Been done many times. The model that holds pretty well is energy peak between 50 -250 Hz and a 3 dB per octave roll off.
Now for a simple question. When using an audio power amplifier with a differential input stage do you get more or less distortion if you inject you test signal into the inverting input instead of the non-inverting one?
Actual measurement please no simulations. I do think The Rev. Pass has a patent somewhat related to to this.
Actual measurement please no simulations. I do think The Rev. Pass has a patent somewhat related to to this.
recognition of CM input nonlinear issues is very old but somewhere you have to come up with the low Z drive
Cherry, Pass both mention it in PA, Cordell inverting "super gain clone"
in op amps: Jim Williams "always invert", Pease, Groner published input CM nonlinearity plots...
Cherry, Pass both mention it in PA, Cordell inverting "super gain clone"
in op amps: Jim Williams "always invert", Pease, Groner published input CM nonlinearity plots...
Summing or inverting input is lower distortion in general. It removes the CM component from the input.
Not sure what that has to do with distribution of slew rate, the crest factor of recordings varies quite a bit RIAA pre-emphaized even more (JC's problem).
Doesn't that just demonstrate that MM don't create huge amounts of spurious wibbling in response to a transient?
It just means that a MM cartridge usually has a 4 pole low pass filter around 20KHz, and of course, this reduces the need for high slew rate. However, I design for all phono cartridges, not just a Shure or its equivalent.
The Aleph X production amps could care less it seems Scott is correct once again.
Why not use CFA and VFA I have works just fine just feedback.
Just pick a volume level and you can model the energy per hertz if you want to for a typical music playback. Dynamic range is that the peaks for a live symphonic orchestra is 30 dB above normal level. recordings almost always limit that to 20 dB.
More popular music may have as little as 10 dB. headroom.
The difference between a pianissimo and a forte can be 100 dB but is more often around 60 dB.
There actually is a MIDI standard for the weighting.
Now for slew rate if you know the energy level and frequency for a given efficiency reproducer you can calculate the slew rate required.
15 V/uS is actually just adequate for many of my systems. I do need more in some cases and home reproduction may use much less efficient loudspeakers but since the loudspeaker to ear distance is much less probably doesn't ever exceed that.
I presume you made the actual measurements, so that brings up the issue of symmetry in the circuitry. Is the circuit fully symmetric?
100V/us usually sounds better Ed. We found this out in the 70's, and worked hard to 'justify' this conclusion. However, high slew rate is only the 'tip of the iceberg', because to get such a high slew rate, improved input linearity (usually achieved by a largish series resistor to reproduce the input transconductance) is almost mandatory, higher open loop bandwidth which CFA provides more easily, and perhaps other factors.
Actually no the new integrated uses CFA pre for simplicity and space combined with a .8 power amp. Maybe cheating a bit but it works well and sounds good.
Yes I measured and it was a symmetrical summing junction on that series of product.
Vinyl ticks and pops deliver the highest slew rate figures, considerably beyond
program material. I am not at all certain that I want to bother reproducing
them with perfect fidelity.
You and I discussed it once, and agreed that 10X peak program material was
probably an adequate margin - that puts us at something like 15 V/uS for a
100 watt amplifier.
😎
The slew rate is on a sample to sample basis a time domain thing you can create a multi-tone signal with exactly the same energy at each frequency with a large range of peak slew rates.
- Status
- Not open for further replies.