as well as I thought it should as it needs to be on the horizontal listening axis.…….the lobe between the TM drivers was obvious………..if slumped down in the couch, the dialogue intelligibility collapsed…..same thing if I stood up( not ideal or recommended for viewing but hey….worth mentioning) But I noticed the issue was the same on the horizontal 15 degrees left and right of center. At that point, the left and right content dominated and the dialogue intelligibility suffered again. The concentric KEF doesn’t suffer from any of this. But what really turned out to be the main advantage was the specific voicing the KEF has……a few db suckout in the 500hz range to eliminate the boxy sound found in male voices and another at 2k which smoothes out female voices. From there, the KEF response tilts up slightly to give you the 6.3khz range where Ssss and Ttttts are so intelligibility is highlighted and 8k which gives you the vertical location we humans need……..it’s right below the screen so content is locked to the viewing angle out to 30 degrees each way……it’s kinda eerie that it worked out that way…..and the best $200 I’ve spent on a bandaid solution ever.
My observational conclusion from the above experiment is that for domestic home theater?…..the coax/coincident driver is by far the best center channel solution.
Designing a passive 3 way speaker is no small effort…..not saying you’re not up to the task at all…..just emphasizing that the reward may not justify the effort and cost. If you’re going to be using/adding subs to your HT, placement options will assist.
Either way, good luck with your build…….the SB active mains sound like a fun project and I bet those Textreme drivers will be incredible!
I think before designing, you should consider the position of center speaker first, because most of the time, your ear level is on the same height as position of the TV, not center speaker. The height position of the center, and its relative position to your ear level is important factor, as it will indicate which off-axis you should make the best optimization for. In my opinion, the most benefit of diy is that you can cater the design to fit your need (or your WAG), similar to the bespoke clothing.
After choosing this design specification, the design process can begin. You can have two options. 1st option is Jim & Dave suggest (WMTW), 2nd option is WCW with C is 6.5" coaxial driver. I think the second option is also worth checking. Firstly, it avoid lobing if crossover at 200-300 Hz. Secondly, since center speaker have tendency to be lower than ear level, it avoid one of big problem with coax driver, that is some irregularity in frequency response on-axis.
After choosing this design specification, the design process can begin. You can have two options. 1st option is Jim & Dave suggest (WMTW), 2nd option is WCW with C is 6.5" coaxial driver. I think the second option is also worth checking. Firstly, it avoid lobing if crossover at 200-300 Hz. Secondly, since center speaker have tendency to be lower than ear level, it avoid one of big problem with coax driver, that is some irregularity in frequency response on-axis.
My experience has shown me two things in this regard
1......the horizontal WW with either the MT or Coax is better......but still exhibits some issues in 200-400hz range
2......a 6.5 coax isn't nearly as effective with phase coherency as the 5.25" class coax.....at least the driver's from KEF.
Quite frankly, the 6.5" UniQ driver is junk and unlistenable
I did not have a center for a long time. I picked up a typical WTW center a few years ago.
I fight with that center all the time. It sounds boxy / closed in. Just replaced the tweeter with a nicer Seas and that helped, redid the crossover to blend it better.
For sports and regular TV, I prefer to run in Direct mode (main speakers only). For movies with lots of vocals, especially in dynamic movies with lots of action, the center helps maintain clarity. But it feels lize a laserbeam of sound, exactly the opposite of my mains.
If I were to build a center, I would pick a tweeter with really good off axis performance like an AMT tweeter. And a midrange with similar good off axis.
The issue isn’t the tweeter, it’s the comb filtering that happens between the two side by side midwoofers
Thanks for your opinion.
1. Yes, compared to a one concentric driver, then horizontal lobing of coax with WW is inferior, but the frequency range of lobing effect is not too large if appropriate crossover frequency is chosen. If woofer is 6.5" driver (SEAS L18) and all driver are kept closely, in crossover frequency c-t-c distance will be 34.5 cm, which is equal to lambda/4 of 200 Hz. So even though the lobing still occurs effectively somewhat to 400 Hz, it is not too much problem here. In addition with WCW format, at least we can force a symmetric lobing. And non-linear distortion will be lower than one coax driver only.
2. Yes again, 6.5" coax is inferior to 5.25" coax in terms of coherency, but 6.5 coax can allow 200 Hz crossover, to reduce the lobing. After all, speaker design is always trade-off games. And the Sica 6.5 coax is not too bad compared to Sica 5.5 coax.
If output is not problem then WCW with 5.25" coax, 5" woofer with crossover at 300Hz is also fine. But the distortion trade off is still there
Depends……on the center to center spacing of the midwoofers and the crossover point to the tweeter.
For the sake of arguement, let’s say 5” midwoofers…..and a tweeter in between requiring 2” of space located as close as possible….thats a center to center distance of 7”…..or a full wavelength at 1800hz. At 1800 hz, comb filtering would be severe and everything at and above will be highly directive……listening off axis by anything more than 12” will suffer greatly. Keep in mind this is best case scenario where you could find and use a tweeter that can cross at 1.8 and fits in a 2” space.
But the problem doesn’t start at the full wavelength distance…..it starts at the 1/4 wave distance and gets progressively worse.…..in this case 500hz or so…..right where all the fundamentals of speech lie.
By design, this alignment was developed and proposed for vertical use PRECISELY because of its directivity pattern….in effect it greatly reduces early reflections from both the ceiling and the floor……highly desirable for a stereo pair in an untreated space…….But turn it on it’s side and it’s an unmitigated DISASTER for horizontal directivity unless for a single viewer positioned directly in line with the array.
A point source coax suffers from none of this for the obvious reasons……there isn’t two drive units separated by distance playing the same passband AND for the mid/tweeter response range that DOES overlapp in the crossover region…..near zero phase cancellation due to the concentric placement AND it’s essentially the same horizontally and vertically……this means that content directivity will be pretty much emanating from the screen display area…..the most effective use case possible for a center channel speaker with a solid surface screen where the sound cannot be produced from behind.
I made the argument earlier that for even the Coax driver in an MTM configuration…..ANY woofer to the left or right will STILL result in a combing effect when placed close to the coax…..it’s the design nature of the beast. But at these lower frequencies, we get some relief from distances as to where the woofers CAN BE placed. If we examine the other limitation of center to center spacing at say 350hz ( to where a coax could effectively play quite loudly without distortion products in an average home environment)……now we’re talking a distance on near 40”…..and a 1/4 wavelength distance of 10” or so…..perfect for a horizontal alignment of 3 5” drivers. Such a speaker would likely handle nearly all the current you could throw at it in a domestic multi use Audio/Video application.
My FINAL argument would be for a pretty typical modern AV multi use scenario where we would have a 60” display and an overall stereo sound field of say 80” inches. In this case, assume the simple KEF single Q150 positioned centrally and crossed to subwoofers at 100hz. Now we have a 1/4 wavelength distance of around 40”……….and if we re examine our overall soundfield of 80”…….well…..the near perfect scenario snaps into focus with instead of a purpose built center channel with woofers……we’ve got two independant sub-woofer applications serving two purposes MUCH more effectively. The only downside here is the power handling of the little Coax KEF suffers. To a degree of reason though, KEF offers the Q150 with port plugs, which in effect reduce the low frequency extension to 80hz while increasing power handling greatly by supporting the woofer movement through reflex/box compliance.
For the sake of arguement, let’s say 5” midwoofers…..and a tweeter in between requiring 2” of space located as close as possible….thats a center to center distance of 7”…..or a full wavelength at 1800hz. At 1800 hz, comb filtering would be severe and everything at and above will be highly directive……listening off axis by anything more than 12” will suffer greatly. Keep in mind this is best case scenario where you could find and use a tweeter that can cross at 1.8 and fits in a 2” space.
But the problem doesn’t start at the full wavelength distance…..it starts at the 1/4 wave distance and gets progressively worse.…..in this case 500hz or so…..right where all the fundamentals of speech lie.
By design, this alignment was developed and proposed for vertical use PRECISELY because of its directivity pattern….in effect it greatly reduces early reflections from both the ceiling and the floor……highly desirable for a stereo pair in an untreated space…….But turn it on it’s side and it’s an unmitigated DISASTER for horizontal directivity unless for a single viewer positioned directly in line with the array.
A point source coax suffers from none of this for the obvious reasons……there isn’t two drive units separated by distance playing the same passband AND for the mid/tweeter response range that DOES overlapp in the crossover region…..near zero phase cancellation due to the concentric placement AND it’s essentially the same horizontally and vertically……this means that content directivity will be pretty much emanating from the screen display area…..the most effective use case possible for a center channel speaker with a solid surface screen where the sound cannot be produced from behind.
I made the argument earlier that for even the Coax driver in an MTM configuration…..ANY woofer to the left or right will STILL result in a combing effect when placed close to the coax…..it’s the design nature of the beast. But at these lower frequencies, we get some relief from distances as to where the woofers CAN BE placed. If we examine the other limitation of center to center spacing at say 350hz ( to where a coax could effectively play quite loudly without distortion products in an average home environment)……now we’re talking a distance on near 40”…..and a 1/4 wavelength distance of 10” or so…..perfect for a horizontal alignment of 3 5” drivers. Such a speaker would likely handle nearly all the current you could throw at it in a domestic multi use Audio/Video application.
My FINAL argument would be for a pretty typical modern AV multi use scenario where we would have a 60” display and an overall stereo sound field of say 80” inches. In this case, assume the simple KEF single Q150 positioned centrally and crossed to subwoofers at 100hz. Now we have a 1/4 wavelength distance of around 40”……….and if we re examine our overall soundfield of 80”…….well…..the near perfect scenario snaps into focus with instead of a purpose built center channel with woofers……we’ve got two independant sub-woofer applications serving two purposes MUCH more effectively. The only downside here is the power handling of the little Coax KEF suffers. To a degree of reason though, KEF offers the Q150 with port plugs, which in effect reduce the low frequency extension to 80hz while increasing power handling greatly by supporting the woofer movement through reflex/box compliance.
Last edited:
@mayhem13 : I think you are wrong here. A 5" midwoofer have diameter of 5" so the radius is only 2.5". So if a tweeter have total diameter of 2" (like SB26SCTN with some cut-out) then radius is 1". So the total c-t-c distance between midwoofer and tweeter is only 2.5" + 1" = 3.5" = 8.9 cm, not 7" like you said.
c-t-c is center-to-center distance, which calculated between the center points of drivers, not farthest-point distance. if 3.5" is quarter wavelength to avoid lobing at crossover frequency, then this frequency is 34300 cm/(4 * 8.9 cm) = 963 Hz, not 500 Hz
c-t-c is center-to-center distance, which calculated between the center points of drivers, not farthest-point distance. if 3.5" is quarter wavelength to avoid lobing at crossover frequency, then this frequency is 34300 cm/(4 * 8.9 cm) = 963 Hz, not 500 Hz
But you have two woofers and it's their distance that creates severe lobing.
If your midrange is 5" and two woofers are 6.5", then if they are close enough, the highest c-t-c distance will be around 6.5"/2 + 5" + 6.5"/2 = 11.5"= 29.2 cm. If you don't want lobing at crossover frequency, then 11.5" must smaller than quarter wavelength, so crossover frequency must be smaller than 34300 cm/s /(4 * 29.2 cm) = 294 Hz. Even then, since crossover is not brick-wall filter, lobing still occur but not as a big problem until 2 * crossover frequency ( if you choose at least 2nd order crossover filter LR2)
Last edited:
I would be wrong in the conntext of a vertical TM section with flanking woofers.....and not really wrong as my reply addressed a single central Coaxial speaker instead of an TM
The TM would be better than the horizontal MTM that's so common to commercial center channels (useless IMO) but not as effective as the Coax option.