Hi jcx,
Thank you! I can always count on you to liberate a smile from me.
BTW, I got me some of those radios. Some even older than that.
Hi Anatoliy,
Hi a.wayne,
I have an extremely strong dislike for designers who are more ego than substance, and there is no shortage of those people out there. I see expensive poor designs as outright theft. There is no sugar coating that one. You know, I do respect equipment that is what it is advertised as. Just as I respect really good designs as well. I think it comes down to honesty and getting what you pay for.
Do you know that equipment that forgoes any kind of protection network for the speakers really do create fires. Most speakers burn (big surprise). The callus disregard some people have for others just so they can save a few bucks on missing protective circuitry garner my intense dislike. There are technical ways to accomplish speaker protection without using relays or fuses if they must. A Carver amplifier represents an excellent example of this.
-Chris
Thank you! I can always count on you to liberate a smile from me.
BTW, I got me some of those radios. Some even older than that.
Hi Anatoliy,
Well, as I said. It only matters when you approach the limits. In this case, maybe at 1/10 of the frequency where you begin to run into slew rate limits. If the amp is not being pushed, it will not matter. There are many asymmetrical responses in nature, and certainly in power amp designs. I think that making an effort to create symmetry in a design may require a compromise in more sensitive areas. It's really up to how you choose to design a circuit. It might be easier to stick with op amps selected for symmetrical behaviour if this is very important to you.
Hi a.wayne,
Bright eyed optimism can't survive when you have worked in audio service for 35+ years. Believe me when I say, there are more aborted audio efforts produced by small high end companies than you really want to know about. Consider that I see equipment laid bare. I see the construction, the schematic (design) and components used. Then, it gets heard and measured. Over the years you get to see the weaknesses and design snafus, plus the rhetoric used to explain things away. That and the pure make-believe. I have stronger words for that, but this is a family site.
I have an extremely strong dislike for designers who are more ego than substance, and there is no shortage of those people out there. I see expensive poor designs as outright theft. There is no sugar coating that one. You know, I do respect equipment that is what it is advertised as. Just as I respect really good designs as well. I think it comes down to honesty and getting what you pay for.
Do you know that equipment that forgoes any kind of protection network for the speakers really do create fires. Most speakers burn (big surprise). The callus disregard some people have for others just so they can save a few bucks on missing protective circuitry garner my intense dislike. There are technical ways to accomplish speaker protection without using relays or fuses if they must. A Carver amplifier represents an excellent example of this.
Oh, absolutely! However, engineers each have things they do better than others. They have a knack for some circuit types and understanding how each part work. You have a design leader, and individuals that contribute circuits they know best that are integrated properly. There is far too much money at stake to mess up when you create thousands upon thousands of each model. There is no "let's fix it in the field" thinking that is acceptable. So while these products may not represent the pinnacle of audio art, they do represent good value, good performance and a reasonable expectation that they will not burn your speakers (and house) to the ground. We also did insurance inspections and repair. We did see equipment that did nasty things by design. They didn't fail safe, which is an important concept in industrial, military and consumer product development. Amateurs who think they design audio equipment do not generally design in "fail safe" concepts. They are not even aware of it for the most part.
-Chris
Chris;
symmetric slew rate does not mean symmetric topology. And symmetric topology does not guarantee symmetric slew rate. I put in this thread already some pictures with explanation, may be half of hundred pages ago.
Also, Doppler effect, non-linearities: of air, walls, whatever, we don't hear as distortions. We hear them as cues to surrounding. But as soon as it differs a tiny bit our ears pop up: "Was ist das?!"
symmetric slew rate does not mean symmetric topology. And symmetric topology does not guarantee symmetric slew rate. I put in this thread already some pictures with explanation, may be half of hundred pages ago.
Also, Doppler effect, non-linearities: of air, walls, whatever, we don't hear as distortions. We hear them as cues to surrounding. But as soon as it differs a tiny bit our ears pop up: "Was ist das?!"
Provided the heat source has constant dissipation....
The variation will be integrated by the thermal inertia , though...
Hello Chris
Other than the standard current limiting circuit ?
Can you say more about that, any practical exemples ?
Thank
Bye
Gaetan
Last edited:
C6 interacts with R3, R4, and the dynamic emitter impedances of Q1 and Q2. A crude calculation suggests the resulting phase lead is maximum around 10MHz.
edit: The effect of R5 is negligible.
Name a high feedback back amplifier considered to be good sounding , contrary name a low or zero feedback amp considered bad sounding . The Audio Note UK amp discussed previously , failed every test known to man , yet it got a pass for sound ...
PS: Please quantify high feedback and low feedback ..........
For the first question, I'll name Laverdine. Their feeling is that feedback is good, and assuming your amp satisfies a list of criteria in open loop, it doesn't matter how much feedback you use. One of the criteria is that the amp's open loop full power bandwidth must be 30 kHz or more.
For the second, please re-read my original post, I was quite specific: up to 20 dB is low, up to 28 dB is medium, and above is much. This is, quite obviously, a very personal classification, no universal truth.
By popular request, I have placed this article online here:
Linear Audio | Online Resources
(2nd article down).
Opinionated perhaps, but Bruno is one who shores up his opinions with facts and figures. If someone would not agree on something with Bruno, he would have to show the flaw in his reasoning or his facts and figures. So far, no callers
jan didden
Last edited:
Rgardin slew rate, my feelings are paired.
One is that it has been accepted that if your slew rate is 0.5 V/uS per each PEAK volt of output power, you should not have problems with that aspect. That would mean, for example, 20 V/uS for a nominally 100W/8 Ohms amplifier.
My personal feeling is that I am far happier with double that, I don't like to just get by, I prefer a nice safety margin. But that's just me.
The other is that we talk of VOLTAGE slew rates, but hardly anybody even mentions CURRENT slew rates, expressed a xx A/uS. What use is an amplifier which can do say 200 V/uS, but only say 5 A/uS, when driving difficult load speakers?
To be fair to the industry at large, capacitor manufacturers consistently fail to declare their products in both of those factors. In my life, I have seen only one manufacturer who did it by the book, that was Siemens regarding their Sikorel range; they stated a speed of 100 V/uS and 10 A /uS.
The point being that if you have electronics capable of 200 V/uS, but capacitors which can do say 40 V/uS, your EFFECTIVE overall slew rate is just 40 V/uS.
So how do we get wild figures like 300 V/uS and such like? We get them because a long time ago, manufacturers decided that the input stage was the determinant - if it can slew fast anough, all our problems are solved. Which is obviously not so, but it looks much better on paper when you use a dual FET, cascoded by some fast trannies, and with a current mirror, and voila! you have the magic numbers. This was started in the late 70ies by Japanese manufacturers, notably Sansui and Kenwood. They even had papers published by the AES on the subject, look it up.
This is, as I see it, diddling the specs, thus giving them even less credibility in real life.
But, there's more. Ever wondered how come an amp with a THD spec of say 0.001% actually sounds worse than another amp rated at say 0.05%? On the surface of it, this is illogical.
But it isn't. What they tell you is what the amp's AVERAGE THD is and they do so by measring it over say 1 hour of steady state state test signal, so as to include the period when it's cold upon switch on and the time it heats up to its normal temperature, when THD will be naturally somewhat lower.
Now, here comes a transient, which causes one of the above example amps, the one with a rating of 0.001%, to pass it on through, but with a THD figure of say 3%. Because the impulse is short, say 100 mS or less, this excursion is buried in the averaging process, but you cannot hide it from human ears. We may not know exactly what's going on, but we can hear that the transient is not what is should be. Whereas if the other amp should happen to pass it on just fine, we will notice by lack of irritation, or even as a more lively sound.
Which is why I declare all my amps with an OVERALL (input to output) slew rate, and of course, have much less impressive figures to read, but much better figures to listen to. And I repeat the process with 8, 6, 4 and 2 Ohm loads, because I have no idea what will eventually be connected as its operating load. I think a realistic figure of smaller absolute value, but applicable under all nominal operating conditions, is a better deal than a more or less theoretical case of a pure lab 8 Ohm resistor. It's certainly more fair.
Just as I give more moderate THD specs than the industry at large, because I give them for PEAK distortion, not average.
Please understand, this not a "hooray for me" post, I'm simply trying to grasp more fully why specifications do a poor job of describing a product's actual, real world performance. Why we have an amp which has out of this world specs on paper, yet a bland sound in situ. God knows I have had that experience many a time, and I have no doubt you have as well.
One more thing. Over the decades, I have had to replace quite a few capacitors. My experience has taught me that when I replace highly regarded caps from Japan by less well known caps from Germany, e.g. Fisher & Tausche, I invariably get better to much better bass, and I get more life from the device. So, in the end, I started to seriously experiment with this, and to cut a long story short, it was confirmed literally EACH AND EVERY time. I realize it's not polite to generalize, but I cannot avoid it, after so many times of confirming it. The only caveat in that is that I have not tried every cap by every manufacturer, just those built in by the manufacturers themselves, which is mostly ELNA and and some Nichicon.
A short test of the amp's reaction to 2 Ohm loads showed a much better response from F&T caps than from Japanese caps. Now, where does this take you? Perhaps to the subject of current slew rate?
The joke is that I have loudspeakers (locally made as per some of my ideas) which are, to the best of my knowledege, the cleanest and most easy to drive dynamic speakers ever. Nominal impedance is 8 Ohms, minimum 6.5 Ohms, worst case phase shift -25 degrees, efficiency 92 dB/2.83V/1m, 3 way speaker, all drivers by Son Audax. Which means that even the most lowly of amps will drive them with ease, as well every tube SET ever connected to it. But not everybody is that lucky.
Анатолиј, it was a good try, although the original should be made with sliwowitz, plum brandy, preferably domestic rather than industrial. But I noticed you said you are not fond of strong stuff, and I can undersatnd since I am the same. Beer and cider are as far as I will venture, my subsistence being guaranteed by milk and milk products.
One is that it has been accepted that if your slew rate is 0.5 V/uS per each PEAK volt of output power, you should not have problems with that aspect. That would mean, for example, 20 V/uS for a nominally 100W/8 Ohms amplifier.
My personal feeling is that I am far happier with double that, I don't like to just get by, I prefer a nice safety margin. But that's just me.
The other is that we talk of VOLTAGE slew rates, but hardly anybody even mentions CURRENT slew rates, expressed a xx A/uS. What use is an amplifier which can do say 200 V/uS, but only say 5 A/uS, when driving difficult load speakers?
To be fair to the industry at large, capacitor manufacturers consistently fail to declare their products in both of those factors. In my life, I have seen only one manufacturer who did it by the book, that was Siemens regarding their Sikorel range; they stated a speed of 100 V/uS and 10 A /uS.
The point being that if you have electronics capable of 200 V/uS, but capacitors which can do say 40 V/uS, your EFFECTIVE overall slew rate is just 40 V/uS.
So how do we get wild figures like 300 V/uS and such like? We get them because a long time ago, manufacturers decided that the input stage was the determinant - if it can slew fast anough, all our problems are solved. Which is obviously not so, but it looks much better on paper when you use a dual FET, cascoded by some fast trannies, and with a current mirror, and voila! you have the magic numbers. This was started in the late 70ies by Japanese manufacturers, notably Sansui and Kenwood. They even had papers published by the AES on the subject, look it up.
This is, as I see it, diddling the specs, thus giving them even less credibility in real life.
But, there's more. Ever wondered how come an amp with a THD spec of say 0.001% actually sounds worse than another amp rated at say 0.05%? On the surface of it, this is illogical.
But it isn't. What they tell you is what the amp's AVERAGE THD is and they do so by measring it over say 1 hour of steady state state test signal, so as to include the period when it's cold upon switch on and the time it heats up to its normal temperature, when THD will be naturally somewhat lower.
Now, here comes a transient, which causes one of the above example amps, the one with a rating of 0.001%, to pass it on through, but with a THD figure of say 3%. Because the impulse is short, say 100 mS or less, this excursion is buried in the averaging process, but you cannot hide it from human ears. We may not know exactly what's going on, but we can hear that the transient is not what is should be. Whereas if the other amp should happen to pass it on just fine, we will notice by lack of irritation, or even as a more lively sound.
Which is why I declare all my amps with an OVERALL (input to output) slew rate, and of course, have much less impressive figures to read, but much better figures to listen to. And I repeat the process with 8, 6, 4 and 2 Ohm loads, because I have no idea what will eventually be connected as its operating load. I think a realistic figure of smaller absolute value, but applicable under all nominal operating conditions, is a better deal than a more or less theoretical case of a pure lab 8 Ohm resistor. It's certainly more fair.
Just as I give more moderate THD specs than the industry at large, because I give them for PEAK distortion, not average.
Please understand, this not a "hooray for me" post, I'm simply trying to grasp more fully why specifications do a poor job of describing a product's actual, real world performance. Why we have an amp which has out of this world specs on paper, yet a bland sound in situ. God knows I have had that experience many a time, and I have no doubt you have as well.
One more thing. Over the decades, I have had to replace quite a few capacitors. My experience has taught me that when I replace highly regarded caps from Japan by less well known caps from Germany, e.g. Fisher & Tausche, I invariably get better to much better bass, and I get more life from the device. So, in the end, I started to seriously experiment with this, and to cut a long story short, it was confirmed literally EACH AND EVERY time. I realize it's not polite to generalize, but I cannot avoid it, after so many times of confirming it. The only caveat in that is that I have not tried every cap by every manufacturer, just those built in by the manufacturers themselves, which is mostly ELNA and and some Nichicon.
A short test of the amp's reaction to 2 Ohm loads showed a much better response from F&T caps than from Japanese caps. Now, where does this take you? Perhaps to the subject of current slew rate?
The joke is that I have loudspeakers (locally made as per some of my ideas) which are, to the best of my knowledege, the cleanest and most easy to drive dynamic speakers ever. Nominal impedance is 8 Ohms, minimum 6.5 Ohms, worst case phase shift -25 degrees, efficiency 92 dB/2.83V/1m, 3 way speaker, all drivers by Son Audax. Which means that even the most lowly of amps will drive them with ease, as well every tube SET ever connected to it. But not everybody is that lucky.
Анатолиј, it was a good try, although the original should be made with sliwowitz, plum brandy, preferably domestic rather than industrial. But I noticed you said you are not fond of strong stuff, and I can undersatnd since I am the same. Beer and cider are as far as I will venture, my subsistence being guaranteed by milk and milk products.
Just out of interest (and on the theme of this thread), has anyone here ever learned to doubt their own 'hearing' i.e. clearly 'heard' something that they later found to be false? For example, I once 'heard' a very clear difference when comparing two digital sources using headphones, switching between them with the source button on an amplifier. I then found that my headphones were directly plugged into the headphone socket of one of them - not the amplifier...
It's a shame really. If it hadn't been for that incident, I might also have gone through life believing that I had Golden Ears, too.
It's a shame really. If it hadn't been for that incident, I might also have gone through life believing that I had Golden Ears, too.
If you are 'in audio', doing experiments and 'playing' with your equipment, something like that will happen to all of us sooner or later.
But it is easy to 'forget' the experience
jan
There quite likely was a real difference. E.g. changing the source selector was changing whether the device you were listening to was being loaded by the input impedance of the amp.
I go and have my ears analyzed once a year. I like to know what's going on, and it does rid me of some foolish notions.
Last year, my hearing was down to somewhere between 16 and 17 kHz, actually quite good for my age (I'm 58). Ten years ago, it was over 18 kHz. I'm on my way down, and there is no cure. The biological machine is grinding us all down.
What exactly are "Golden ears"? Is this related to the quantity, i.e. the bandwidth of our hearing, or quality? If quality, how do you assess it? Who is qualified to test for them and declare them present? Personal mega egos and financial gain excluded?
True, some people do indeed hear more than others, can pick up details others may pass over, but how sure are we that this not a matter of training, rather than absolute ability?
I have had a professor from the School of Music, University of Belgrade, sit next to me and pick apart the sound of a Krell vs. a Levinson. We ended up disagreeing, he said Levinson was the best he had ever heard, and to me, Levison was just that little bit more detached than Krell. We could both pick out which was playing at any moment with our backs turned to the stage. So, who is right? I don't know, but I still like Krell that little bit more.
I don't think I have anything special about my own hearing, except for one thing - training. I've been in audio since I was 11, since 1966, at the time intersted only in having the latest hit number, and not until 1970 did I start to pay attention not only to what, but also to how. That is 42 years now.
I firmly believe that most people here would do the same, with the same training. Ears are like muscles, exercise them and they will do better than otherwise.
In the end, who cares? Why should you concern yourself with that at all, so long as you truly, honestly enjoy the music you listen to? Why try to introduce absolutes into what is a very personal hobby?
How many times has love for music been judged by what gear you have, and worse, by what kind of music you listen to? I like Bob Seeger, you might not; I like The Marshall Tucker Band, you might think Southern Rock is nonsense. You may like Richard Wagner, I don't even if I draw my roots from Vienna, I like Mozart. I get high on Irish jigs and Enya, you might prefer Metallica. So long as we enjoy what we like, all is well.
I learnt some thing about my hearing many years ago. The first incident was when I was temporarily playing an MM pickup through a tape recorder input. The sound seemed OK, but just a bit toppy. Then I realised that the RIAA was turned off - this input doubled as a flat mike input too. I was surprised that I did not immediately hear and correct such a major frequency response error.
Fast forward some years to when I was developing/debugging a DIY FM tuner. Again it sounded good, but not quite right in the treble. The deemphasis was not quite right. I then rechecked the circuit and realised that I had miscalculated the capacitor values. Trying different values I could clearly hear the difference. We are talking about an HF shelf of around +- 1-2dB above 2-3kHz. The cap value which sounded best also turned out to be the one which was right according to calculation. This was not exactly a blind experiment, but the relationship between cap value and deemphasis rolloff was not so simple that it could have seriously biassed the result - otherwise I would not have made the original error.
So how come I could hear a small shelf, but not notice (at first) a major LF rolloff and HF rise? My conclusion is that my/our sensitivity to error may be logarithmic.
Fast forward some years to when I was developing/debugging a DIY FM tuner. Again it sounded good, but not quite right in the treble. The deemphasis was not quite right. I then rechecked the circuit and realised that I had miscalculated the capacitor values. Trying different values I could clearly hear the difference. We are talking about an HF shelf of around +- 1-2dB above 2-3kHz. The cap value which sounded best also turned out to be the one which was right according to calculation. This was not exactly a blind experiment, but the relationship between cap value and deemphasis rolloff was not so simple that it could have seriously biassed the result - otherwise I would not have made the original error.
So how come I could hear a small shelf, but not notice (at first) a major LF rolloff and HF rise? My conclusion is that my/our sensitivity to error may be logarithmic.
Regarding Golden Ears, my girlfriend has the best ears I know when it comes to low level sounds, very high frequencies (>16 kHz) or following a conversation across the room filled with 100+ people. It's almost scary what she can hear. She definitely has the better sensors. But even after 8 years being exposed to high end audio, when it comes to judging sound quality she is completely useless. She simply does not care, and the same goes for 90% of the people I know. They are all perfectly happy with the performance of their cheap surround sets that make me want to leave the room.
I know the first time when I went to a live classical concert 25 years ago I got tears in my eyes the very first second, and that was when they were still voicing their instruments. To some people sound quality does matter.
So when it comes to golden ears I suspect it's a combination of interrest and experience.
I know the first time when I went to a live classical concert 25 years ago I got tears in my eyes the very first second, and that was when they were still voicing their instruments. To some people sound quality does matter.
So when it comes to golden ears I suspect it's a combination of interrest and experience.
- Status
- Not open for further replies.