TS > LS> of 2925 watts plugged into a worst case wall 
TS > LS> plug.  Both amps are
TS > I would estimate that your 2400 W amp can only sustain 1500 TS > W in a 
worst case wall plug.  If you can prove otherwise,
 LS> That's what the FTC claims in these specs.  But that differed when i 
You've since said that this amp is wired to take two 20 A 120 circuits via 
dedicated twistlocks.  That's hardly an "ordinary wall plug" worst case, 
sometimes a flaky 15 A circuit.  Transient peak ability doesn't make an amp 
capable of greater than 100% I/O efficiency.  
 LS> upgraded the toroid and the capacitors FROM it's 
 LS> 6800mfd.  PLus i had to up the bridge rectifiers from 
 LS> 30 amperes to 60 amperes because the thermocouple kept 
 LS> killing the power because the rectifiers were the 
The usual rule of thumb for caps is 1,000 uF per amp of intended PS capacity. 
 Power supply parts are expensive and take space, and so some designers cut 
corners.  
TS > LS> watt amplifier like the one above, but i modified the TS > LS> power 
supply, and i measured it can hold 4,000 watts TS > LS> total power, RMS, but 
i needed a 35 ampere breaker
TS > 4,000 Watts is a rate of energy transfer, not an amount of TS > energy.
 LS> Well everytime i tried to make this amp huskier on 
 LS> it's wattage, what you said above is pure BS.
Wrong.  Might I suggest you study basic physics, and international standard 
units?  "Watts" are not a quantity that can be "held".  Care to try for 
watt-seconds, Coulombs, or some suitable unit?  
 LS> is pure lethal because it supplies 100 amperes with a 
 LS> 50 percent reduction in damping factor at 167 volts.  
I presume you're talking about some undefined momentary peak measurement 
here, and a coupling resistance or diode drop between upper and lower supply 
rails as to the nominal damping factor?  
 LS> I can ARC weld with this amp in bridged mode without 
 LS> clipping and a bit of heat off the heatsinks in the 
 LS> amp.
I can't see that sustaining arc welding, as the current ability would be too 
low for anything but momentary surges.  Real welding uses lower voltages that 
sustain what you're tlaking about as peak currents only.  
TS > Do you know what the ESR of the filter capacitors in that TS > amp is, 
and what their capacity and no/full load rail
TS > voltages are?
 LS> Yes, explained below.  This amp uses dual 
 LS> complementary rails instead of trying to compare it 
 LS> to some cheapo single complementary rail amplifier 
 LS> like you are trying to do.
Ever see the Stewart amp design?  By comparison, you could argue that dual 
rail voltages are a kludge.  Dual rails have been used as the basis of 1 kW 
AM broadcast transmitter designs popular from 1976 to 1985.  Amp designers 
may have borrowed the idea there, while such TX's have moved on to quadrature 
and digital waveform synthesis designs using MOSFETs.  Every design is some 
kind of compromise, as there are different distortion modes and design 
complexities to most fancier circuits.  
 LS> The ESR of the caps is 6800mfd per rail.  It uses a +145v, +78v, 0v -
ESR is equivalent series resistance, the internal impedance of a cap which 
causes heating and limits current delivery ability.  It usually decreases 
with larger capacity, but is also affected by other design actors.  Low ESR, 
of a few hundredths of an ohm or less, is typical of better quality caps used 
for amp filters, strobe or snubber circuits, etc.  
 LS> 78v, and -145v rails out of the power supply to the 
 LS> rails.  6800mfd per rail is the capacitance in the 
 LS> amplifier, which is not much.  The rails were built 
30,000 uF or more would seem desirable for such an amp IMHO, unless it was of 
a supersonic DC-DC converter design.  With the dual rail bridgable mono 
design it appears you're describing, 4 such caps would be required, not 
exactly a small chassis real estate allocation.  Using separate supplies per 
channel, you'd be looking at 8 caps, though dividing the current handling.  
 LS> this way in the amplifier to prevent harsh clipping 
 LS> on the output rails and causing the back EMF from the 
 LS> speaker from interefereing with the damping at high 
 LS> levels of power.  A wall plug supplies 1875 watts at 
That sounds irrelevant.  It has to do with thermal efficiency and idling 
power considerations.  You hit headroom limits at the same point as if the 
amp had single rails only of identical upper volatge and current capability.  
 LS> breaker panel, it can sustain exactly what the 
 LS> manufacturer said it would - 3,400 watts.  And yes the 
 LS> amp now NEEDS a 35 ampere breaker from the main 
That's pushing it, as while 4200 watts in is credible for a high efficiency 
amp, and likely wouldn't trip a breaker of that rating for a few hours, it's 
exceeding normal sustained loading codes.  
 LS> sustained 80 percent duty peaks.  Energy transfer is 
 LS> 75 percent efficient in the most worst case (plugged 
 LS> into a 18 or 16 guage line cord).  This is what i 
That's not a measure of the amp, but is a safety hazard.  
 LS> measured out of the amplifier, total, each channel 
 LS> sustained at 2.67 ohm loads (3 8 ohm speakers in 
 LS> parallel) with a bit of reactivity and back EMF from the speakers.
Actual speaker Z isn't the same as nominal at most frequencies.  Any power 
claims you make based on voltage measurements into speakers are suspect.  
One reason amps need more current capacity than a nominal power output should 
demand, and benefit from low damping factor, is that speakers do have 
reactive quirks plus resistances and impedances below nominal at some 
frequencies.  
Terry
--- Maximus 2.01wb
---------------