LS> I meant amp in parallel with 8 ohm speaker, and the 
 LS> ammeter in between the two.
Do you really mean two nominal 8 ohm speakers in parallel, yielding a 4 ohm 
nominal load?  (Still, dummy loads are usually used for power ratings, as 
speakers have so many quirks.)  
 LS> The building's wiring is ok, except for line losses from the 6 guage 
 LS> wire they feed the house on AFTER i installed the 
 LS> proper 0/3 cabling on my 200 ampere meter feed.  
 LS> Still hydro insists on using 6 guage from the pole to 
 LS> the house.
That's not possible for a feed of more than 60 Amps.  I've seen 4 AWG Cu THHN 
inside a meter housing catch fire in a 100 A service loaded to 70-80 Amps.  I 
wouldn't be surpised to see 1 AWG Al (service wire is often high temp rated, 
and not subject to the cooling restriction or flammables exposure of indoor 
wiring).  1/0 Cu would be nice.  You have to allow for xfmr drop and breakers 
too in figuring service losses.  
TS > LS> speaker did not like it, i measured 167 volts at 100 TS > LS> ampers 
into the speaker.  I noticed amps are getting
167 V P-P into 4 ohms would yield 870 Watts and a current of 42 A P-P, or 
about 15 A RMS.  Either your load is way below 4 ohms, or you have a huge 
reactive current flow, or inaccurate metering.  What are the plus and minus 
upper and intermediate rail voltages?  
TS > LS> all my amplifiers with this class AB dual rail system.
TS > It would be hard to spec meaningful efficiency on such a TS > design 
without the context of the measurement conditions,TS > as it would be highly 
variable.
 LS> Yes, i figured that much out talking to an amplifier 
 LS> company named Mark Levinson and Krell.  Class A/B has 
 LS> distortion which i did not know about until now.  The 
 LS> distortion between the puch pull rails known as 
 LS> crossover distortion when the transistors are not 
 LS> working to their linear maximum.
Crossover distortion is a bias matching problem, often most noted at low 
levels and high frequencies.  All analog circuits have some distortion, 
though the best I've seen are around 0.0002% THD.  That wasn't consumer 
electronics.  
BTW, Mark's company and Krell are both local to me.  They're both a little 
strange, and not what is practical for most apps you're discussing.  
Terry
--- Maximus 2.01wb
---------------