Peng
Thanks for your comments. Please note that I originally thought this amp (an old B&K EX4420) had 650 watts maximum AC power consumption based on a manual. I later saw the tag on the back of the amp, which said 900 W 64U (what does 64U mean?) by the AC inlet and fuse.
With a 900 W maximum power consumption, does a power rating of 200 wpc at 8 ohms, 350 w at 4 ohms for a stereo amp with class A/AB design seem reasonable? The manual never specified what type of rating it was, FTC, EIA, etc., but it did say power was measured at 1 KHz, not 20 Hz to 20 KHz.
I don't think 64U is directly related to any power or energy units. It may just be some code that B&K used to supplement the serial number, such as date and/or place of manufacture, revision etc., I am obviously guessing though, but again, that is not an energy, or power unit.
You can still download some of the old B&K unit manuals such as the reference 4420, as well as pictures of the back panels of the EX or Ref 4420 and I have seen 600 and 650 W but not 900 W. So again I am guessing depending on their date of manufacture, they might have used different standards.
If the so called 650 W is specified as maximum average power consumption on continuous basis, then the peaK power would be 2X that because for sine wave peak=2X average, would be 1300 W. In that case it is reasonable to assume the amp can be rated 350 W into 4 ohms for a short duration that is long enough for it to be rated as "continuous". In fact that would be in line with how Adcom, Anthem, Marantz, Parasound and even some higher end amps rate some of their amps. I know so because I do own a few amps.
Please also note that as KEW alluded to, the rated power did not specify both channels driven and no THD, freq bandwidth specs cited either. That means the 350 W into 4 ohms could be for one channel driven. In that case, if rated for both channel driven, and at 1% or higher THD, it could have been rated less, such as 300 WX2, still pretty good for a 40 lb amp.
There are no shortage of amps, some Krell, Passlab, Bouder models would be safe assumptions, that are rated continuous in truly and literally sense, but I am sure that B&K is not one of those based on the information provided in their manual and pictures of their innards that are googleable.