What makes a good center channel? (1 Viewer)

[Jeff_Blanch]

I've been thinking about some options for building a complete set of HT speakers, and have been a little confused with respects to the design of the center channel. I was looking around and I found this link that tells us that MTMs are bad for centers (of course they push their own design). Is there any merit to what they are saying?
What is the worst possible way to design a center channel speaker - a way that practically ensures hard-to-understand vocals in a speaker whose main job is to render vocals intelligibly? Well, you could design a speaker like most of today’s commercial center channel speakers, with two woofers located on opposite sides of the enclosure and a tweeter centered between them.
Yet when I look at a majority of center speaker designs, they are exactly that. Essentially an MTM on its side. I know there might be issues with phase cancellations. How far apart or close together do two woofers need to be to minimize any adverse affects from putting them side by side?
To give you guys an idea of what I was thinking, I am currently planning on making three identical MTM speakers, one each for L, C, R (all shielded). The other 4 would be TM arrangements with the same drivers. What would be the drawbacks to something like this? My listening would probably be 90/10 movies/music.
thanks
EDIT: OK I figured out the frequency/distance calculation to determine at what frequency the two speakers begin to interfere. If interference begins at about 1200hz and the crossover is at 2500hz, how big of an impact is this really going to be? Isn't this also an issue with regular tower MTMs? What is the difference?

 

[Patrick Sun]

The dispersion characteristics of a MTM on its side is not good when used as a center channel. By using a tweeter/midrange in a vertical orientation, the dispersion for the speaker will be better, especially for vocals, which is good since center channel speaker carry a lot of dialog/vocals in DVD soundtracks.

You could always test out the disperion of sideways drivers by flipping your current speakers on their sides, and listening to them and how they sound different as you move around the room.

 

[JayDaniel]

I'm no speaker expert, but I've seen a lot of center channel speakers lately in a M-T-M design, but with the tweeter at the top portion of the enclosure, not just in the middle on the same plane as the midwoofers. My center channel is designed like this, with the tweeter about and inch or 1.5 inches higher than the midwoofers. Perhaps this improves the performance/dispersion/dialogue somewhat. Does anyone know?

 

[Chris Carswell]

Good question. I'm designing my MTM mains & surrounds with the tweeter off center but the center I wasn't sure about at first. The theory with the off center tweeter on "Vertical" speakers is that it helps dispersion (also lowers the CTC distance). The part that hung me up a bit was where they say it's not a good idea to take a "vertical" speaker (your typical book shelf, MTM, or tower) and lay it on its side. The thing is when they are referring to that, I believe they assume the tweeter is "in-line" with the woofers. This is why it wouldn't be a good idea. So I designed my center like my MTM mains with the tweeter off center, about 2" above the woofers.

 

[Jeff_Blanch]

Patrick, thanks for the advice. I figured there was something to it, but wasn't sure. So is dispersion the main problem? Does that just mean that the off-axis response will be poor? I will have to see what turning my speakers does, but I've only got a small two way right now, so I'm not sure if it would really show me a difference.

JayDaniel, I noticed the same thing. I'm not sure what raising the tweeter accomplishes besides possibly bringing the two woofers closer together. I did take this into consideration, but wasn't sure what difference it would make. Anyone care to comment?

 

[Bill Fagal]

The problems with side-by-side drivers operating at frequencies above mid-bass are very real: narrowed dispersion and ragged of-axis response. These factors have strictly to do with the acoustic interactions of multiple transducers operating within the same passband--tweeter placement is a separate issue.

Vertical MTMs work OK because they narrow only vertical dispersion which can even be a plus as it reduces floor and ceiling reflection amplitude. as frequency rises, focused amplitude lobes aimed at the floor and ceiling multiply but do not affect response in the all-important horizonal plane.

Now take that system, rotate it 90 degrees, and call it your new center channel. Now you you have a narrow vertical band of smooth response flanked left and right by lobes interspersed with suckouts. Instead of evenly illuminating a wide horizontal swath of earholes, that window of smooth response is aimed only at those in the center, while wasting the rest of its beam on the floor and ceiling. That's the sad theoretical truth (though, of couse, your mileage may vary).

For these reasons (and a few others), I'm personally sold on coaxial alignments where a single treble driver is centered in a single mid-bass. Gone is the frequency and position-dependant raggedness, and sweet spot real estate multiplies, even through crossover overlap. Ahhh.

Bill

 

[Zac_F]

There is also the Audax center, with the w/tm/w design. This has a tweeter above a 5" midrange flanked by 6.5" woofers. It is said to avoid the lobing problem in the critical frequencies since these are produced by the single midrange. I've built one myself, and I sit off-axis frequently and never have detected a problem.

 

[Ben Ven]

I agree with Bill. I just verified what Bill said. I ran some quick SPLs. You just move over 1 foot and you will get a huge suckout at certain frequencies and then move over another foot and you get huge peaks. Kinda neat. Why not use a full range driver like a Fostex or Jordan?

 

[ChristopherD]

Hey Ben,
Can you run some tests with the center in a vertical orientation if you have time? I'd like to know how great the diffeences are. I will be building soon and have a great deal of interest in this issue. Thanks in advance!
Chris

 

[Greg Monfort]

>Why not use a full range driver like a Fostex or Jordan?
====
It's certainly moving in the right direction IMO, but where you run into trouble is with power handling if you want to get to DD reference, and why I prefer co-axial drivers, such as the Adire HE10, in small to medium rooms and relatively large horns coupled to high efficiency pro sound midbass drivers in larger rooms.

GM

 

[Bill Fagal]

Since I don't have a huge budget, I've been considering the cheapish Selenium and Eminence coaxials Parts Express carries. The Selenium 12CO1P (12") and 15CO1P (15") come with a crossover, but can be bi-amped too. The Eminence Beta CX woofers (8"-15") seem to be a performance bargain. The 8"-12" models are suggested to mate with the APT-50 super tweeter (Basically the Adire HE-10 hardware) with a crossover at 3K. However, I believe you could thread in the superior PSD2002 compression driver and use a lower XO. Jon Risch has some interesting series XO ideas that seem promising with this setup.

As I say, I haven't personally tried these, so I'd be interested to hear your impressions if you have.

Bill

 

[Jeff_Blanch]

Ok, so the problem is very real. Do commercial speaker builders who make these MTM centers just count on it that the people who use them will be sitting right in front of the speaker?

Here is another theoretical question. If the interference between the two speakers occur at the frequency range whose wavelength is smaller than the distance between the two sources, what would happen if I moved the two sources together, so that the distance between them is less than the wavelength of the highest frequency that I would ask it to reproduce (the crossover)? If the distance is less than the wavelength of the crossover frequency, there would be no interference between them, right?

Or I could just throw in a small midbass driver.

 

[Patrick Sun]

Commercial makers of MTM center speakers count on aesthetics to carry them through in the purchasing decision by the general public.

 

[Bill Fagal]

Yes, the majority of speaker manufacturers place a much higher premium on aesthetics than acoustics--and I do not shrink from implicating high-end companies either. Sadly, it's a very successful business model. Our human love of symmetry has led to the axiom: horizontal MTM means center channel, and vice versa. (Ask yourself--why do you want so badly to make an MTM center work? Doesn't it just SEEM right?) He who has ears, let him hear.

If the distance is less than the wavelength of the crossover frequency, there would be no interference between them, right?

A point to consider is that interference between band-sharing drivers is a matter of degrees, it's not a brick wall on/off switch. At wavelengths many times the C-C distance, go ahead and consider them a point source. But as frequency rises, directionality and lobing creep in. Even an octave+ below that critical C-C distance, I believe your sweet spot will be horzontally narrowed to some degree, with lobes and suckouts beginning to say hello.
Also, what's your crossover slope? Most aren't very steep, and none are brick walls (we await the digital domain to bring us those ) So, an octave+ above your crossover point, continually multiplying lobes will still influence your polar response, though from a few dB down.
So, by all means try to shorten the horizontal spacing of band-sharing drivers. Just realize that you're mostly playing with degrees of compromise.
Bill

 

[Jeff_Blanch]

Actually the reason I wanted an MTM for the center is more convenience than anything. I wanted to take an available crossover and use it for the center, so I wouldn't have to design a specific crossover for the center, as I have little experience in this area.

I guess I was being a little optomistic with my theoretical idea there. It does seem correct that the closer you get to the limit, that there would begin to be some interference due to the crossover slopes.

As to adding a midbass driver, would I be able to take an existing MTM crossover, add a midbass like the PE 5.25 woofer with a textbook crossover? Or would that really mess it up. Or could I take a TM design and add dual woofers again using textbook crossovers? I don't have funds for fancy software and measuring equipment. What do you all think?

 

[Ben Ven]

Christopher,
I can run some sweeps with the speaker in the vertical position. My RLC channels are MTM arrangements. All identical. I did have a nice spreadsheet with some nearfield measurements using my Rat Shack SPL meter. But my hard drive went for a
Anyway, I probably won't have time until at least the weekend. If I do it, I will post my crude results.
Anyway, there was one full range driver I did not mention and that is, Babb. I am actually looking at becoming a distributor. These may get to DD reference levels, but I have not heard them yet. I actually ordered a few sets to try out the 514, 534 and 1034. I have not received them yet.
They make a 10" and it may fill medium to large rooms. Power handling is up there, although not out of this world. 100 W RMS and 96 dB efficiency. Not bad. There is some interesting technology that goes into these drivers. Apparently, six patents.

 

[ChristopherD]

Thanks, Ben.
It would be nice to be armed with some raw numbers when doing my decision making, let alone combatting the wife and friends when they try to explain that my CC is set up "wrong".
I'm sure you've got plenty of more important things to do, so if you don't get around to it, no worries. I'll just have to tell them, "It's better, Ben said so."
That should be good enough...

 

[Greg Monfort]

>The Eminence Beta CX woofers (8"-15") seem to be a performance bargain.
====
They are, and the Adire HE10 is an even better one IMO as it's been tweaked for HIFI Vs the made for high SPL Betas.
====
> The 8"-12" models are suggested to mate with the APT-50 super tweeter (Basically the Adire HE-10 hardware) with a crossover at 3K. However, I believe you could thread in the superior PSD2002 compression driver and use a lower XO. Jon Risch has some interesting series XO ideas that seem promising with this setup.
====
At the expense of power handling I presume, but don't have any experience with these or the Seleniums. My background is primarily with Altec/Urei duplex, and various quality whizzer coned FR drivers.

Really though, if only slightly louder than normal TV viewing HT volume is acceptable (i.e. B*** levels), some very good sounding bipole/shielded HT monitors can be built using relatively cheap FR drivers. I recently built an all RadioShack set of five monitors using ten discontinued drivers, five cheap piezo super tweeters, and five discontinued inductors for >$100, capacitors, quality plywood, connectors, and all. Even using currently available drivers they could still be built for ~$250 and get superior imaging/soundstaging compared to a typical two-way since there's no electrical XO/driver overlap in the 500-7kHz critical hearing BW.

I had to take a bit of extra time tweaking the drivers a bit to smooth them out, but considering how much I saved, it was worth it. You can get around this by using pricier Fostex or Jordan drivers, but then you have $150-300 in each monitor. :^(
====
>Commercial makers of MTM center speakers count on aesthetics to carry them through in the purchasing decision by the general public.
====
And/or size. How it performs is at best a secondary consideration.
====
>My center channel is designed like this, with the tweeter about and inch or 1.5 inches higher than the midwoofers. Perhaps this improves the performance/dispersion/dialogue somewhat. Does anyone know?
====
Yes it does. Butt the midbass drivers together and either use a small tweeter or cut off as much flange as possible to get the closest fit between them.

With a 2.2kHz XO, this limits the midbass to 6.25" drivers, but a pair will get you to the needed 80Hz. With some foam to control the tweeter's off axis response to that of the midbass combo, the power response is fine for a typical HT that doesn't need to be quite at DD reference.

A better solution is to make a bipole midbass so that the driver's magnets touch. This solves three problems, excessive gap between the midbass/HF at the XO point, baffle step compensation (with an inductor on the rear driver to roll it off), and the coupled magnets makes them a shielded design.
====
>Anyway, there was one full range driver I did not mention and that is, Babb. I am actually looking at becoming a distributor. These may get to DD reference levels, but I have not heard them yet.
====
Me either, though there are some folks on the fullrange forum that swear by them. Just looking at specs, they have abysmal efficiency and a very high Q, though Babb specs them much higher and has stated that their drivers can't be defined by T/S due to their unique design. Yeah, right, if it has an impedance peak and compliance it can be defined by T/S. Me and some others have hammered Alan some on this point, but without hearing/measuring them in-room we're not sure how much it affects the speaker's response.

Anyway, I guess since I'm a 'doubting Thomas', Alan has singled me out as one of the folks he wants to subjectively review a pair of the new high zoot Lorelei speakers. If he sticks to the timetable I should be able to bring them to the 10/25 Atlanta DIY2002 meet. According to Babb, this driver is the only one you need to make a small 20-20kHz phase coherent speaker, though I don't see how it can approach DD reference in a typical HT.

GM

 

[Chris Tsutsui]

I don't know if anyone mentioned this but you could consider my idea of a variable center channel.
It would be separate enclosures for the midranges and tweeter so they can be moved around and clamped down to preference.
Another idea would be to use 2 separate speakers as a center since they'd both be playing mono it would sound like it's coming from between them and provide better off axis response. You could then space them apart or screw them into eachother to change the width of the stereo or you could stack them, stagger them, or angle them one to the left of the theatre one to the right.
Just some food for thought on designing the "perfect" center.

 

[Jeff Rosz]

howsabout cloning something like this