Going further with the current source example, a resistor has some simple current vs voltage vs temperature relationships, while a CCS can start to do some interesting things. As Vce changes in response to the signal, power dissipated obviously changes, which affects Vbe, so you start getting those bizarre audiophile-ish complaints like the amplifier "sounding slow". Well yeah, that's because it is slow — slow to recover from transients that cause the amplifier to temporarily become de-tuned while its discrete transistors drift back to their normal operating temperatures.
I've seen simulations where a tightly tuned LTP can cancel even harmonics by an order of magnitude, or 20dB, relative to the odd ones. But if you lose that tuning, they go back up by 20dB or more. Is it still just "harmonic distortion" (or worse, "hearing things that aren't there") if the harmonics themselves move around and change their character dynamically? Douglas Self disputes the existence of "thermal distortion", but IMO it's just coy semantics because there clearly is some kind of distortion at play, even if it's just the usual THD or IMD undergoing secondary modulation at a slow rate.
This then brings to mind the idea that a source of distortion might not be so audible on its own, until something happens that brings our attention to it.
Returning to the CCS, it's very tempting to optimise it with a cascode. The CCS gets a constant voltage load at the expense of a few extra parts. But the risk of a "failure of imagination" is real, so it's very difficult to categorically state whether a simpler solution like a resistor is duller for not capitalizing on some neat circuit, or actually smarter for avoiding all the pitfalls.
I use a separate housing for the power supply when I don't want to damp the amplifier to death. But then without plugs and sockets, without connectors. Vibrations from the transformer would not be a problem.
In general, I recommend aligning the transformer so that the listener is exposed to the lowest possible EMF: Toroidal transformer - the listener looks through its hole;-) But this alignment can also make sense in terms of electronics - if the listener is far enough away from the transformer: > 3 meters in most cases. I can feel some transformers for many meters)-; These EMFs are not healthy)-;
And please don't forget the neighbors: they often sit in the EMF of some TV or amplifier - and after a few years slip off their couch with a heart attack or stroke - or don't wake up one morning;-)
Things like noise, bandwidth, whatever distortion measure is most suitable for the application. I wrote "supposedly" because I don't think you can actually prove that the solution is close to optimal, except in the case of noise.
ok that's pretty much a good summary of iteration steps, but how about listening? Try correlate from the listening impressions and measurements, learn what different types of distortion pattern or the lack of for the ear detectable distortion, ie. "very low distortion" gears, do to the sound impression... be honest with yourself whether it's is objectively correct VS subjectively pleasing, accept both strengths and weaknesses on equal terms, well... that is.. do you trust your senses? :)
Just an observation from my side with regards to so called low distortion SS amps (comparing to my little old tube amp) is that, at first glance some euphonic audio gears may have a sound appeal it's easy to fall in love with at first "sight", at least on simple type of simpler music it may fit quite well, but with more complex music it often falls apart, also, although the low distortion amp and other audio gear may initially appear cold and emotionless, but giving oneself a "burn-in" period, just forget about the gear, relax and enjoy the music, after a month, two or three.. you start to appreciate all the details that are there woven into the sound fabric not veiled by some naively appealing euphonic distortion, well, that is how I come to appreciate "low distortion" audio gear in the long run.. the constant run after the audio nirvana solution is only snaring in oneself... the chance is you will self-deceive and never find your way out of the maze, beware, the hidden high-end hand is there to keep you fenced within a riddle wrapped in a mystery inside an enigma. ::)
Good luck experimenting with ad811, it's an opamp intended for video signals and uses "current feedback", a true beast, but who knows, it may do well for audio too. :)
The method can also be applied differently: With the knowledge of the tonal differences of parts - they sound as different as wines or beers taste - I look for suitable or desired tonal characters. And then comes the circuitry. By the way, there are also projects in this forum that work in this way: see for example Nelson Pass and the JFet stories - only here the components are not listened to beforehand in order to determine their audio suitability, but one starts from a few measured visual values that, according to experience, correspond to a tonal quality.
However, experience has shown that mixing parts like mixing colors: the colors blend together and it becomes gray and distorted. Unlike with cuvees or beer mixing: here, the aromas and colors usually remain perceptible. So the only thing left to do is to develop the simplest possible circuits with as few components as possible.
Which, of course, is not feasible for highly complex tasks such as D-A conversion;-)
And not overlook: I have already achieved the desired sound by choosing the part, or some parts - which I can’t achieve through complex setups)-;
However, experience has shown that mixing parts like mixing colors: the colors blend together and it becomes gray and distorted. Unlike with cuvees or beer mixing: here, the aromas and colors usually remain perceptible. So the only thing left to do is to develop the simplest possible circuits with as few components as possible.
Which, of course, is not feasible for highly complex tasks such as D-A conversion;-)
And not overlook: I have already achieved the desired sound by choosing the part, or some parts - which I can’t achieve through complex setups)-;
hi thank you very much for your precious advice and apologize for the belated reply
My dream goal is to get a volume control that does interefere a minumum with the rest of the chain
Every element in the chain will have an impact starting from the final transducers headphones or speakers
The question about the listening test of a preamp implies the use of ancillary equipment that could influence the overall sound a lot
I have seen so many reports about blind tests to be skeptical of the process I am looking always for low THD+noise mainly
For now i have decided to study more the sim softwares I understand almost all designers use them at least in the first stages of the design process
A tip regarding tests: double-blind is not a prerequisite for scientific validity.
Experience shows that
a) it makes sense for listeners to hear their known system, i.e. mostly at home, and that
b) there should be a recovery, remembering, and adjustment break for the ear of at least 10 seconds - it can also be many, many minutes, and
c) there can sometimes be very long phases of familiarization with a particular sound, i.e. device. This is another reason why a home practice makes sense,
d) the expectations of those involved should not be misleading. It must therefore be communicated in advance, for example, that there need not be any differences to be heard if that is part of the test, sometimes not to serve any differences;-)
And more...-)
If it is to be "scientific", then it must be a suitable method;-)
Experience shows that
a) it makes sense for listeners to hear their known system, i.e. mostly at home, and that
b) there should be a recovery, remembering, and adjustment break for the ear of at least 10 seconds - it can also be many, many minutes, and
c) there can sometimes be very long phases of familiarization with a particular sound, i.e. device. This is another reason why a home practice makes sense,
d) the expectations of those involved should not be misleading. It must therefore be communicated in advance, for example, that there need not be any differences to be heard if that is part of the test, sometimes not to serve any differences;-)
And more...-)
If it is to be "scientific", then it must be a suitable method;-)
It may well be that an "ideal" volume control must be positioned at the loudspeaker. Directly at the driver;-)
A complete active preamp is far from a good volume control. Even an epassive version with its own box and sockets is a piece of junk. So could be the best to use some kind of thing at the power amp input. Couldn't the input resistors be replaced?
A complete active preamp is far from a good volume control. Even an epassive version with its own box and sockets is a piece of junk. So could be the best to use some kind of thing at the power amp input. Couldn't the input resistors be replaced?
The "volume pot" is an inescapable necessity (smoking pot too..?) and so are the attached culprits, and.... imaginary "culprits", which needs a specialists intervention.. :)
I found this ASR thread on volume pots, and the reference in post #3 to Benchmarks page is a good read, warmly recommended.
https://benchmarkmedia.com/blogs/application_notes/13095789-volume-control-technologies
https://web.archive.org/web/*/https...on_notes/13095789-volume-control-technologies
Mainly there are four different types of volume controlling technologies, from their blog post:
- Digital attenuator - digital signal processing (DSP)
- Integrated analog volume circuit - (IC)
- Passive attenuator - resistor network, or passive potentiometer
- Active gain circuit - amplifier and variable resistance
https://www.nisshinbo-microdevices.co.jp/en/MUSES/series/MUSES72320.html
https://www.nisshinbo-microdevices.co.jp/en/MUSES/series/MUSES72323.html
See here: https://www.diyaudio.com/community/threads/muses-volume.322983/
Can it get any better? Ü
I see what you mean but at least for easy of use I still prefer a conventional attenuator
I used in the past a stepped series attenuator and it was very good indeed but it was 50k
I intend to buy a 5 or 10klog dual pot of decent quality
I like cermet ones in particular
They are linear but that can be modified
Good morning to Everyone
Just to fix one fundamental principle i quote the introduction i read on the book
https://m.media-amazon.com/images/I/61YocYF7cuL._SL1360_.jpg
Can we agree on this ! that SIM comes always first ? ok after the idea of the circuit course
Just to fix one fundamental principle i quote the introduction i read on the book
https://m.media-amazon.com/images/I/61YocYF7cuL._SL1360_.jpg
Can we agree on this ! that SIM comes always first ? ok after the idea of the circuit course
Why always?
Unless you plan on selling an audio product more cheaply by skipping all listening verification / validation tests, then development will probably include multiple round-trips of listening. And in some cases, like guitar effects, distortion, soft-clipping, modular synths, or hard-to-simulate components like transformers, a "rapid prototyping" kind of workflow could be more effective.
It depends on what you are trying to achieve. 'Correctness' according to a theoretical recipe, or some kind of real-world result.
Unless you plan on selling an audio product more cheaply by skipping all listening verification / validation tests, then development will probably include multiple round-trips of listening. And in some cases, like guitar effects, distortion, soft-clipping, modular synths, or hard-to-simulate components like transformers, a "rapid prototyping" kind of workflow could be more effective.
It depends on what you are trying to achieve. 'Correctness' according to a theoretical recipe, or some kind of real-world result.
It's a quote from a book about SPICE, particularly LTSpice. There is no obligation to use a simulator at all, but if you want to use one, it indeed makes sense to get your circuit to work in both theory and simulation before building a prototype.
If, after building a prototype, you find that the circuit that works so well in simulation works poorly or not at all in real life, you can go back to the simulator and try to figure out what was missing from the simulation model. Hopefully that will then also give you an idea what is wrong with the real-life circuit and what you can do about that. Besides, it will teach you something about the possibilities and limitations of your simulator.
If, after building a prototype, you find that the circuit that works so well in simulation works poorly or not at all in real life, you can go back to the simulator and try to figure out what was missing from the simulation model. Hopefully that will then also give you an idea what is wrong with the real-life circuit and what you can do about that. Besides, it will teach you something about the possibilities and limitations of your simulator.
Last edited:
Hi I would like to thank you All for the very kind help and precious advice
Let me put the question this way
I agree completely that what works even very well on sim could not work maybe at all once built as prototype
But I am a beginner and I need to know above all other things if something that doesn't work on sim could work in reality
I really hope that the answer is negative
If this is true for a beginner like me to be able to use wisely a Sim software is extremely valuable because the sw will tell me if a circuit cannot work
It is like a pre condition
Instead if some circuit doesn't work at all at Sim and instead works as a prototype I will stop immediately my efforts and I will not bother you anymore
Let me put the question this way
I agree completely that what works even very well on sim could not work maybe at all once built as prototype
But I am a beginner and I need to know above all other things if something that doesn't work on sim could work in reality
I really hope that the answer is negative
If this is true for a beginner like me to be able to use wisely a Sim software is extremely valuable because the sw will tell me if a circuit cannot work
It is like a pre condition
Instead if some circuit doesn't work at all at Sim and instead works as a prototype I will stop immediately my efforts and I will not bother you anymore
Thank you very much indeed
This is a small step for a competent audio designer but a giant leap for me
I have to fix some points at this initial stage
I am trying to learn the behaviour of parts using the SW
It's hard but intriguing
I can only say that I have checked some circuits of commercial units and they Indeed work also on simulation
I am focusing the source to power amp link with the volume control
Next step I would like to understand how much I can vary the gain of a power amp without causing oscillations
I have already fried a Rotel integrated amp some years ago in the attempt to reduce its gain
And also I burnt my fingers
Bad thing ignorance
This is a small step for a competent audio designer but a giant leap for me
I have to fix some points at this initial stage
I am trying to learn the behaviour of parts using the SW
It's hard but intriguing
I can only say that I have checked some circuits of commercial units and they Indeed work also on simulation
I am focusing the source to power amp link with the volume control
Next step I would like to understand how much I can vary the gain of a power amp without causing oscillations
I have already fried a Rotel integrated amp some years ago in the attempt to reduce its gain
And also I burnt my fingers
Bad thing ignorance