As someone who would love to own the full set but keeps getting waylaid with other must haves from the disposablte income fund, the thing that makes LA interesting for me is/was that the writers were given a fairly free reign to do what interested them, even if the end result was something no one would actually build (well on here we have people who will build pretty much anything). New boundaries were pushed and new things were learned and in some cases a surprising number of units have been built. But most of these are very advanced stuff. [/quote]
I have a 1965 Radford STA35 which is pretty much as good as you need within it's power envelope. Odd for a tube amp in that it uses 35dB of NFB.
You are missing a key part of the Psyche of Engineering/Physics types. They like to do things for the intellectual challenge even if they know there is no audible point. For example DF96 uses home built tube amps. And doing stuff for sh*t and giggles or for discovery doesn't cause any raised eyebrows. It's people who have either an agenda (e.g. 8 legs bad, nfb bad) or are convinced that they can hear the difference between components when totally sighted and with no controls.
Thanks Marcel, Jan and Bill for the replies. I appreciate the transparency and information.
As I said, nothing but goodwill.
But I think you all can see and lend some credence to where I’m coming from here.
I don’t “trust my ears”. But then again, I don’t have anything else to trust either because I can’t trust data I can’t fully comprehend / interpret. So, the safe bet given the circumstances I suppose is to trust nothing.
I’m also not defending that camp. You all expressed measured frustration and I see it thread in and thread out of despair with this kind of building a bridge between EE and hobbyist.
I’m simply suggesting that the communication breakdown isn’t what is normally barked down the party line.
As a non-EE, fresh on the shores how would you interpret the couple hundred plus amplifier designs in the forum and thousands of pages spilled on the same topic?
To me, an intellectual pursuit usually has a greater utility or purpose in its crosshairs. Without that it’s much more like a sudoku puzzle- entertaining, but self-serving.
If I was in a car forum and saw a couple hundred different engine designs I’d make assumptions along similar lines as I did here in the beginning, not knowing any better.
It does not help that many of the more “EE approved” designs around here aren’t very novice friendly..... ive found myself doing archeological excavation level digging for parts and PCBs.. many times leaving empty handed. Also, easily digestible, well organized documentation doesn’t seem to be an EE strong suit. Many in the beginning want a friendly kit with a nice fat pdf. There are of course exceptions to all of this, but that’s my take.
I don’t think the general population here is as stupid or gullible as is commonly suggested, we are just trained in widely disparate fields.
Again, you might feel this is totally not the case, but you do come to the table from a very different universe than the average hobbyist.
Given this, in my mind it’s no wonder people buy the snake oil and keep coming back for more. It’s easier and they don’t know any better. EE’s are allergic to marketing ( and sometimes people skills! ;-) ) and so the cycle persists.
I don’t think it’s stupidity to assume that thousands and thousands of pages written on amplifier topologies would suggest that they all sound the same as long as they weren’t broken and that no real improvements have been made in multiple decades.
In my mind, I would consider those continuing to write articles about amplifiers very misguided? Again, not a dig... I’m genuinely perplexed as in many other instances this would be viewed as pathological behavior to reassemble the same thing in different ways ad infinitum with no improvement?
If you had some EE’s do an LA for dummies with extremely practical general advice in a welcoming tone I think it would go a long way. Doubly so if there were kits available so people could see the results in action.
You can put me down for a dozen, in advance.
I see many who have gotten so frustrated with folk engineering logic that I think they may have lost their perspective.
In short: my suggestion would be rather than use logic that will sail right over their heads, lessen the price of admission to sensible behavior and make it an attractive and simple option for the weekend hobbyist. Indoctrinate them thoroughly and early and on their own playing field, just like how it works in every other effective social system.
If not I think it will be more of same.
As I said, nothing but goodwill.
But I think you all can see and lend some credence to where I’m coming from here.
I don’t “trust my ears”. But then again, I don’t have anything else to trust either because I can’t trust data I can’t fully comprehend / interpret. So, the safe bet given the circumstances I suppose is to trust nothing.
I’m also not defending that camp. You all expressed measured frustration and I see it thread in and thread out of despair with this kind of building a bridge between EE and hobbyist.
I’m simply suggesting that the communication breakdown isn’t what is normally barked down the party line.
As a non-EE, fresh on the shores how would you interpret the couple hundred plus amplifier designs in the forum and thousands of pages spilled on the same topic?
To me, an intellectual pursuit usually has a greater utility or purpose in its crosshairs. Without that it’s much more like a sudoku puzzle- entertaining, but self-serving.
If I was in a car forum and saw a couple hundred different engine designs I’d make assumptions along similar lines as I did here in the beginning, not knowing any better.
It does not help that many of the more “EE approved” designs around here aren’t very novice friendly..... ive found myself doing archeological excavation level digging for parts and PCBs.. many times leaving empty handed. Also, easily digestible, well organized documentation doesn’t seem to be an EE strong suit. Many in the beginning want a friendly kit with a nice fat pdf. There are of course exceptions to all of this, but that’s my take.
I don’t think the general population here is as stupid or gullible as is commonly suggested, we are just trained in widely disparate fields.
Again, you might feel this is totally not the case, but you do come to the table from a very different universe than the average hobbyist.
Given this, in my mind it’s no wonder people buy the snake oil and keep coming back for more. It’s easier and they don’t know any better. EE’s are allergic to marketing ( and sometimes people skills! ;-) ) and so the cycle persists.
I don’t think it’s stupidity to assume that thousands and thousands of pages written on amplifier topologies would suggest that they all sound the same as long as they weren’t broken and that no real improvements have been made in multiple decades.
In my mind, I would consider those continuing to write articles about amplifiers very misguided? Again, not a dig... I’m genuinely perplexed as in many other instances this would be viewed as pathological behavior to reassemble the same thing in different ways ad infinitum with no improvement?
If you had some EE’s do an LA for dummies with extremely practical general advice in a welcoming tone I think it would go a long way. Doubly so if there were kits available so people could see the results in action.
You can put me down for a dozen, in advance.
I see many who have gotten so frustrated with folk engineering logic that I think they may have lost their perspective.
In short: my suggestion would be rather than use logic that will sail right over their heads, lessen the price of admission to sensible behavior and make it an attractive and simple option for the weekend hobbyist. Indoctrinate them thoroughly and early and on their own playing field, just like how it works in every other effective social system.
If not I think it will be more of same.
Last edited:
An intellectual pursuit can also be a hobby, especially in some cases where circumstances mean that not enough solder sniffing can be done in the day job. It's good for you. Keeps mind active, gives satisfaction and might even stave off issues in older age.
DIYaudio is quite clear on the title banner. 'By fanatics for fanatics'.
.
Now if you want newbie friendly designs that will sound great, there are plenty of those around. B1 preamp (boards available from the store) and one of the LM3886 amps will get you something exceedingly good for not much outlay and no hand made by virgins capacitors need get injured in the process.
More power, build a honey badger (again available from store here). Safe supported solution.
Or you can go off piste as many do
But recipes for happiness are doomed to fail in the same way that active speakers are not popular amongst audiophiles (and esp audiophile reviewers).
DIYaudio is quite clear on the title banner. 'By fanatics for fanatics'.
. Now if you want newbie friendly designs that will sound great, there are plenty of those around. B1 preamp (boards available from the store) and one of the LM3886 amps will get you something exceedingly good for not much outlay and no hand made by virgins capacitors need get injured in the process
.More power, build a honey badger (again available from store here). Safe supported solution.
Or you can go off piste as many do
But recipes for happiness are doomed to fail in the same way that active speakers are not popular amongst audiophiles (and esp audiophile reviewers).
On the contrary, there is a level of feedback at which all competently-designed amplifiers become virtually indistinguishable, because they all damage the sound so slightly that the damage is inaudible.gabdx said:
My understanding is that we can't "all" hear the differences between opamps, and those differences are quite small - which is why we can't all hear them. Don't exaggerate a tiny effect. Some opamps suffer from common-mode effects in the non-inverting configuration which no amount of feedback will remove.
The reason for this is quite simple: existing tests show that amplifiers are (or can be) about as good as they need to be, while speakers are still a long way off from this. Hence the genuine research effort is concentrated on the topic which most needs research. Only in the confused brains of audiophiles do genuine hi-fi amplifiers possess night and day faults.oivavoi said:
Good post Spaceistheplace, but has a lot of angles and I'm not inclined to exhaustively address all of them.
Just a couple of comments: Doing the same thing over and over with the same result, yes that would be senseless. But 'the result' is not necessarily sound quality. Example: the millions upon millions poured into class D. Does that lead to much better sound, compared to other classes? Some say it does, some say it doesn't. Probably you wouldn't notice it if your great AB amp is changed for a class D if nobody told you. But boy, has class D advantages, and is it worthwhile to put effort into it!
Communication between novices and old hands. That is indeed something that often is not handled well here. If someone expresses confusion between series and parallel connection of resistors, and you start to explain parasitic capacitances and voltage coefficient, you've lost him. Explain instead that series means that both have the same current, and parallel means both have the same voltage, and in either case Mr. Ohm determines the voltage and current on the part.
Unfortunately, very often simple questions are answered by inundating the helpless victim with very advanced stuff (which even is often wrong), apparently in an effort to establish someone's great understanding. Worse than no answer at all.
OTOH, if you are a novice and really interested in stuff, there's no way around putting in some hours and legwork yourself, and that often is also sorely missing. If you just look for free info, that's what it's worth to you in the longer run.
Jan
Just a couple of comments: Doing the same thing over and over with the same result, yes that would be senseless. But 'the result' is not necessarily sound quality. Example: the millions upon millions poured into class D. Does that lead to much better sound, compared to other classes? Some say it does, some say it doesn't. Probably you wouldn't notice it if your great AB amp is changed for a class D if nobody told you. But boy, has class D advantages, and is it worthwhile to put effort into it!
Communication between novices and old hands. That is indeed something that often is not handled well here. If someone expresses confusion between series and parallel connection of resistors, and you start to explain parasitic capacitances and voltage coefficient, you've lost him. Explain instead that series means that both have the same current, and parallel means both have the same voltage, and in either case Mr. Ohm determines the voltage and current on the part.
Unfortunately, very often simple questions are answered by inundating the helpless victim with very advanced stuff (which even is often wrong), apparently in an effort to establish someone's great understanding. Worse than no answer at all.
OTOH, if you are a novice and really interested in stuff, there's no way around putting in some hours and legwork yourself, and that often is also sorely missing. If you just look for free info, that's what it's worth to you in the longer run.
Jan
Last edited:
I can only answer for myself, but I enjoy designing and building electronics (both audio and radio). I understand electronics. I don't understand speakers, at least not at the level where I can design something good enough to satisfy me.spaceistheplace said:
1. Yesgabdx said:
2. Not necessarily.
3. Yes - that is why feedback cannot usually be applied over more than about three stages
4. Not sure what you are saying
5. no idea what you are saying
6. no idea what you are saying
You seem to have confused yourself. It is true that the closer you get to perfection the harder it gets to make further improvements. That is not an argument for not even attempting to get near perfection. The same rules of nature apply to audio as to the much more demanding applications you mention.
No, this is not the truth. Feedback largely removes from the output anything which is not in the input. In doing this it unavoidably adds a bit more re-entrant distortion to the output, but this then is also reduced by the feedback. Feedback does not single out 2nd for special attention; it just happens that in most systems 2nd already dominates so the feedback has more 2nd to remove than other distortions. Feedback does not "destroy" low level signals; on the contrary, when signal levels are low and so distortion is low feedback works at its best.
No, faithfull reproduction is the goal (for me, and some others here). That may sound "dead" to someone who has other goals. As usual, THD is used as an accusation, as though we are all THD fans. An amp does not know that it has to "track multiple signals"; it just sees a voltage. Only Maxwell demons and Fourier deniers have to track multiple signals. Poorly designed amps allow the PSU to be heard; I guess a really bad PSU might even break through a good amp.
Thank you for trotting out all the usual 'true believer' confusions and misunderstandings.
The disputes here are not between EEs and hobbyists. They are between those who understand audio electronics (including feedback, Fourier etc.) and those who do not. Some of the former are hobbyists (some of which may also be EEs); some of the latter are professionals.spaceistheplace said:
Thanks for taking the time to comment on that post. I saw so much wrong I couldn't muster the courage.
Jan
This is the general consensus of the "engineer" here as a story:
1. build amp
2. sounds bad, high distortion (thd 2% whatever)
3. I will add more gain, lets add another gain stage, or use higher gain devices.
4. GNF, yes! now the amp performs perfectly, 0.001% because I used a lot of it.
5. Wow all those amps sound the same.
DF96, feedback attacks differently each harmonics generated. The formula is not the same for 2n, 3rd, etc.
And the correction removes 2n harmonics not from the music signal as you interpreted my text, it removes 2nd of the amp distortions, as it does, it introduces 3rd harmonics which are killing music details.
As for signals, there are many signals, noise signals, feedback signal and the node input signal.
1. build amp
2. sounds bad, high distortion (thd 2% whatever)
3. I will add more gain, lets add another gain stage, or use higher gain devices.
4. GNF, yes! now the amp performs perfectly, 0.001% because I used a lot of it.
5. Wow all those amps sound the same.
DF96, feedback attacks differently each harmonics generated. The formula is not the same for 2n, 3rd, etc.
And the correction removes 2n harmonics not from the music signal as you interpreted my text, it removes 2nd of the amp distortions, as it does, it introduces 3rd harmonics which are killing music details.
As for signals, there are many signals, noise signals, feedback signal and the node input signal.
You have your definition of an amplifier, it's just not a very accurate one, no problem
This is all so not true, it isn't even wrong. Even the first step is wrong. You left out 'design' which should be the first step in any sensible undertaking.
Everything after that is downhill.
Jan
No. Have you ever actually read a textbook?gabdx said:
No, I did not misunderstand you. The 2nd removed is 2nd introduced by the amp. The 3rd created is small in comparison, and is itself reduced by the feedback. It does not kill music details; however, distortion in an amp without feedback can artificially create 'detail' which is not present in the signal and some people confuse this with genuine detail and so miss it when it is absent.
You are just playing with words.
Your confusion is so profound and so boldly proclaimed that I am beginning to suspect that you may be commercially involved with audio.
where the field is light is psycho-acoustics knowledge. Once one gets exposed to it one gets humbled and starts to moderate their engineering deduction (and this is coming from an engineer).
In the last ten years it has been shown that when it comes to tonal noises the brain is all about looking for patterns. In one study for example a group of harmonics was played and one harmonic out of the group was turned on and off while subjects always heard it, (basically the brain perception fills in the missing harmonic which physically is not even in the air).
it was also shown that if there is a periodic structure to a tonal group, the brain can detect it even if it is buried 3dB under the noise floor (pretty amazing eh). I have access to the best microphones ever made and they still do not match the dynamic range of the human hearing (ear drum to neural processing outcome), and occasionally I get to work on a sound quality issue where we can hear it but it takes days of flipping data on its head to see anything (and this in software for which maintenance fee is 20k+/yr).
The people on the committee writing the ECMA-74 standard for audio testing (a precursor of ISO), only got to a point where they show that "landscaping" a spectrum with gentle slopes has an effect on perceived tonality since even a wide band of say 1kHz can have a tonal quality if it rises suddenly against the background; (and they made a ppt presentation with embedded audio clips to demonstrate it).
Moreover, as some one already hinted here, the goal(s) behind this hobby are not always the same. For me (similar to Linkwitz and Pass and many others) it goes as follows: I accept that the recording is some random mixing outcome made by a non-standardized method and to somebody's individual taste, while using a different equipment to play back the sound and create a pleasant (to them) experience. The origin of the signal is not even the band playing together; the chances are all instruments and the voice(s) were recorded one at a time in a room with tbd reverb, mic position etc.
Now my goal is to put together a set of equipment which can create a listening experience which sounds very natural, pleasant and engaging. I may use any means available in the process. Starting with the fact that people listen with two ears with simultaneous processing of the two incoming streams, it turns out that creating a sound field by managing dispersion pattern of the speaker drivers, maintaining the same tonal balance off axis while avoiding early reflections, using amps with a particular distortion pattern, etc etc (all the things I learnt from others and then experimented with) is the best way to accomplish this. And one will never have the best outcome across the board for all the different music material as there is no standard way of creating it to begin with.
as you can see in this picture, a published single number for one, wanna be measure of an amplifier performance (THD, THD+N or whatever) is completely irrelevant and judging by the brain's supercomputing capabilities inadequate.
In the last ten years it has been shown that when it comes to tonal noises the brain is all about looking for patterns. In one study for example a group of harmonics was played and one harmonic out of the group was turned on and off while subjects always heard it, (basically the brain perception fills in the missing harmonic which physically is not even in the air).
it was also shown that if there is a periodic structure to a tonal group, the brain can detect it even if it is buried 3dB under the noise floor (pretty amazing eh). I have access to the best microphones ever made and they still do not match the dynamic range of the human hearing (ear drum to neural processing outcome), and occasionally I get to work on a sound quality issue where we can hear it but it takes days of flipping data on its head to see anything (and this in software for which maintenance fee is 20k+/yr).
The people on the committee writing the ECMA-74 standard for audio testing (a precursor of ISO), only got to a point where they show that "landscaping" a spectrum with gentle slopes has an effect on perceived tonality since even a wide band of say 1kHz can have a tonal quality if it rises suddenly against the background; (and they made a ppt presentation with embedded audio clips to demonstrate it).
Moreover, as some one already hinted here, the goal(s) behind this hobby are not always the same. For me (similar to Linkwitz and Pass and many others) it goes as follows: I accept that the recording is some random mixing outcome made by a non-standardized method and to somebody's individual taste, while using a different equipment to play back the sound and create a pleasant (to them) experience. The origin of the signal is not even the band playing together; the chances are all instruments and the voice(s) were recorded one at a time in a room with tbd reverb, mic position etc.
Now my goal is to put together a set of equipment which can create a listening experience which sounds very natural, pleasant and engaging. I may use any means available in the process. Starting with the fact that people listen with two ears with simultaneous processing of the two incoming streams, it turns out that creating a sound field by managing dispersion pattern of the speaker drivers, maintaining the same tonal balance off axis while avoiding early reflections, using amps with a particular distortion pattern, etc etc (all the things I learnt from others and then experimented with) is the best way to accomplish this. And one will never have the best outcome across the board for all the different music material as there is no standard way of creating it to begin with.
as you can see in this picture, a published single number for one, wanna be measure of an amplifier performance (THD, THD+N or whatever) is completely irrelevant and judging by the brain's supercomputing capabilities inadequate.
Last edited:
I prefer music obtained from a genuine performance. I recognise that this is getting rarer.koja said:
It is unclear what can be meant by "natural" when the sound never existed. My requirement is different: I want reproduction which sounds like what I would hear in a concert hall. I realise that this is an illusion and cannot be fully achieved, yet I find that transparent (what some have called "dead") systems get closest to this goal.
I have yet to meet anyone who genuinely believes that a single number can characterise all aspects of an amplifier's performance, yet I frequently come across people who accuse others of believing this. I believe this can be called 'tilting at windmills' or 'erecting Aunt Sallys'.
Fully agree with your post (snipped to conserve bandwidth ;-).
Some of this knowledge is used to advantage in the industry. For example, there were at one time Bose products (don't know if it is still the case) where they examined the spectral content of the signal and if it had low bass, that the build-in speaker could not reproduce, they synthesized a couple of harmonics of that tone and what do you say, the bass tone was clearly heard!
Even common stereo is nothing more than a cunning illusion. Secret tip: there is no singer between the speakers, it's all make-believe ;-).
This fact lead the great Richard Heyser to the statement that an audio engineer and a magician are in the same line of business: creating illusions!
Jan
No. It is easy to make amplifier with global negative feedback that have 2nd harmonic distortion dominant or 3rd harmonic distortion dominant.
I wonder whether some find the reality even more unbelievable than what the snakeoil vendors pedal, for example, it does seem magical to me that so much information is packed into a single varying voltage. And then how it can be dissected and made sense of. It almost takes a leap of faith to accept it, it certainly takes the will to want to and maybe the audio mystics are afraid to even try because they fear the magical power will be lost. For me, the opposite is true, I'm even more in wonder.
+1.
Part of the hurdle I think is that it takes a sizable mental jump to recognize that intuition is often right only in the most simple and straightforward cases.
Once you go into sufficient depth and complexity, you find that our precious intuition is mostly wrong. That can be unsettling.
I mean, whom can you trust if not yourself? Yet...
Jan
Part of the hurdle I think is that it takes a sizable mental jump to recognize that intuition is often right only in the most simple and straightforward cases.
Once you go into sufficient depth and complexity, you find that our precious intuition is mostly wrong. That can be unsettling.
I mean, whom can you trust if not yourself? Yet...
Jan
an exotic opinion
You may be served the specialty of a 3 star chef , by the chef himself , he will provide you also salt and pepper. The 2A3 SE amplifier-user enjoys 5% distortion at 2.5W . The Black Ortophone provides on its vertical mouvments, a generous 3.5% distortion to enjoy the innocent vinyl-lover. The low distorting AR3 did satisfy the most demanding ears of H. V. Karayan with only 2% distorted sound . A golden ear submariner cannot discern down to 0.1% distortion. Since Matusalem, Williamson is delivering with 20db NFB, 0.1% F-3 100khz.
An amplifier is a voltage source . It has a voltage generator and an internal impedance . A loudspeaker is a giant microphone with a series resistance . The resistance linearizes the currents both of the amplifier who's impedance is much lower and the speaker's ,while damping . Although the speaker is a voltage generator ,I don't know how ,somtimes it behaves as a current generator . The amplifier than answers the current by a voltage in accordance to its open loop tranconductance transfer function . The Williamson will provide not 0.1% but with 1% distortion with resistive component up to 10khz . If an amplifier lambda with identical performance but with 40db NFB had been used ,it would be 10% with 1khz.
Kokoriantz
---------------------------------------------------------------------------------------------------------------------------
Lebanees adage : "Amel kheir ou zettou al baher". Act in goodness and throw it into the sea.
You may be served the specialty of a 3 star chef , by the chef himself , he will provide you also salt and pepper. The 2A3 SE amplifier-user enjoys 5% distortion at 2.5W . The Black Ortophone provides on its vertical mouvments, a generous 3.5% distortion to enjoy the innocent vinyl-lover. The low distorting AR3 did satisfy the most demanding ears of H. V. Karayan with only 2% distorted sound . A golden ear submariner cannot discern down to 0.1% distortion. Since Matusalem, Williamson is delivering with 20db NFB, 0.1% F-3 100khz.
An amplifier is a voltage source . It has a voltage generator and an internal impedance . A loudspeaker is a giant microphone with a series resistance . The resistance linearizes the currents both of the amplifier who's impedance is much lower and the speaker's ,while damping . Although the speaker is a voltage generator ,I don't know how ,somtimes it behaves as a current generator . The amplifier than answers the current by a voltage in accordance to its open loop tranconductance transfer function . The Williamson will provide not 0.1% but with 1% distortion with resistive component up to 10khz . If an amplifier lambda with identical performance but with 40db NFB had been used ,it would be 10% with 1khz.
Kokoriantz
---------------------------------------------------------------------------------------------------------------------------
Lebanees adage : "Amel kheir ou zettou al baher". Act in goodness and throw it into the sea.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.