How many of the amplifiers suffered from crossover distortion?? Before the cause was discovered??
Aside from that, transistors do produce harsher overdrive distortion than tubes. The Saturation and cutoff are sharper curves. But who runs their amplifiers at maximum possible output?

 

Can't wait to crank Who's Next for the sons-in-law.

 

Can't wait to crank Who's Next for the sons-in-law.

1. Ya gotta break 'em in right.

2. Their reactions to Behind Blue Eyes should be interesting.

3. For an encore, might I suggest Live at Leeds / one of the re-issue CD's / A Quick One

In round numbers, my sibs and I saw that performance. It was very good.

ak

 

How many of the amplifiers suffered from crossover distortion?? Before the cause was discovered??

Not many. Crossover distortion was "discovered" in the 20's (1920's) and the two most common approaches today were standardized in the 30's, so it wasn't any kind of surprise. What *was* a surprise was thermal runaway. It wasn't as widespread a problem as the press would have had you believe, but it did slow the acceptance of solid-state amps.

Fun fact: The output video amplifier in the first broadcast-legal videotape machine ran on plus *and minus* 300 V supplies. IIRC the output stage was a cathode-follower working into a constant-current sink, just like class A audio, with global feedback to stabilize the output blanking level at 0 V. Designed by Ray Dolby in 1956.

ak

 

I hold that others are certainly welcome to enjoy that music, live, or perfectlt reproduced from recorded as perfectly as they can afford.
I want none of it!! you are certainly welcome to listen to whatever pleases you at whatever sound level does not rattle my house or car. I find "perfect fidelity " music rather boring and totally uninspiring. I do ask that you extend me the same courtesy of not commenting further on the requirements needed to satisfy your musical tastes, as I will not criticize yours.

So is it that you do not listen to recorded music, or that you aren't particularly concerned about the fidelity of the sound?

 

So is it that you do not listen to recorded music, or that you aren't particularly concerned about the fidelity of the sound?

I listen to recorded music, tape, CDs, Vinyl. The fidelity is "good" but certainly not the 0.01% THD required by some folks. AND, how can you sense which distortion is added on playback when the performances were using overdriven tube amps back in 1971??

 

The whole concept of "Feedback loop delay" in amplifiers is far stupider than "just a bunch of hooey", it is, in reality, screamingly STUPID!!!! Feedback happens cycle by cycle, it can not be delayed.
On top of that, the big "advantage" of the direct-coupled amplifiers is that drift in the first stage will take out the expensive transistors in the output stage.. And possibly others in between.

What kind of feedback is being discussed in this thread now?
Feedback actually happens instantaneously in a linear circuit. The adjustment due to that feedback could be delayed, but it would be so fast or else the feedback would not work as intended, and if it did not work as intended, then the amplifier would not sound like it had the right feedback and it would be noticeable. However, if by chance it was noticeable, it would have to have been on purpose and part of the design.

I could not imagine a modern amplifier without feedback. That's because of the way feedback corrects for defects in the rest of the circuit such as nonlinearities.

 

2. Their reactions to Behind Blue Eyes should be interesting.

Well I'm sure they've heard it. But then there's actually hearing it. The one I used to crank on the dorm floor was Goin' Mobile.

I've got some other classic vinyl to share, too. Pretty sure Dark Side of the Moon is in there, as well as Rubber Soul, and at least a couple Stones albums.

 

Hello there,

Well those sound like some very profound statements about audio. However, I am very leery about hearing statements that profess one type of audio amplifier over another or really about how listeners perceive the sounds they hear to be either good or bad. That's especially since I read about the Stradivarius listening tests that had shown that even the best soloists preferred modern violins over million-dollar vintage Italian designs in blind tests.
It seems almost like the old proverb, "beauty is in the eye of the beholder", converted into "beauty is in the ear of the beholder", which makes a lot of sense.
Also, "perfect" may not translate to "desirable" in all cases, although we would need a precise definition of what we want to deem as being "perfect".

Perhaps you might reveal where you get your information from, and personal experiences with audio.

Capacitors though were always coming into question when used in audio work.

I never stated anything about what one may prefer, but reproducing what was recorded.
As recording quality improves, don't you wish the most accurate and natural sound possible?
If you wish different, that is your prerogative. However, misleading and misinformation, that some
have provided, is not conducive to improving the audio experience.

Blind testing as pushed in audio and other entities, are skewed 100% of the time towards no sonic
difference and are not scientific in any way, shape, or form. It is one of the common forms of scams, and
trickery possible. Simply addressing one confound variable, sight, is a marketing technique used by
manufacturers and organizations for years. One must delve into the medical field to find some of the
confound variables to be addressed. A few questions for any to answer.

1. Prove that back and forth blind AB testing for 10 times, or even 5 times equates to the normal, non sighted
experience comparison.
2. With a group test, with 50% "seated" in bass increasing and 50% in bass decreasing, how does one obtain
a 95% confidence conclusion?
3. Prove that the switching mechanism(s) does not negatively affect the sonic character, such as masking of
inner detail?

It is simply amazing how natural the sound can be from a system that uses all polypropylene decoupling
capacitors vs electrolytic capacitors. One cannot force recording component manufactures to upgrade,
but at least the playback can be improved. As the saying goes, two wrongs don't make a right.
The question is do you wish the most natural, or simply manipulate a degraded musical signal?

One of the difficult question is what brand caps, and other parts are accurate/natural, and how good is the
design.

cheers
pos

 

The whole concept of "Feedback loop delay" in amplifiers is far stupider than "just a bunch of hooey", it is, in reality, screamingly STUPID!!!! Feedback happens cycle by cycle, it can not be delayed.
On top of that, the big "advantage" of the direct-coupled amplifiers is that drift in the first stage will take out the expensive transistors in the output stage.. And possibly others in between.

Sorry to be so blunt, but that is a point blank false statement Bill. Every device delays the signal
to some extent, which gives the device a figure of merit, an upper frequency limit, along with the
associated circuitry. Miller Capacitance is a negative for sure, limits the bandwidth, and delays the
input to output. Each stage is involved. This is first semester electronics engineering.

How is your second paragraph a big "advantage" if the output stages are destroyed? Might want to
rephrase? Actually, unless designed properly, direct-coupled between two stages will cause a bass
boost. But that is if you understand the thevenin equivalent circuit, how a basic circuit works.

cheers
pos

 

"perfect fidelity" is only available at a live performance. The very good reproduction, which came across to me, was at the home of one we were visiting. I did not ask about the details of the system, as it would not have been appropriate in the context of the visit.
The need for "perfect fidelity", like other expensive things,is an acquired taste, not appealing to me. Perfectly OK for others, as presently we do not all have to be alike. THAT FREEDOM is precious indeed!!

Yes, you won't be seated where you might like. However, you will never know how good the sound of the
recording is until you truly obtain a great audio system. It is not just slightly better.
You will think you were there. Here is a recording to check your system, just how good it is.

Just better fidelity is an improvement. Real, extraordinary fidelity is an acquired taste? If that is the case,
why go to the live event at all?? I take it you do not. No sweat, but please don't bully others into not
going for obviously better fidelity.

The start of this string was posting/teaching obviously misleading, false information.

Cheers
pos

 

You know this how? I'm not doubting it, just wondering if you found a schematic somewhere.

45 years, starting in 1980, and seen many schematics.
Next, open any piece of gear and electrolytic capacitors abound
as decoupling capacitors. The the power supply decoupling capacitor at the
analog section of the dac chip as well as in the following analog devices
(chip or discrete).

Electrolytic capacitors actually become inductors at low frequencies. High frequency
type electrolytics are better, but still not nearly as good as poly caps. Attached is a
graph of typical electrolytic capacitors. When the curve starts, that is when the cap
is becoming an inductor. Not good.

Next, Layout is quite important as common wires or foils cause bleeding, channel
separation problems. And it is frequency sensitive, so the actual dual channel response
is Not flat.
Layout because frequencies as low as 1khz transmit through the air, actually
measured, impedance sensitive. So while just one channel might measure flat,
lack of channel separation will cause non linear frequency response of both
channels together. (Specs typically only show one channel.)

One more thing. Even a 1n914/4148 diode, with one pf (1 micro micro farad) of junction capacitance
will bleed audio when reversed biased/shut off.

Hope this answers your questions.
pos

 

Sorry to be so blunt, but that is a point blank false statement Bill. Every device delays the signal
to some extent, which gives the device a figure of merit, an upper frequency limit, along with the
associated circuitry. Miller Capacitance is a negative for sure, limits the bandwidth, and delays the
input to output. That is each stage. This is first semester electronics engineering.

To be blunt in return, the delay of a signal through an amp has absolutely nothing to do with the fidelity of the signal at the output.
And Miller Capacitance simply provides a rolloff of the high frequencies, nothing unusual, and it's readily compensated.
These days it's quite easy to design a good audio amplifier with much greater than a 20kHz bandwidth.

Perhaps you should go past first semester electronics of 40 years ago.

Electrolytic capacitors actually become inductors at low frequencies. High frequency
type electrolytics are better, but still not nearly as good as poly caps. Attached is a
graph of typical electrolytic capacitors. When the curve starts, that is when the cap
is becoming an inductor.

That's why any good amp has a smaller non-electrolytic capacitor with low inductance in parallel with the electrolytic.

 

I listen to recorded music, tape, CDs, Vinyl. The fidelity is "good" but certainly not the 0.01% THD required by some folks. AND, how can you sense which distortion is added on playback when the performances were using overdriven tube amps back in 1971??

The difference between your system and the best is quite large, not .01% THD.
You give your audio system way too much credit. You still do not understand the problems.
Design, Parts sonic quality, Layout, Switch type, even type of wire used.

Switching from 11 parallel per leg 3N copper speaker wires (typical hardware store) to
11 parallel 6N pure copper wires required altering in speaker crossover part's values
(both capacitor and inductors/chokes/resistor) as well as updating the dac.
That is how large the sonic change was.

Why parallel speaker wires? Two reasons. Total gauge to adjust the woofer damping while
parallel to lower the reactance as the frequency rises.

cheers
pos

 

The start of this string was posting/teaching obviously misleading, false information.

Actually, the start of this *thread* was a question.

ak

 

To be blunt in return, the delay of a signal through an amp has absolutely nothing to do with the fidelity of the signal at the output.
And Miller Capacitance simply provides a rolloff of the high frequencies, nothing unusual, and it's readily compensated.
These days it's quite easy to design a good audio amplifier with much greater than a 20kHz bandwidth.

Perhaps you should go past first semester electronics of 40 years ago.
That's why any good amp has a smaller non-electrolytic capacitor with low inductance in parallel with the electrolytic.

I see more generic comments with no education, just misinformation. Science is
science and is timeless.

Of course it does as it alters the input signal from the preceding source output.
Not only that, but that original feedback signal, plus altered output signal "travels"
through the amp a second time, looping, thus altering the input signal even more.
Besides that, the amp parts degrade the original feedback signal multiple times, as well,
again altering the input musical signal. You cannot alter real science.

Secondly, the HD increases as the frequency rises, over multiple octaves,
thus altering the harmonic structure. (Rane discusses bandwidth of peaks, valleys and effects on
perception.)
Again I suggest obtaining the RCA Radiotron Designers Handbook, 26 engineers. It applies to solid
state just as with tubes, as does my engineering textbooks and degree.
In fact, SS amps use more devices than tube amps, have much more internal resistances/ poorer quality
junction capacitances than tubes.

Next, rolling off the high frequencies affects the delay time and phase. Let me quote with more explanation.

"Phase Shift: In an ideal audio system, all frequencies should pass through with their time relationships intact. However, in reality, all electronic components (amplifiers, filters, etc.) and even cables introduce a "phase shift," which is a delay that varies with frequency. This means different frequency components of a complex musical signal can be delayed by different amounts."
"Group Delay: This is a measure of how much the different frequency components of a signal are delayed as they pass through a system. If the group delay is not constant across the entire audio spectrum, it can lead to "smearing" of the soundstage and a less defined audio image. For a system to be considered to have a linear phase, the group delay should be constant."

A smaller poly cap in parallel with an electrolytic capacitor is not close to optimum. It certainly
does not cancel the negative effects of the electrolytic caps except under carefully controlled
conditions, and not 100% even then.
Electrolytic capacitors negatively affect the bass just as much as the mids and highs. The smaller
parallel poly cap cannot be as effective at bass frequencies vs mid and highs, thus is
frequency sensitive.
You made another common assumption and mistake. Please do not push off your misinformation
on others.

pos

 

Actually, the start of this *thread* was a question.

ak

You are correct. I meant some of the replies I read on the first page.

Thank you for bringing that to my attention AnalogKid.

cheers
pos

 

Hope this answers your questions.

Indeed it does. It tells me you have no freaking idea about the device in question.

 

Sorry to be so blunt, but that is a point blank false statement Bill. Every device delays the signal
to some extent, which gives the device a figure of merit, an upper frequency limit, along with the
associated circuitry. Miller Capacitance is a negative for sure, limits the bandwidth, and delays the
input to output. Each stage is involved. This is first semester electronics engineering.

How is your second paragraph a big "advantage" if the output stages are destroyed? Might want to
rephrase? Actually, unless designed properly, direct-coupled between two stages will cause a bass
boost. But that is if you understand the thevenin equivalent circuit, how a basic circuit works.

cheers
pos

Hello again,

What delay are you talking about?
If you are going to talk about a delay in the feedback you are going to have to show some sort of circuit because this would be the only way to start to understand this delay you talk about. I am asking about this because any 'delay' in the feedback would have to be just part of the system, and feedback is what makes it work in the first place.

If it is purely electrical, then any delay would have already been accounted for in the frequency response, and if the specs are 20Hz to 20kHz then that's the response regardless what the 'delay' is.
Maybe you are talking about the acoustic feedback ideas that involve sensing the audio out of the speaker and using that as feedback?

Pure electrical feedback has been used for a long time and it always improves the reproducibility of the system or what we also refer to as compliance. If you input a pure sine wave then you should get a pure sine wave out. If the feedforward part of the circuit changes the wave in some way then the feedback will correct that to a huge degree by means of the way negative feedback works. In the limit with an infinite gain, the output signal is PERFECT and this is easy to show. In a real life system, the gain is not infinite but is usually very high ... high enough to get the error down to an incredibly low value. Without that feedback though, the signal could be very distorted.
All of this is not hard to show mathematically using circuit analysis.

Perhaps you are also referring to less than adequate components. You talk about decoupling capacitors but could it be you mean coupling capacitors, or is it both?
Coupling capacitors have been brought into question, we know that, but for decoupling an electrolytic plus some good quality ceramics have been used. Without the electrolytic caps the bass response would suffer in high power systems. If you know something inferior about this kind of setup then you would have to state what it is specifically, and it would help if you stated what the solution might be so we can see with no difficulty what the issue was.

Not sure how you feel about class D audio amps and related. The benefit here is very small size vs power handling capabilities.

It may not help much to say this is better than that, or this is worse than that, because anybody can say stuff like that. It's only when we get down to the circuit level that we can truly evaluate what is good and what is bad. This could be as simple as describing what makes a component good or bad, or it may involve showing individual circuits that coincidentally exploit these problems and some that don't. We could start with a single circuit and see what we can find out, just to start with.

 

45 years, starting in 1980, and seen many schematics.
Next, open any piece of gear and electrolytic capacitors abound
as decoupling capacitors. The the power supply decoupling capacitor at the
analog section of the dac chip as well as in the following analog devices
(chip or discrete).

Electrolytic capacitors actually become inductors at low frequencies. High frequency
type electrolytics are better, but still not nearly as good as poly caps. Attached is a
graph of typical electrolytic capacitors. When the curve starts, that is when the cap
is becoming an inductor. Not good.

Next, Layout is quite important as common wires or foils cause bleeding, channel
separation problems. And it is frequency sensitive, so the actual dual channel response
is Not flat.
Layout because frequencies as low as 1khz transmit through the air, actually
measured, impedance sensitive. So while just one channel might measure flat,
lack of channel separation will cause non linear frequency response of both
channels together. (Specs typically only show one channel.)

One more thing. Even a 1n914/4148 diode, with one pf (1 micro micro farad) of junction capacitance
will bleed audio when reversed biased/shut off.

Hope this answers your questions.
pos

Hi,

I did not see this post at first, that's very good that you posted something more concrete.

Electrolytics do have problems as you note, but they are not usually used alone. Maybe in older designs they were, but today they are almost always used with a smaller value ceramic in parallel. That's been known for a long time now, even back in the days of base line Transistor Transistor Logic (TTL) circuits.
If you know of a reason why this might not work for audio it would be good to mention here and interesting to read.
I assume you are talking about power supply bypassing.