Center Channel Design Questions

[weinstro]

Re: Center Channel Design Questions

Thanks - that's helpful.

This center channel will be driven by an Onkyo NR737 AVR. The HP cross frequency internal to the AVR is adjustable, but I don't think the slope is. Are AVR HP slopes typically LR2 or something else? I can't test this AVR - it's 2,000 miles away, but I'd like to model the effect of the HP crossover via Response Modeler.

Thanks and regards,

Rob

[jhollander]

Re: Center Channel Design Questions

I don't know what most AVRs are. However if you add a 1st order hi pass in RM the resulting driver plus hi pass slope is in the second to third order range. I find it easier to look at slopes in PCD.

[make_some_noise]

Re: Center Channel Design Questions

You could use some of these:

http://www.parts-express.com/fountek...8-ohm--299-022

one of the ND20 Dayton tweeters.

and maybe add a couple stout 4" drivers to pick up the bass section.... DA 115's would look good along with the founteks. Maybe a wmtmw.

Or if you can set the center to small and route the bass to the sub, just use an array of the founteks with the little dayton tweeter.

Something else to think about is that... Since the founteks are so small and so wide in bandwidth.... You could essentially make the box shorter and put feet on it. Cross them over around 150hz and put a down firing 4-5" bigger driver(s) in the cabinet to help the very bottom end. Since the woofers would be down firing and essentially .5's you could use anything that will dig deep enough no matter what they look like.

[weinstro]

Re: Center Channel Design Questions

I wonder if my approach for "matching"'is flawed. The LR main channels will run out of steam (xmax) at roughly 103 dB w/ 15 watts. I'm trying to select drivers to match that capability although there is no real requirement to play that loud. For a quad of woofers, this is relatively easy. For a 3-4" full range driver, I'm having difficulty finding a driver that can play at this level and have decent 30 degree dispersion >10 kHz. By my calcs, I need 88 dB (@1 w) sensitivity or better.

So, assuming that I'm making things way more complicated than necessary, I'm curious to learn what you folks do to match drivers, within a speaker, and across channels?

The driver below is nice, but it won't play at the level I'm trying to match:

http://www.parts-express.com/fountek...8-ohm--299-022

I did manage to find one driver that met all goals - the TB W4-1805S. But, my paper tiger has taken an unexpected change in direction as that driver went NLA yesterday. Ugh! I'm sure that's happened to others before.

Thanks and regards,

Rob

[jhollander]

Re: Center Channel Design Questions

You may want to consider a chambered tweeter crossed low and accept the dip at the x-o point.

[Chris Roemer]

Re: Center Channel Design Questions

ND105-4s will outplay DA115s by +5dB, +10dB at and below 50Hz.

[weinstro]

Re: Center Channel Design Questions

ND105-4s will outplay DA115s by +5dB, +10dB at and below 50Hz.

Thanks for the suggestion, but it's also too big - I'm constrained to 4.5" baffle height (114.3mm). The DA115 is 115.6mm wide.

[weinstro]

Lining and Stuffing-

I'm in the process of completing the cabinet, and the height has come in at 4-3/4", a bit over 5" deep, and 47" wide. As a reminder, this is an MM(FR)MM center channel using ND91 woofers and a Tang Band W4-1757SB for tweeter/mid duties. The woofers are slot ported while the TB driver is in a separate sealed chamber.

I'm thinking I should line the inside of the cabinet with something, but I'm wondering what folks use when the cabinets are very small spaces? There isn't a whole lot of space left over for 1" thick material, even with ND91 woofers. The cabinet is amply braced so I'm not all that worried about panel resonance, but the length of the cabinet and port is reminiscent of a transmission line where the driver offset is around 40-50%. So, there could be acoustic resonances to worry about.

Similarly, the W4-1757SB will reside in a 1-2 L sealed box. Should I bother to throw some stuffing in there?

Thanks and regards,

Rob

[jhollander]

I usually stuff all my enclosures as I worry about reflected sound inside the cabinet coming through the cone. For vented I keep the path to the port free then stuff to catch the rear wave. May be only an inch or so of batting stapled on top of Roxul or fiberglass at that short depth. Sealed I like layers of different densities finishing with poly batting near the magnet if the pole is vented. If you did make a MLTL might be difficult to eliminate those resonances.

[weinstro]

Out of curiosity, I modeled the as-built box in HornResp, Unibox, and Transmission Line (Leonard Audio). The schematic for the HornResp model didn't really reflect what I built and I was unable to configure the model for a better fit, so I set it aside.

Unibox predicts a small resonance at around 730 Hz (shown below) due to the port, but this package doesn't account for acoustic wave effects.

Transmission Line predicts a more aggressive scenario. What's neat about this package is it will model stuffing densities. Most of the resonances can be damped out, but not the resonance associated with the port at 750 Hz or so. This shouldn't be a huge surprise as I don't intend to stuff the slot port.

I'm hoping to cross from the ND91's to the fullrange driver at 800 Hz or so, but that remains to be seen (and heard). So, what do you think?

[jhollander]

Sounds like a good plan to me. Are you going to measure or do simulations?

[weinstro]

Both, although the cabinet is getting to the point where measuring will be upon me soon. Another learning curve...

I keep finding myself in scenarios where the common baffle diffraction and room boundary simulators run into some limitations. So, measurement capability is pretty much a requirement.

[weinstro]

So, I managed to tear out about 25% of the plywood finished face when routing the ends square.

Where do I register for my free T-shirt?

Box is otherwise ~90% complete.

[jhollander]

Ouch

[weinstro]

Relative Acoustic offsets....

For my design work to be sufficient, I need to determine relative acoustic offset between the ND91s and the W4-1757SB.

My question is this: for an MMTMM design, should I tie all woofers together and test as if this were a single average woofer?