Of course, otherwise we would all agree on the minimum sr number. But Richard has presented his theory regarding the indirect effect of high slew rate (may be what we hear is not the slew rate itself but what may come along with high slew rate).
I agree. If we made slew rate high then distortion will be rise too (in the same amp). And then we can not find exactly how slew rate will influence to the sound. We only analyze the input signal (music) and can not find music with slew rate more than 1V/uS/1Vp. And we stop all experiments. This is enough, nothing to improve. We already became master of electronics and biology too 😀
Since some people pretend to ear differences when a 2V/us
signal is slewed by a 200V/us amplifier rather than a 20V/us one
then just changing the RC input network frequency corner
would be audible when shifting the corner from 2MHz down
to 200KHz , so you can use two amplifiers with this only
difference and still fulfill the condition of same amp but
different SRs.
Don't you forget internal distortions, specially high order ones (MHz) from Class AB output stage ( and their IM with the signal ) supposed to be canceled (being part of ) by the closed loop !
I agree with your reasoning about SR for every amp (or preamp) with no GNFb, and just look at the response curve, HD and IM for them.
Last edited:
I have experiment to made RC filter around 300kHz to my VSSA. But I can not find any change of the sound. It is not about the input signal. With RC filter around 300kHz, my VSSA still sound different than my VFA.
Yes that is accepted, the problem is when we slow some amplifier down again and again until some audible differences start appearing. Then when we measure, the slewrate is still sufficient for 20k signal.
Just like THD, it doesn't need to be ultra low, the slew rate also not necessary to be placed as a design goal. Just like I said before, ulta high slewrate isn't affecting the sound, also ultra low THD. There are "another things".
A quick search did not reveal any free scholar authoritative references. I'll keep searching, meantime here's a nice description of IM3 cancellation and it's relationship with the bias. You'll of course note in fig. 6.30 (b), and the associated text, Re=1/2gm as one of the two possible conditions for IM3 cancellation.
Thank you. No local library has a copy where I can read the full text so it will take some time to understand from the limited excerpt.
Best wishes
David
Of course, hi-slew-rate it is not about having high bandwidch.
Most of the input signal after 20K are distortion products and parasitic RF.
You were wise to add this filter, i always try to add one as low FC as possible.
I believe one of the reasons why some prefer input trafo and symmetrical wiring is their ability to bring a very efficient high pass filter and reject cabling induced noise.
At this point, if somebody is expecting me to jump in, I won't. The amount of subjective garbage and engineering nonsense is exceeding any limit that can be processed by a rational human.
So keep up the good work, it's a safe method to keep me away from this thread. Any attempt to return to reality will be immediately sanctioned with my contributions.
So keep up the good work, it's a safe method to keep me away from this thread. Any attempt to return to reality will be immediately sanctioned with my contributions.
Just keep away your fanaticism to single very limited topology, because it prevent your self to figure out the things behind the amplifier quickness. You don't even mentioning your spec limitations to face my 0.01%THD limit.
(underlined by me)
Amen!
Do you really think I'm that stupid ?🙄
If you had read post #3135 an itty-bitty more carefully, than you should have known I was talking about the noise (or hum or whatsoever) from only one current source. It is this noise, split in the next stage (and still correlated!) that will finally be cancelled out when combined by the TISes.
The purpose of a cascode is to eliminate the Early and Miller effect. Regarding the first (common emitter) stage, this is (best) done by keeping Vce constant (as in my schematic, not yours). Regarding the next (common base) stage, this is done by feeding the signal into the emitter and recycling the base current (a la Hawhsford). As a result, the output impedance is much much higher than from single high Early voltage transistor (contrary to your explanation).
From day one, these are the basic principles which determine merits of a cascode. Selecting proper transistors for a cascode is just a side-issue, and in no way tells the whole story (talking about engineering nonsense 🙄).
Apart from noise, it does make a practical difference. PSRR is at least 20dB better (it increased from 70 to 90...130dB, depending on temperature, for which the test circuit has not yet been compensated).
Last edited:
PS: Something went wrong during editing, as the max. editing time expired.
The following text:
Regarding the next (common base) stage, this is done by feeding the signal into the emitter and recycling the base current (a la Hawksford). As a result, the output impedance is much much higher ...
Should be (and equal to the original version):
Regarding the next (common base) stage, this is done by feeding the signal from a high impedance into its emitter. As a result, the output impedance is much much higher...
So, just forget Hawksford (for the moment).
The following text:
Regarding the next (common base) stage, this is done by feeding the signal into the emitter and recycling the base current (a la Hawksford). As a result, the output impedance is much much higher ...
Should be (and equal to the original version):
Regarding the next (common base) stage, this is done by feeding the signal from a high impedance into its emitter. As a result, the output impedance is much much higher...
So, just forget Hawksford (for the moment).
Last edited:
How could we configure the distortion cancellation method shown by Dimitri (Elect Design Jul 19, 2012) into a complimentary amp circuit?
Thx-RNMarsh
Thx-RNMarsh
Here is the link to the article in Electronic design:
JFET Follower Amplifier Cancels Distortion | Analog content from Electronic Design
JFET Follower Amplifier Cancels Distortion | Analog content from Electronic Design
at best its a variation on "replica cancellation - where you are "copying" the load current, "adding" its effect back in a Class A stage
these characteristics make for some difficulty in translating the principle to Class AB output power amp with varying Z with frequency, noticeable nonlinearity in voice coil heating, inductance modulation...
these characteristics make for some difficulty in translating the principle to Class AB output power amp with varying Z with frequency, noticeable nonlinearity in voice coil heating, inductance modulation...