different amplifier configurations

[budget minded]

i'm quite familiar with the "every time you double your drivers, you get 3dB more output" rule, but that rule doesn't take frying your amp which might not like 4 ohm loads into account.

does the same rule apply with doubling your amp channels? eg. would adding a second pair of speakers with the same amplifier output as your originals ALSO give you 3dB more output? if you have a pair of speakers that can put out, say 105dB at 100w would you then get 108dB by adding another amp & pair of speakers?

i have a 100wpc panasonic class D receiver that i want to use for small parties with 4 x 93dB satellites each being driven by a channel at 8 ohms. if the rule works the way i think it does, there would be no significant improvement in buying a 300wpc @ 4 ohm stereo amp that can handle the impedance worth a $300+ investment. i can't double my speakers either as that would create the dreaded 4 ohm load i don't want to fry my receiver with.

i'm not interested in class H rail switching pro amps as they weigh a ton and probably don't have the amazing detail of my panasonic which spanked the daylights out of my overrated & muffled NAD. i like EVERYTHING about class D, the light weight, the efficiency & especially the speed & effortless detail with zero treble grain. other than $300/300w pro amps, anything else class d & high power is going to cost way more than i can afford if i want to be up & running before summer is over.

i'm thinking, for right now, just sticking with my panasonic until i can afford a 500wpc+ unit than can do 4 ohms is the best value for my system. sound quality is something i absolutely refuse to compromise in any way. it wasn't easy finding a small/fast driver combo for the satellites that could do 93dB. for what it's worth, i'll also have 4 X 12" sealed subs wired for an 8 ohm mono load driven by a 500w BASH class D subwoofer plate as soon as those become available again. i was going to run the 2 X 6 1/2" & 1" titanium satellites in "small" mode with my subwoofer for a little extra headroom.

any feedback that can either confirm my assumption or clarify it, or add another "free dB trick" is appreciated.

[budget minded]

i forgot to mention, a coverage trick i intended to employ was spacing each speaker in a 20 foot by 20 foot square (outdoors or as space permits) so that only those in the center of the "dance floor" are more than 10 feet away from a speaker. that, and all of the mains will be mounted on posts above everyone's heads so that the treble is totally unobscured. before upgrading amplification over v1.0 (whatever that turns out to be), i'll be getting a behringer digital EQ for tweaking the sound at every given location (minus the effect of a crowded room on acoustics) for optimal sound quality.

i refuse 200% to go with hideous sounding high efficiency PA technology. i hate EVERYTHING about that sound. louder is NOT better! i'd rather listen to whatever without the distortion you get turning it up EVERY TIME. i'm not DJing for money either so i don't need to sell out.

[GM]

Originally Posted by budget minded

does the same rule apply with doubling your amp channels? eg. would adding a second pair of speakers with the same amplifier output as your originals ALSO give you 3dB more output?

sound quality is something i absolutely refuse to compromise in any way.

Summing multiple identical sound sources:

SPLmax = 10*LOG10(dBn)+dB

So from this we see that in theory, two channel pairs adds ~3 dB, three pairs = ~4.77 dB, four pairs = ~6 dB, etc..

SQ of such arrays degrade with increasing number of discrete systems until it has enough 'n' sources to overload our internal processor, forcing it to auto sum this assault on our senses into tricking us into perceiving it as a coherent point source. For smaller arrays then, it's necessary to get far enough away for it to seem as one.

For a spaced array, I recommend you study up on distributed sound system design to determine the number of discrete systems and directivity required for each app.

GM

[gedlee]

GM

Not quite right. If you are talking about SPL then you have to use 20 Log, not 10. What you show is for power not SPL. So for SPL you would need to double your numbers.

[GM]

Greets!

It's my understanding that SPL is 20 LOG only when dealing with quantities that are not power ratios and since we want to sum dBs of powered systems where some percentage of it is turned into acoustic power, 10 LOG is the correct multiplier. In short, doubling acoustic power (Lp) in theory = +3.01 dB.

GM

[gedlee]

Gm

If you are talking about "Acoustic Power" then 3 dB is correct, but SPL is not a power unit - its like voltage, or current. So to get power you would need to square it, or taking the square outside of the 10 Log10( p^2) makes it 20 log10( p ) = SPL.

[budget minded]

when i say SPL, i mean the same thing as dB.

i didn't get any of the mathspeak, but gleaned that doubling the number of amps & speakers has the same effect as doubling the speakers on one amp. i was kind of hoping that the doubling of power driving the speakers would give a 6dB increase. i'll be wasting power for the same output, but not putting a 4 ohm strain on my receiver at least going in 4 channel mode.

i'm not as enthused as i once was about 300wpc technical pro & vocopro "professional" amps as after a couple years, i still can't find any reviews for them. i think when i upgrade my mains, i'll go with a dual mono connexelectronic DIY rig as i've seen them get respect for sound quality & engineering. i actually wanted to build a sub amp with one of their units, but can't find any info on building line level lowpass filters anywhere. i guess there aren't a lot of people DIYing subwoofer amps.

i'm surprised that a good 5 years after class D consumer gear, it's still hard to find ANY kind of class D amps, but tons of technical paper spam.

[gootee]

Line-level lowpass filters are probably not a hot topic because they are an extremely-well-known and relatively easy case. But there is a staggering amount of information available about them, on line. For example, there are two good chapters about filters in general in this excellent free book from Analog Devices: ADI - Analog Dialogue | Op Amp Applications Handbook . And there are plenty of Electrical Engineering sites with more than you might want to know about them. (It'll give you a good reason to appreciate learning a tiny bit of math.) And you can download TI.com's free FilterPro software and have at it, if you are talking active filters. Or you could even try simple passive 1st-order analog low-pass filters and just plop unity-gain buffers after each stage. Crude but actually quite effective. You might want to download LTspice (free), from linear.com (hint, hint).

Do you know what you need, in terms of the specs of a low-pass filter? There are many here who could probably almost-instantly give you a circuit if you could define exactly what you want it to do, and in what system context.

[Pano]

gootee, did you post in the wrong place? or am I missing something?

[gootee]

Originally Posted by panomaniac

gootee, did you post in the wrong place? or am I missing something?

"budget minded" wrote:

i actually wanted to build a sub amp with one of their units, but can't find any info on building line level lowpass filters anywhere. i guess there aren't a lot of people DIYing subwoofer amps.