You don't believe it - and that's it. No argument will change your opinion. And this is the base of most postings here in that forum - the strong beliefs of its members. It is the never ending debate about science and religion. My life experience with audio technique is quite similar to the one Gnobuddy just described: Doing many experiments, reading many textbooks dealing with technical aspects and human sound perception as well trying to understand everything. At the end, i.e. in the eighties, it was clear that sound performance of amps were good enough for the restricted capabities of human sound perception. Since them I lost the interest in HiFi amplification, it is a minor aspect.
I believe in science - and that's it.
I believe in science - and that's it.
Could it be that science haven`t revealed everything yet and we therefore cant do the rellevant measurements? I can alway hear a clear difference when I push direct on any quality of amplifier when tone controls are in neutral. I never see measurements compairing amplifiers or sources done with musical signals or speakers hooked up. Those who says all amps sound the same may be right but maybe they or noone knows how do do the measurements. Electronics is young science. Im divided.
Cheers!
Not at all, electronics, specially audio amplification, is old science. Older than me!
The knowledge to answer your questions is available since decades. You have to find it
Because full osmosis has not yet materialized, I still have not full understanding of this complex circuit. But I just realized increasing the gain of the first stage will also cause the same increase when the unit is in bypass. Then it would be a fixed boost pedal with no adjustment, unless circuitry is added to compensate.
Off the cuff, it looks like the addition of a divider immediately after C003 would do it. Q001 would buffer the increased resistance. This mod will require trace cutting on the PCB. which is undesirable yet doable.
I might try the opamp in stage one as unity gain buffer and see what happens first, but any noise in stage two will not be buried as recommended previously, especially the 220k FB resistor 044.
A respondent to an EE Times article said this:
"Many audio circuit designers seem to ignore the fact that, with regard to noise, all resistors are the same EXCEPT when DC flows through them - which, of course, causes the "excess noise". Most applications of resistors in audio circuits see no DC, yet many audiophiles believe that an expensive resistor will result in lower noise. Of course, low excess noise resistors are useful in applications where there DC does flow - the plate load resistors in vacuum-tube circuits, for example."
Is not DC flowing on each half of the cycle? Can anyone expound on what this man is saying?
Off the cuff, it looks like the addition of a divider immediately after C003 would do it. Q001 would buffer the increased resistance. This mod will require trace cutting on the PCB. which is undesirable yet doable.
I might try the opamp in stage one as unity gain buffer and see what happens first, but any noise in stage two will not be buried as recommended previously, especially the 220k FB resistor 044.
A respondent to an EE Times article said this:
"Many audio circuit designers seem to ignore the fact that, with regard to noise, all resistors are the same EXCEPT when DC flows through them - which, of course, causes the "excess noise". Most applications of resistors in audio circuits see no DC, yet many audiophiles believe that an expensive resistor will result in lower noise. Of course, low excess noise resistors are useful in applications where there DC does flow - the plate load resistors in vacuum-tube circuits, for example."
Is not DC flowing on each half of the cycle? Can anyone expound on what this man is saying?
This is a very good observation, and it illustrates the vast gulf between what the typical magazine reviewer does, and what real audio researchers did.
Peter Walker founded QUAD in 1936. By the mid 1950s or so, the amplifiers he was building - even the P.A. amps - were beginning to measure really well, and sound really good. He was beginning to suspect the loudspeaker was more of a problem than the amp, but encountered the same question/criticism that you just raised. Maybe the amp is flawless on the test signals (sine waves, etc), but how do you know it is equally flawless with music?
So Walker invented a brilliant (and yet very simple) measurement technique. Tap a little signal from the output of an amp. Subtract it from the input signal. Adjust the signal levels to match, (i.e. to cancel as much as possible), and look at the result.
If the amp is perfect, the output is a perfect amplified copy of the input. If the amp has a voltage gain of 10, and you take the output and divide by 10, the resulting signal should exactly match the input signal. Subtract one from the other, and the result will be exactly zero.
If the amp is not perfect, it will add distortion and noise to the signal. Now when you do the subtraction, that added distortion and noise will not subtract away (because those things are not in the input signal, only in the output signal.)
Mathematically, if the input signal is "x", and the amplifier gain is "A", the output signal of a perfect amp is A*x. Divide by A, you get x. Subtract x from x, you get zero.
If the amp is flawed, it adds distortion "D", so the output is A*(x+D). Divide by A, you get x+D. Subtract the input x, and you're left with just (D) - the distortion alone.
So Walker's measurement directly picks out any flaws in the amplifier's output signal. You don't have to use sine or square test signals - you can play Mozart or Metallica or whatever through the amp. You can connect the output to an actual loudspeaker, not a dummy load, to eliminate concerns that the amp is misbehaving when an actual loudspeaker is connected. And the measurement will tell you exactly what the amplifier is doing in real, live, music-listening conditions. Is it flawless, or is it not?
In those days, all amps were tube amps, and all were AC coupled, so there was inevitably some phase shift at the very low and very high ends of the frequency range. So Walker added the ability to add small compensating phase shifts to his subtraction network, the procedure once again being to tweak amplitude and phase for the best possible subtraction, i.e., the smallest signal coming out of the subtracting amplifier.
What Walker found was that his best amp designs were already so good that this difference signal was below the noise floor of the amp - it was not even measurable, never mind audible. In later years, both noise floor and difference signal dropped to where they were less than the hiss of air molecules bouncing off human eardrums - less than the shot noise of air in the room!
(Walker also found many early amp designs that failed this test - for example, later solid-state amps that ran the speaker signal through an electrolytic capacitor, failed the test.)
I don't recall exactly when Walker began using this test method, but I think it was already in place by 1960, maybe earlier. That would be some sixty years ago, now.
And the audio and Hi-Fi magazines? They NEVER used Walker's method. How can you sell magazines if every amp you test in your "shootout" is perfect? So they kept using sine waves and square waves and completely worthless "golden ear" listening tests. That way, you can always have a "winner", and people will continue to buy your worthless magazine.
Amazingly, in all the years I've been tinkering with audio, I have only ever encountered one other person who has even heard of this very simple - and incredibly effective - test method. That person is Rod Elliott, whose name and website is probably familiar to everyone reading this thread.
Somewhere on Elliott's website, he has a description and schematic you can build, and use to perform Peter Walker's brilliant amplifier test.
Because today's audio amplifiers are invariably direct-coupled except for maybe one cap at the input, the phase errors of 1955 are mostly gone, so the test methodology is even simpler.
Sadly, because amplifier designers and magazine testers alike have ignore this simple and bullet-proof testing method for sixty years, we probably still have some bad amplifiers out there - for example, ones with marginal stability, or over-protective protection circuits.
Not every amplifier is perfect. Sadly, many of today's very expensive tube Hi-Fi amps would fail this test badly (while most cheap solid-state amps would pass it.) And if the tube-amp does pass Walker's test, it will sound exactly like the solid-state amp: there will be no "tube sound", because the amp is audibly flawless, and a flawless amp has no sound of its own.
So here we are, sixty years on, we have worse amps than the ones Peter Walker was building in the '50s and '60s, and people are paying far more for them. That's what happens when designers and consumers ignore science, and turn to superstition instead.
So now you can buy an overpriced, poor-quality tube amplifier, then scatter amazing magic-unicorn-pixie-dust Mpingo Discs around the room to make the sound better than perfect: The Magic of Mpingo . Who needs science when you can have magic instead?
-Gnobuddy
nobody plays metal with their volume knob halve way. Typical Gibson humbuckers have about 8K real impedance, so you can expect about 10nV noise density.
Shorting the 10k input resistor might lower the noise about 3dB.. nothing you can do beyond that, except practice on guitar pick atack, and uae a noise gate.
Do not mistake DC resistance with real part of humbucker impedance. It is not 10kOhm. If you measure an impedance peak of 500kOhm at resonance, 500kOhm is your real part.
Without that impedance would be infinite. .
Remember, half-resistance is not half-way down on the volume knob - we're talking about a logarithmic pot.
The typical "10% taper" log pot has one-tenth resistance at half rotation. Half-resistance occurs at about 85% rotation, or less than one-sixth of the way down from maximum. (See attached LTSpice simulation of one linear and one log pot.)
ITPhoenix - who started this thread - said in one of his posts here that he uses his guitar volume control when playing metal, in order to create playing dynamics. (I think that means he plays a less extreme form of metal.)
-Gnobuddy
Which is why I quoted a typical range of DC resistance, and didn't quote reactive impedance.
Some pickups designed for metal are heavily overwound compared to classic Gibson humbuckers from the 1950s - I've seen 18k humbuckers advertised by well-known aftermarket pickup manufacturers, and some of the cheap pickups from China are also in this resistance range.
Pickups like this have a very dull sound when plugged into a clean guitar amp, but that's not their intended usage.
-Gnobuddy
Attachments
My error, I meant "dynamic range", i.e., varying degrees of amplitude, but in my mind extends to tempo, technique, timbre changes, etc. Off the cuff, in rock: Led Zeppelin; in metal: Metallica. Toccata and Fugue in D Minor when played correctly would necessitate all this for few listeners would be able to last the whole 15 minutes, in my opinion.
Again off the cuff, try "Java" by Al Hirt. For him, the volume pot was his breath control. YouTube
My compressor operator could not sleep at night shift if the compressors are not running. He becomes uneasy. So is the chain saw music for some people. When music stops, they become uneasy and wait for the next chain saw song anxiously. After injecting chain saw sound in to their brain for an hour they are happy. Some compare this chain saw is not like the other chain saw I heard. So what is the fuss about some fuzz making more noise in to already chain saw noise? Just make it as it is. It might click and becomes popular as new noise. They will debate.. this chain saw adds natural sounds also and is better.
It is a Resonant Peak. AT resonance it is a "pure" resistance. And for typical g-pickups, it happens where the ear is extra sensitive.
I knew that (zero phase shift on resonance), but it never dawned on me that this would have consequences on thermal noise.
Thanks!
-Gnobuddy
Honestly, I don't think you were in error in any way - I say playing dynamics, you say dynamic range, I say potayto, you say potahto...
I'm completely unfamilar with Zakk Wylde's body of work, but I stumbled across his "Farewell Ballad" instrumental some years ago. It's an excellent demonstration of how a master guitarist uses the guitar volume control.
Watch as Wylde repeatedly makes tiny, almost imperceptible changes to his volume control every few seconds, gradually increasing the amount of distortion and emotional intensity in his guitar sound as the piece progresses towards its climax: YouTube
This sort of playing dynamics has almost become a lost art in popular music, thanks to the "loudness wars", canned drum-tracks and bass riffs, heavy use of compression, and cut-n-paste music "recording".
I play regularly with other amateur musicians, and playing dynamics are not a part of anybody's else's repertoire. I struggle to remind myself to incorporate them in my own playing.
Some forms of metal use so much gain that turning down the guitar volume knob won't make any perceptible change in loudness. Metal and the more extreme forms of rock aren't really part of my normal music listening, but the harsh power-tool guitar sounds in Evanescence circa early 2000's - say in "Bring Me To Life" - are an example: YouTube
At least the singers can actually sing well, and the overall piece has playing dynamics, even though the guitar parts don't.
-Gnobuddy
It is possible, even with those kind of setups, with lots of practice, to use only technique, like palm muting and varying the force the strings are struck. Personally, I prefer to get as close as possible live to what would be released on media, which is another reason I hate the hiss.
There is one female rocker that mounted a thin, camel hair brush just ahead of the bridge so noise from the unwanted, vibrating strings are silenced, yet does not adversely affect the strings played. Some of these vibrations are caused by __________ excitation in addition to direct vibration caused by the player (I forgot what goes in the blank).
This thread went off topic a few times but the content is interesting, and could help someone come up with a new idea, or one that he/she has not discovered yet, short of submersing the amp in a cryogenic bath.
Last edited:
"people"----WHAT people?? Doctors? Shoe salesmen? Lab technicians? Audiophile shop clerks? Musicians? Schoolteachers? Who? Are you saying that EVERYBODY's ears are the same? Some "people" simply have better ears than others, and can hear things that others can't. Or don't. And just because a device may have inaudible THD, that doesn't mean that there aren't OTHER deficiencies in an amplifier that do make them sound different. I once voiced a control room at a prominent studio to be FLAT using pink noise as a source. Then the mixing engineer ( a true "golden ears" guy who had engineered many hit records of very successful artists) came in and listened; first to music and then (to my surprise) actually listened to the pink noise. He said that he felt there was a small dip in the response at about 6 KHz. I re-hooked my voicing gear and re-measured. Sure enough, there showed a 1 db dip at 6.3KHz on my real-time spectrum analyzer. I corrected it and he was satisfied. I was stunned. This guy really COULD hear a LOT better than I. Not all ears are equal.
I do very much like your idea of the Peter Walker test; it allows use of real music through an amplifier hooked to a real speaker to measure. I am wondering if he was able to put a number on it; after all, if an amplifier has the 0.1% THD, there must be SOME difference between input and output signals that can be quantified with the Walker test. I couldn't find the article(s) you referenced on the Elliott Sound Products website, so I wrote to Rod and asked him. I am still awaiting a reply.
You can simplify this differentials measuring setup in case your test amp is inverting: The residual at the input is your error signal. This can be amplified with any hi-gain hi-input impedance amp.
That argument is overstressed here many times. I agree that there are people who have trained their audio perception to a very high degree - but there are physical limits to us nobody jumps over. Or can you name any person on this globe who runs the 100m distance in 5sec? But there are some who achieve 10sec. Maybe you do it in 20sec? All in all, this is not a big range of numbers that holds for the entire human race. This is one example of a relevant, real number describing our limits. As is 20kHz for the audible band.
I put an opamp as the first stage in a guitar amp to lower the noise of the 12AX7 first stage. It worked with a gain of 2.5 and reducing the common 68k RF resistor to 10k. I rolled amps and noticed the difference between the TL072 and the NE5532. The 072 was chucked immediately is sounded so bad--to me.Well, I , for one, have noticed an IMMEDIATE and UNDENIABLE improvement in the sound of MY system after applying copious amounts of SNAKE OIL to my ears! You SIMPLY must TRY IT!!
Also, I could not notice any difference between PP and PET capacitors in the guitar amp.
I did an online frequency test and heard until just before 18kHz. I also read a paper that claimed the people who invariably listen to mp3 music media through $20 earbud amps do not appreciate CD quality recordings through high end audio, and may actually sound worse to them.
An associate of mine had a halfway decent stereo system. He grew tired of it over the years and upgraded to one 3 times the cost. This required him to roll and modify speakers since the sound was not quite what he wanted.
There seems to be the added dimensions of psychology and physiology involved in sound perception.
Thanks PRR for, "It is a Resonant Peak. AT resonance it is a "pure" resistance. And for typical g-pickups, it happens where the ear is extra sensitive," I would never have found that I don't think.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.