Amplifiers - how high a slew rate?

[syswei]

I was looking at the Bryston 7B-SST, which has a 60 volt per microsecond slew rate, and the Crown MacroTech MA2402, which has only a 13 V/usec slew rate.

Is a difference like that going to be even remotely audible? How high a slew rate is desirable in amps in this power range (500-600W into 8 ohms)?

Also, if I were to bridge the MacroTech, does the slew rate remain the same or does it double or something?

Thanks in advance to anyone who might help.

[nosebleedaudio]

I would put the Bryston several notches above the Crown.
The distortion is far lower.
And the slew rate or speed of the amp is very important.
It's a truly high end amp.
My .02cents

[Snatchman]

SOT..I always sorta vision the term " slew rate" in an analogy like this. ( this my stupid way.....) " You leave a six pack of beer in the car outside and it starts to pour down rain. So...you can either run or walk to get it. "Fast slew rate": You run outside and you get a little wet, but you are back inside quickly, settled down and drinking beer. "Slow slew rate": You walk outside, getting soaked, all pissed off,and takes a while longer to get back and settle down.... I know it's not this cut and dry, but you use what you can to remember stuff. Especially at my age...

[kenjkelly]

The slew rate will affect how well you can hear transients. If you have a slow slew-rate amp you can miss those quick impulses. Of course your speakers would have to be fast too.

On the flip side, some folks like microphones with slower responses (like LDCs) because it smooths the signal. But for monitoring you probably want to hear what is really going on transient-wise.

Doing some simple math P=(E*E)/R, so 600W*8Ohm=E squared. E=69.28 volts to deliver that power to 8 ohms. (Note that speakers are only 8ohms at specific conditions and Watts can be RMS or peak, but we have to do the math with something). So at 60V/ms you are sure to hear those 1ms pops.

[Francis Vaughan]

Small error in the above.

The slew rate is in volts per microsecond - a microsecond corresponds to 1MHz, not 1kHz.

So, why is slew rate important? Clearly it isn't about audible frequencies. You could power a small AM radio station with one of these amps.

The answer lies in issues with the feedback loop in a conventional power amplifier. In an amplifier with a feedback loop (which means pretty well all modern power amps) the output does not directly follow the input signal, but follows the difference between the input and a scaled version of the output. The trick is that any distortions in the output are seen as a difference between the output and the input and the output is driven to close the gap. Which in a naive way gets you the ability to arbitrarily reduce distortion.

However the devil in in the details. In order to reduce distortion, the core amplifier (that is without the feedback path) needs to have very high gain. In essence we trade gain against distortion. The usual trivial analysis of feedback amplifiers assumes infinite gain. Reality makes the analysis a bit messier. This need for high gain implies that the amplifier is able to slew the output at a rate needed by this gain in order to follow the distortion correction. If it can't slew this fast, the distortion reducing mechanism starts to fail in the face of demanding signals. Worse, the failure causes the entire distortion reduction mechanism to fail during this period, creating vastly worse distortion. There are a number of design mechanisms that can be applied to avoid this happening, these include limiting the bandwidth of the input signal (something which is really essential) and some sophisticated nested topologies. But in the end, a high slew rate is something worth having.

But then we get into the specmanship wars. And suddenly a high slew rate becomes of itself something that is considered a good thing. But by itself it isn't direct evidence of sonic benefit. It is a second order specification that may well be evidence of a superior amplifier, but only if used appropriately in consort with the rest of the design. Similarly a lower slew rate (so long as it doesn't get so low as to clearly compromise the ability of the amplifier to operate correctly) isn't, of itself, evidence of an inferior design.

[kenjkelly]

Dooh, thanks for catching that!