LS> watts.  A year ago i measured 2925 watts out of it 
 LS> total very near to clipping, and recently i replaced 
 LS> the line cords with 14 guage instead of 16 guage and 
 LS> i remeasured with the same test and same impedance i 
 LS> got 1650 watts per channel into its rated 4 ohms 
With 120 VAC mains at about 40 A?  Other than the fact you'd likely never 
sustain that in actual use, NEC calls for 8 AWG wire at that current.  
 LS> instead of the manufacturers 1200 watts per channel 
 LS> into 4 ohms per channel, both channels driven at 
 LS> once.  I used a 100 HZ tone to test my amplifiers, and 
 LS> very high temperature Alpine subwoofers.  Now, is 
 LS> there a way of calculating headroom?
Buy real dummy loads of stable, known resistance (non-inductive too).  
Pick the fine print for headroom specs you feel like using.  
Generate suitable burst signals to meet those specs.  
Measure the levels of sustained and burst clip points or distortion spec 
limits.  
Compare the difference as a dB ratio.  
Points to study:  time constants of caps.  
Complex and dynamic impedance of speaker loads.  
Terry
--- Maximus 2.01wb
---------------