DS> I just read some information about system PA balancing, in some
 DS> information from Apogee, that I never thought about before, which made
 DS> me realize I probably have been adjusting things wrong all these years
 DS> if I'm to believe what I just read.
Giving summary impressions first, it sounds like you read some silly 
marketing BS, with some slightly confused suggestions that aren't quite right 
outside being helpful simplified procedures within some possible context you 
didn't describe as being detailed in what you read.  Overall, Bonnie's 
comments on balancing sound output are more realistic.  
 DS> The article dealt with fixed gain on various amplifiers in a system. It
Most amp stages other than mic inputs on a board are fixed gain, and levels 
adjusted with pads of some sort at key points.  Functionally, so long as 
we're not discussing noise and headroom optimization, what's the difference 
between a 50 dB gain amp preceded by a 0-20 dB pad, and an amp with a 
variable resistor controlling gain from the feedback loop from 30-50 dB?  
 DS> said all amplifiers have a fixed design gain, and that mixing different
 DS> amplifiers in a system with different gains would cause mismatched
 DS> sound. For example: one amplifier might have a gain of 40, which means
 DS> for a 1 volt input it will produce 40 volts output (assuming it has
 DS> that much output), while another amplifier can have a gain of 35,
 DS> producing 35 volts output. 
There are different vendor standards for producing full rated power at a 
standard input level like one volt, producing one watt with the same input 
level regardless of power rating (such that a 1,000 watt amp will need 10 dB 
more drive for full output than a 100 watt amp), and many variants or 
arbitrary designs.  Within the same segment of the same array, all amps 
should be operating equally in most cases.  If the compromise of defective 
amps (even if brand new from shoddy vendors, though even Crown has had bad 
runs now and then) or mismatched brands and models must be made, it makes 
sense to set them for as similar performance as possible.  
In an ideal system this can mean setting all knobs on identical model amps 
within a unit full open or to some standard point, at which all outputs 
should be the same for the same input (with inputs from the same crossover 
output or limiter).  When this is not the case it means that either something 
about the equipment is substandard, or that within that band a compromise has 
been made to run similar speakers and amps on the same crossover but balance 
different venue areas by adjusting groups of amp inputs for some cluster 
segments.  
 DS> This suggests to me that the system balance
 DS> would change as the overall system gain was increased at the mixer. The
Not at all.  That relationship should be static within other system limits.  
 DS> article stated that the input volume controls on each amplifier needs
 DS> to be balanced - without speakers attached - so that all amplifiers
 DS> produce the same voltage output for the same voltage input.
Buzz.  Ideally all power amps are ideal current sources.  In practice they 
all have some design source and optimum load impedance based on rail 
voltages, current capacity, and other design limits.  Checking for uniform 
voltages might better be done to compare amps with all speakers connected, 
within a single bank of a given frequency, throw depth or coverage area, etc. 
 Ideally all amps within such a cluster segment should be same brand and 
model, as they may have various differences in impedance, phase performance, 
etc.  Good amps should at near full power change voltage little if you lift 
speaker loads, while less capable ones may rise in voltage without speaker 
loads.  These changes if present may also be nonlinear with volume.  If 
possible you stick the junk amps someplace else where they stand alone, in 
less critical applications.  
 DS> efficiencies. For example: a bass speaker might have a 96 db rating at
 DS> 1 watt, while a midrange speaker might have a 98 db rating at 1 watt. I
 DS> would adjust this 2 db difference at the power amplifier, so that my EQ
 DS> setting would stay closer to a straight line across the bands. With the
 DS> matching of gain concept, the EQ curve would reflect the speaker
 DS> efficiency differences, as well as the room itself, in other words, in
And mid/high horns might have much higher efficiency than that (I'd think 
front loaded mids would too, coverage dependant), as well as different bands 
having different power allocations.  Obviously feeding some 120 watt horns 
800 watts is absurd before even considering system balance.  
 DS> My question is, is the gain matching I described really as important as
 DS> Apogee makes it sound, and are my conclusions correct? In my home
There's got to be some fine print missing, which taken out of context (which 
may not be accurately described) is wrong.  
 DS> theater system, I have an Adcom 545 running a (mono) pair of JBL
 DS> subwoofer's, an Adcom 535 running the (stereo) JBL main speakers, an
 DS> Adcom 535 running the (mono) center and (mono) rear channels. With 4
 DS> distinct speaker sytems, each with different efficencies, and 2 models
 DS> of power amplifiers, with unknown gain and no volume control to adjust
 DS> on thte amplifier anyway, I can't test out this theory, but on the
If you were talking about setting power amp gains to be uniform using 
existing (or other unknown source amps) in a home multi-channel full range 
system with uniform speakers and Apogee supplying the line level decoder, 
what you described could make sense.  In any commercial system, or the home 
theater with mixed speaker types, it's absurd.  
Terry
--- Maximus 2.01wb
---------------