"PStamler" wrote in message
...
On Tuesday, February 12, 2013 9:19:39 PM UTC-6, Arny Krueger wrote:
My calculations show that it takes about 3.55 v/microsecond slew rate to
produce a low distoriton sine wave at 20 KHz and 20 volts rms (equivalent
to
50 watts into 8 ohms).
Just about any even halfways competent amp in those days could do that.
That would imply that an amplifier putting out 7.75V rms (+20dBu) would
need a slew rate of only 1.38 V/us to produce a low-distortion 20kHz sine
wave.
My spread sheet says 1.381737, so check!
Samuel Groner's measurements of opamps suggest otherwise.
I have his 2009 book right here and I don't see that. Page number?
For lurkers, it is a freebie download that you can probably search for and
find.
Arny, I think you're assuming that if an amplifier isn't actually slewing,
it's okay -- in other words, that slewing is an all-or-nothing phenomenon.
Groner's tests suggest, instead, that well before the point of overt
slewing, the misbehavior of the
voltage amplifier stage is producing real and measurable distortion.
Your conclusion seems to be several miles west of what I actually said. ;-)
At this point we know that "real and measurable distortion" can be
fantastically low compared to what it takes to have flawless SQ.
And by the way -- in the 70s we saw a lot of consoles with LM301 opamps,
which had abysmal slew rates under usual operating conditions. And those
consoles sounded like crap, too.
The performance of a 301 can be improved to about 0.88 v/uSec which gives
you a 5 volt RMS signal at 20 KHz. Not +20 but a pretty useful signal
level, especially if this is a consumer device and reference level is -10.
These days there is a fair amount of pro stuff that only goes up that high.
It's only a few dB shy of +20.
I think that those consoles may have sounded like carp but SR wasn't the
reason why if the designers knew what they were doing.
http://www.ti.com/lit/ds/snosbs0c/snosbs0c.pdf page 7