Building a Dinner Theater w/ DIY Speakers, Subs, etc... - Page 5

Been through this before. It is not worth the risk. The difference between 300 and 500 watts per driver will not likely be measurable in real world conditions. The SPL increase at best will only be 2.22 dB, but some of that will be lost due to compression.

(log(500) - log(300))X10 = theoretical increase in dB

(2.699-2.477) * 10 = 2.22

The actual increase WILL BE LESS than this. The risk to your equipment rises quickly as you approach (or exceed) the thermal and mechanical limits. You're applying 67% more power to get maybe a couple of dB under ideal circumstances. This is crazy. The risks outweigh the benefits in my opinion. Applying power that does not make more SPL is a waste - all it does is take the equipment that much closer to the point where damage results. As soon as watts in don't equal SPL out, there is no need to put more watts in.



With the MFW 15, I hit that limit at about 300 watts when I measured it, but I ran out of amp, so I am not 100% sure that I hit the driver limit or the amp limit first. Those sweeps should be separated by 5 dB each, as you can see, the top one is not.



The T-6 with the W-8 740C shows this a little better. I did the plot a little differently too, I overlaid all of the plots with the power offset taken out. By 64 watts (1/2 the driver's rated power handling) I was showing signs of reaching the driver's limit. By 128 watts, the effects are clear. At 256 watts, I heard lots of bad noises... I did not bother including that plot. This time I did not run out of amp, I clearly ran out of driver.

Added power will result in thermal rise to the voice coil, increasing Re, which results in a decrease in the efficiency of the driver (power compression). With the additional power, you also increase excursion by about 40%. With an xmax of 14 mm, 500 watts exceeds this limit (15 mm). More importantly - with an xmech of ~20 mm, 500 watts is well past that limit (32 mm).

300 watts should not damage drivers, and with 4 cabinets fed 300 watts, you're approaching 130 dB at a meter, in a groundplane setting.

I'd certainly hope that's enough.

If the baffle wall is all absorptive material, you will lose some boundary loading for the 4Pis, it will be as if they were more like in free space. Enough to matter? Hard to say. mouting the speakers close to the screen material will help if you use reflective material for the baffle wall, or use a thin absorption layer to eliminate HF reflection problems but still retain some boundary loading lower in freq. I say use drywall, listen, hang the screen, listen. If there is a problem, use 1" of fiberglass in front of the drywall, rehang the screen, listen. If the problem is solved, done. If not, another 1" of glass. 2" of 3lb per cuft fiberglass will absorb any freqs reflected back by the screen and then some.

L and R speakers should be around 24-28 degrees off axis.

I think of multiple subs cancelling room modes either by not exciting them due to position (even order modes), or by cancellation of odd order modes. Either way, you get the same result. Wayne is right, spreading out the subs will get you the largest area of similar (NOT FLAT) response (with large variations in response outside of that area), with one EQ solution (which may require more headroom than you have). What Parham fails to address is that multiple, spread out subs lose efficiency compared to clustered subs, depending on the chosen sub locations, and IIRC you do not have a whole lot of headroom to begin with (with 4 F20s) if you plan to run at reference level in the middle of the theater space.

The ONLY way to see the difference in your room will be to measure and experiment with sub locations, which will take time, and which may not be a part of the overall decor you are going for. I do realize that you have limited placement options, hence my recommendation to cluster the subs under the screen. Sure, in an ideal world, you would have subs with 10dB of headroom above reference anywhere in the seating area, spread out around the room to cancel modes, and even running a JBL BassQ to optimize the system fully.

Some general reading that may help w/ LCR placement:

http://www.jblpro.com/pub/cinema/cinedsgn.pdf


JSS

One amp for multiple surrounds may be better since it might be hard to drive more than one amp from the pre-out on the AVR.

I think the Onkyo PR-SC5508 may be available in a pro version with rack mount ears. Way over your budget though. It also has XLR outputs, that you won't find on a receiver.

dutch,

For a quick angle measurement, a fist at arm's length (knuckles horizontal, not including thumb) is approximately 10 degrees of angle. From the middle of the theater, if you have 2 fists between center and L or R, you should be good. If it is closer, consider mounting the L/R just outside of the screen...

Looks like a pretty good plan, thanks for the 1st post updates! Looking forward to seeing this completed....

JSS

Wiring in parallel may help you reduce the amount of amps you will need. If the amp is stable to 2 Ohms and can provide the needed power, then do it.

JSS

Wire two in series, which will give you 16Ω. Then wire those pairs together in parallel, which will get you back to 8Ω.

Golly this takes me back to my car audio days. I used to be decent at this but now I can't remember much other than loud bass, what does that mean? Can't be good

That's exactly what I had in mind. All of your diagrams look correct, however, I'd wire up the surrounds to load the amp at 8Ω instead of 2Ω. Look at the specs for the amp - distortion rises from .1% to 1% which I would think means the amp is really straining to hit the mark. Also keep in mind that the speaker impedance is not a fixed value. It varies with frequency. Look at the graph below from Wayne's site. The impedance of the 2π drops to 4Ω between about 6kHz and 13kHz. That will put a 1Ω load on the amp with four wired in parallel.

I like the exterior. Nice choice on color. Cables look like they are run well and secured nicely. The only thing drawback to those staples is that if you ever want to pull new wires through that it'll be tough but that might not be a consideration. Nice work.

Crown

In the case of pro gear, most value assessments come from criteria that has little to do with subjective sound quality and more with things like reliability, road worthiness, portability (read weight),etc. As to your choice of amplifiers, I suggest you narrow down your AVR/ preamp processor to determine it's output interface and level. You're going to want avoid having to use output level matching devices between the two or suffer limited dynamic range or output. Amplifiers that have selectable input gain stages are obviously more flexible.

In the case of your need for sound deadening in or behind the baffle wall, Roxul is the preferred material.

+1 to the level-matching. I have some Behringer stuff, but none of the above models.

Are you planning on using an AutoEQ package onboard an AVR for final adjustment/tweaking of the system?

JSS

Since you'll be using discrete amps would it make more sense to use a preamp rather than an AVR since you won't need the amps in the AVR?