Hi thank you very much for the very valuable link
The main problem for me starts at the source and its V level
I listen to various digital sources with different output
No vinyl
Imho a very good volume control followed by a unity gain buffer should be quite doable
This can be achieved with very low THD and good driving ability
When even a small voltage gain is introduced the THD goes up and gets amplified by the power amp
So for me the problem is the line stage
I have to fix that
After that I can pass to the power amp selection
The main problem for me starts at the source and its V level
I listen to various digital sources with different output
No vinyl
Imho a very good volume control followed by a unity gain buffer should be quite doable
This can be achieved with very low THD and good driving ability
When even a small voltage gain is introduced the THD goes up and gets amplified by the power amp
So for me the problem is the line stage
I have to fix that
After that I can pass to the power amp selection
I had a '5G' phone for a while, until it broke and I went back to 4G / LTE, and the difference was curious. I actually found the perceptible radiation to be much lower with 5G. LTE produced a 'burn' sensation like a combination of radiant heat and rheumatism, which was persistent and annoying and put clear limits on the amount of internet time I could tolerate. The 5G phone didn't have that. Maybe it still turns our DNA into obedient nanobots, but for normal daily use it was better.
Hi and good morning to Everyone Just the last rambling and then i will stop on this subject Promised
Let's take a simple circuit that i like aestethically The one attached
Let's say that the goal would be to reach the lowest THD possible for a 3Vout on a 5kohm load (the settings i normally use for sim of buffers)
There are different parameters to "play" with ... npn type, currents, resistor, Vsupply
The possible combinations would be almost infinite
Why almost nobody is interested to assess how this simple circuit could provide in the best conditions i.e. to check its limits ?
Almost everyone would answer ... this simple circuit is clearly limited by its nature
ok but are you really sure of which these limits are ?
I mean before passing to a much more complex design why not investigate the ultimate performance of a simple circuit ?
Lets say that this will provide at max -90dB HD at 2kHz ... then if i want -100db i will pass happily to other designs but with not regrets
Maybe it is my impression but very few designers are stimulated by this challenge ... i.e. to extract the most from the least
You maybe will say that i am obsessed with this thing of low THD Judging from the post i see that i am not alone
Many time projects are pushed just on the basis of their very low THD performance
So my guess is that the interest for this goal is quite popular among designers
Let's take a simple circuit that i like aestethically The one attached
Let's say that the goal would be to reach the lowest THD possible for a 3Vout on a 5kohm load (the settings i normally use for sim of buffers)
There are different parameters to "play" with ... npn type, currents, resistor, Vsupply
The possible combinations would be almost infinite
Why almost nobody is interested to assess how this simple circuit could provide in the best conditions i.e. to check its limits ?
Almost everyone would answer ... this simple circuit is clearly limited by its nature
ok but are you really sure of which these limits are ?
I mean before passing to a much more complex design why not investigate the ultimate performance of a simple circuit ?
Lets say that this will provide at max -90dB HD at 2kHz ... then if i want -100db i will pass happily to other designs but with not regrets
Maybe it is my impression but very few designers are stimulated by this challenge ... i.e. to extract the most from the least
You maybe will say that i am obsessed with this thing of low THD Judging from the post i see that i am not alone
Many time projects are pushed just on the basis of their very low THD performance
So my guess is that the interest for this goal is quite popular among designers
The circuit is intriguing. Driving the emitter in this circuit essentially get you some power gain and no voltage gain. The emitter is both the lowest impedance and most nonlinear load option for a transistor. It would make sense if you are trying to maximize bandwidth (i.e. 300 MHz) but no benefit for audio. And driving the least linear point means you have distortion before you start amplifing.
I work backwards from the load requirements to optimize a circuit. And there are issues to consider that add complications. Short circuit protection, overload recovery etc. And then I look at bandwidth, Selling time slew symmetry etc.
And harmonic distortion is not a cause, its a symptiom of nonlinearity. That same nonlinearity will cause IM when more than one tone is present. Personally I think distortion below -100 dBC is enough to move the focus to bandwidth and transient responce stuff. However these all seem to work in concert each facilitating improvement in the other.
I work backwards from the load requirements to optimize a circuit. And there are issues to consider that add complications. Short circuit protection, overload recovery etc. And then I look at bandwidth, Selling time slew symmetry etc.
And harmonic distortion is not a cause, its a symptiom of nonlinearity. That same nonlinearity will cause IM when more than one tone is present. Personally I think distortion below -100 dBC is enough to move the focus to bandwidth and transient responce stuff. However these all seem to work in concert each facilitating improvement in the other.
Hi ! thank you for the very valuable advice
I cannot get a 2nd harmonic distortion lower than -90dB with 1kHz signal 3V/5kohm from the circuit attached above
I tried hard because i love its simplicity but failed This means to me that topologies have indeed intrinsic limits ?
To be fair i have to say that i am using also low Voltages and common parts
Maybe raising up voltage supply and selecting boutique parts will increase performance I do not know
Since now the best performance i got is from a component similar to the LM317 I do not have the LM317 model unfortunately
I cannot get a 2nd harmonic distortion lower than -90dB with 1kHz signal 3V/5kohm from the circuit attached above
I tried hard because i love its simplicity but failed This means to me that topologies have indeed intrinsic limits ?
To be fair i have to say that i am using also low Voltages and common parts
Maybe raising up voltage supply and selecting boutique parts will increase performance I do not know
Since now the best performance i got is from a component similar to the LM317 I do not have the LM317 model unfortunately
Hi ! wow this is a lot better indeed !!! thank you very much for the very kind and valuable support
I understand well that one thing is the result at sim and another the performance of a physical prototype But i have also understood that a good result at sim is a good start Like instead a very bad result at sim should mean that the circuit need a rework
I am planning to try to build a prototype The nice thing is that the needed parts are quite cheap and common
If the prototype will reflect the sim i wonder how much is needed for a buffer seriously
The test load is already low being 5kohm ... almost no power amp has a Z input lower than this (maybe i should add some nF to the resistance ?)
And in the case of very sensitive power amp than can deliver max power with jut 1Vin or even less the THD will be even lower ... practically zero
To conclude my approach would be to stretch a minimalist circuit to its limits
So when I move on to a more complex circuit I feel more at ease with my conscience
There is an old saying by Albert Einstein: “Make things as simple as possible, but no simpler.”
I ask myself ... simpler of what ?
While I experiment the fact (also because I am verifying it with the simulations) that a typology has an intrinsic underlying limit
My doubt, as I said in the initial post, is what this ultimate limit is
Even a very simple circuit has several variables starting from better active components, supply voltages, choice of working points, etc.
When I see a complex circuit I am on one side admired and on the other shocked. It's an upsetting feeling
Thank you very much again for your precious help
Have a good day and best regards, Gino
I understand well that one thing is the result at sim and another the performance of a physical prototype But i have also understood that a good result at sim is a good start Like instead a very bad result at sim should mean that the circuit need a rework
I am planning to try to build a prototype The nice thing is that the needed parts are quite cheap and common
If the prototype will reflect the sim i wonder how much is needed for a buffer seriously
The test load is already low being 5kohm ... almost no power amp has a Z input lower than this (maybe i should add some nF to the resistance ?)
And in the case of very sensitive power amp than can deliver max power with jut 1Vin or even less the THD will be even lower ... practically zero
To conclude my approach would be to stretch a minimalist circuit to its limits
So when I move on to a more complex circuit I feel more at ease with my conscience
There is an old saying by Albert Einstein: “Make things as simple as possible, but no simpler.”
I ask myself ... simpler of what ?
While I experiment the fact (also because I am verifying it with the simulations) that a typology has an intrinsic underlying limit
My doubt, as I said in the initial post, is what this ultimate limit is
Even a very simple circuit has several variables starting from better active components, supply voltages, choice of working points, etc.
When I see a complex circuit I am on one side admired and on the other shocked. It's an upsetting feeling
Thank you very much again for your precious help
Have a good day and best regards, Gino
Hi Gino,
It's not sure if Einstein actually said that, but it seems possible.
The correct quote should be: "Everything should be made as simple as possible, but not simpler”.
In the sense of "understand".
"made as simple as possible" to understand.
"not simpler" in the sense of no longer having a meaning.
I believe that if confirmed he was talking first of all to himself.
https://graeme-47328.medium.com/simplicity-34dec5e201ac
A further related quote attributed to Albert is the following:
"If you can’t explain it to a six-year-old, you don’t understand it yourself".
which, in some way, closes the circle.
https://medium.com/@struber/if-you-...you-don-t-understand-it-yourself-6a632135127b
Hi thanks for the kind explanation
just to add a little personal anecdote I had a boss at work who systematically gave me the slides he had prepared for a presentation to get my comments
He used to tell me " if you understand them it means that 90% of other people will understand them too "
On at least one occasion I told him that some slides were not clear and needed to be reworked a little to be more understandable
He laughed in my face with a sense of pity
Then he showed them to his boss, the director who made the same comment as me
I did not tell him i told you
just to add a little personal anecdote I had a boss at work who systematically gave me the slides he had prepared for a presentation to get my comments
He used to tell me " if you understand them it means that 90% of other people will understand them too "
On at least one occasion I told him that some slides were not clear and needed to be reworked a little to be more understandable
He laughed in my face with a sense of pity
Then he showed them to his boss, the director who made the same comment as me
I did not tell him i told you
Nice story!
However, those kinds of bosses are the worst.
And I have seen, not directly, many of them do that.
I always thought that the cause of their arrogance was their wifes.
They do to others what they would like to do to their wife, while she does what to them!
The positive side of the story is that you seem to have the makings of a director!
Anyway, thank you very much for confiding that episode to me.
However, those kinds of bosses are the worst.
And I have seen, not directly, many of them do that.
I always thought that the cause of their arrogance was their wifes.
They do to others what they would like to do to their wife, while she does what to them!
The positive side of the story is that you seem to have the makings of a director!
Anyway, thank you very much for confiding that episode to me.
One can answer this question: it is to be developed in relation to a task, e.g. listening to music, audio. If this task is fulfilled...-)
And since we perceive contrastively, compare, we can foresee a development that corresponds roughly to: "as less as possible, as much as necessary"-)
And since we perceive contrastively, compare, we can foresee a development that corresponds roughly to: "as less as possible, as much as necessary"-)
referring to #50
When I started doing electroncs as a kid I built quite simple circuits - everything beyond looked too complex to me at this time.
I also removed all parts of old radios that I considered superfluous.
Certainly I always liked simplicity.
While learning more, my designs became more complex - but I always looked at each single component whether it was necessary or not.
Here is an example to make things better by adding complexity:
Replacing the collector resistor in a common emitter amplifier stage by a current source vastly increases linearity of (unloaded) output signal.
No matter how expensive your collector resistance had been before.
When I started doing electroncs as a kid I built quite simple circuits - everything beyond looked too complex to me at this time.
I also removed all parts of old radios that I considered superfluous.
Certainly I always liked simplicity.
While learning more, my designs became more complex - but I always looked at each single component whether it was necessary or not.
Here is an example to make things better by adding complexity:
Replacing the collector resistor in a common emitter amplifier stage by a current source vastly increases linearity of (unloaded) output signal.
No matter how expensive your collector resistance had been before.
Interesting To be more specific i have just one dream With a great recordings with a great system me blindfolded to get the illusion of having the orchestra in front of me I know It is a lot But there are steps The first is a very clean amplification system Then the final transducers
I think that when the reproductions is very very good the difference with live will be bearable
People are more and more noisy these days Even during music They cannot stay in silence enjoying music
Regarding the DAC: simply use the chip and bridge the following analog amplifiers or buffers. Most DACs give a clean audible analog signal, and also drive all sorts of things;-) Then use the power supply of the source or amplifier instead of its own. This gives even more resolution, cleanliness and finesse. I won't go into the further steps. I usually only use the headphone output - I'm probably too lazy;-)
Regarding current source: my experience is that the best resistors sound better than the best transistors. I have tried cs over and over again, but returned to a resistor for sound reasons. A transistor brings a "technical veil" to the sound. It slows down, slows down the tempo, or even destroys the tempi. And a few other sonic losses)-;
Regarding current source: my experience is that the best resistors sound better than the best transistors. I have tried cs over and over again, but returned to a resistor for sound reasons. A transistor brings a "technical veil" to the sound. It slows down, slows down the tempo, or even destroys the tempi. And a few other sonic losses)-;
I design Guitar preamps.True words!
SS of course.
Lots of times I tweak and tweak and tweak, test them with live Musicians playing by me, test and measure in a Lab setting, etc.
Circuits invariably get more and more complex.
At some point, just to find how much I have "advanced", I play the very tweaked complex latest version side by side with an early simple one.
Lo and behold, often the early "simple" one sounds BETTER 😲