Slew Rate Stuff (Henry Pasternack)
Andre Jute wrote in message
1dsqy28.2ajilg1et37y8N@ts01-117.cork.indigo.ie …
When adequate slew rate current is not available, the input
capacitance cannot be driven, and the high frequency end
of the sonic spectrum is truncated because the input
capacitance forms part of a filter.
No, no, no.
The input capacitance of the second stage in combination with the output resistance of the first stage forms an RC filter that rolls off the HF response at f = 1 / (2 * pi * R* C).
Slew rate limiting is a completely different phenomenon caused by the driver stage running out of current when the instantaneous rate of change of the input signal is high. This depends on signal amplitude and can happen at any frequency. The result is not “truncation of the sonic spectrum”, but gross signal distortion, typically following high-frequency transients.
Henry
Alexander S. Moe thedj@cats.ucsc.edu wrote:
In my personal experience, interelectrode capacitance and slew rates
have a HUGE effect on the overall sound of an amp. The high frequency
response of an amp is mostly dependant on the extension far beyond the audio
range. Even though we can’t “hear” beyond 20K (or even really 15K for most
males over 30) we can “perceive” a shift in the phase of a signal an order
of magnitude beyond the audio range. (i.e. 150kHz-200kHz).
It seems as though everybody is confusing slew rate and frequency response. Interelectrode capacitance in conjunction with driver stage output resistance forms an RC low-pass filter that rolls off the gain and induces phase shift at high frequencies. This is distinct from slew-rate limiting which is a form of overload caused by driving a circuit with a signal that changes too rapidly.
If an amplifier slews, especially a feedback amplifier, the result will be obnoxious distortion. The effect of phase shift in the 150-200kHz region is a lot more subtle, if it’s audible at all. Most single-ended amplifiers and certainly phono cartridges have significant non-linear phase shift at the frequency extremes. Loudspeakers, except those designed for linear phase, are even worse. And yet they are quite able to produce perfectly marvelous sound. I don’t believe flat phase and amplitude response are as important as some people make them out to be.
An amp with every stage running high current and moreover, the
preamp, will have high frequency extension which is limited only
by the output iron (interstages too).
All other things being equal, running a tube at higher plate current reduces the plate resistance and raises the HF cutoff frequency. It also increases the slew rate. So there is a correlation between the two, but it’s coincidental, not causal. High current by itself is not a guarantee of wide bandwidth. Consider a pentode power tube used as a driver and biased at 100mA. It will never slew. But because of the high plate resistance the frequency response will suffer in comparison to a triode running at more modest current levels.
It’s also possible to design a driver with very low output impedance but also low peak output current. Put a 12AX7 biased at 0.5mA in a local feedback loop and you can obtain heroic frequency response driving the grid of an 845. But it’ll slew constantly.
Henry