Mathematics is a concept, a language. At present, no distinction is made between object and concept in mathematics either. Most mathematicians cannot, and therefore neither can contemporary mathematics. "Theoretical physics" is a prime example: unnoticed, a mathematics has found its way in that is based on misobservation (e.g. Michelson Morley experiment), or disregards observations (e.g. solar phenomena, see picture below), or calculates things that physically are not or cannot be (e.g. dark matter) or are not or will not be feasible (e.g. time travel, fusion reactor). More precisely: this mathematics is not mathematics, science, but dreams and fairy tales;-)
I forecast: Ginetto's AUDIO opamp will be based on a lot of trial and error, on practical experience, testing and tuning away from conventional electrical engineering. To become a real and best AUDIO opamp;-)
I forecast: Ginetto's AUDIO opamp will be based on a lot of trial and error, on practical experience, testing and tuning away from conventional electrical engineering. To become a real and best AUDIO opamp;-)
Attachments
This SE "test amp" made 200 watts with one vacuum tube, the 833A. Yes, it came from a 500 watt AM radio transmitting station. The old RCA transmitters used 4 X 833A tubes. Two were used in the parallel wired final RF amplifier and the other two were used in the class AB2 audio amp known as the modulator. The OPT was custom wound by Transcendar. Operation in the red glow region is acceptable, and actually preferred for this tube. Makes for a wicked guitar amp too!
Attachments
Thank you and very kind of you and of course i wish all the best to you as wellYou have to ask it, with your.
From what we can see, you have all it takes to make your dream come true, although naturally the "time" factor is an important variable.
However, in my humble opinion (and not just mine) this is the best forum in the world where you can draw on the experience of others, but you have to know how to "choose" who can really help you.
And for sure, at least for your projects, that someone isn't me! 😉 🙂
I affectionately wish you all the best! 👍
You designers are geniuses
Ginetto, please note that I might even actually be a genius, who knows, but I'm certainly not a designer. 🙂
However, mine was meant to be a "see you soon", certainly not an adieu! 😉
I thank you sincerely for your kind words but my situation is the one of a short guy who wants play basketball
However i am sure that designers follow a method For instance i have seen some circuits that have a sort of symmetry On one lets say branch there is a npn on the other the complementary pnp etc
so i could try to put together a branch and do the other is a symmetrical fashion Problem is the interaction
Some other discrete opamps are quite similar
Some brands have been stuck with proven designs for many of their products
I guess that it is not a 2 minutes task to put down a great performing circuit
Constant current sources seem useful to try good working points
I will measure my progresses measuring the % of the simulations that will provide a nice sine at the output
For now i have had issues even with circuits taken from the web or service manuals They are strangely depicted
Thanks a lot again
I would just like to add that not too recent devices that use discrete and that have had very positive reviews for sound do not seem really outstanding at least in simulation When I see the datasheets in comparison the opamps seem to come from an alien intelligence
Several high end manufacturers have already jumped on the single chip opamp bandwagon.
I'm just sorry that they are so small. I would love them to be bigger and better cooled for example. Like a small power pack.
Several high end manufacturers have already jumped on the single chip opamp bandwagon.
I'm just sorry that they are so small. I would love them to be bigger and better cooled for example. Like a small power pack.
Just wait till AI starts doing circuit design of these things. I bet someone somewhere has fed one lots of those topology patterns so it can learn how to "draw up" something of it's own creation. You know it can simulate whatever it comes up with too, until some desired measurable criteria is hit. I bet it would be a marvelous schematic diagram and something no one has yet designed.
Sorry if I don't agree.
[IMHO]
AI is anything but intelligent.
AI is nothing but data.
The data is processed by CPUs.
AI is processed by many CPUs.
Many CPUs process a lot of data.
The faster you access that data, the more impressive the result will seem.
But all that remains is data.
There is absolutely nothing intelligent about it, other than the evil intentions of those who allowed all this to happen and who still today have not thought of having to regulate this more than strategic sector.
We already talked about it on the following thread about a year ago:
https://www.diyaudio.com/community/threads/chatgpt-ideas-about-amplifier-design.394541/page-27
What still seems to matter today is not what AI really is, but what people think about AI, or to put it another way, how people have been manipulated by the holders of the so-called strong powers.
That data lacks the ability to be intelligently combined.
Is that called human genius?
And people on average instead think that AI can provide you with a high-quality finished product.
If we are talking about the same level of quality, well, it can't.
Can AI create crap like fake videos of people from the past (or present) that look real, but are fake?
Can AI "create" a non-existent photograph of something or someone that looks real?
Can AI create a fake poem, passing it off as true?
Is this AI?
AI cannot create an authentic thought.
It can get you a finished product of mediocre quality.
But who benefits?
In quality AUDIO it is not possible to proceed without human genius.
Why?
Because AI can neither listen nor (if it could, but cannot) evaluate the artistic value of what it has listened to (but cannot listen to).
Furthermore, in that data the human genius is hidden by the same mystery for AI that for me is reading musical notes on a line of a musical score and no CPU will ever be able to recognize it.
Or re-create it.
[/IMHO]
Last edited:
We should learn to think processually, to think logically. Also in terms of complexity, increasing and collapsing complexity, leaps in quantity > leaps in quality. That sort of thing.
We also need to abandon our human-centeredness. An AI does not have to develop any kind of human-audio-compatible amplifier.
Unfortunately, people do not realize that they are only parts of a context, they are bundles of mechanisms: We look back at our competitor to see that we are a nose length ahead of him. So we overlook where we are running;-)
We also need to abandon our human-centeredness. An AI does not have to develop any kind of human-audio-compatible amplifier.
Unfortunately, people do not realize that they are only parts of a context, they are bundles of mechanisms: We look back at our competitor to see that we are a nose length ahead of him. So we overlook where we are running;-)
Yes even if at least for now i heard of some AI generated schematics that dont work
Maybe artificial intelligence already has a sense of humor?
Let's just hope he doesn't enjoy seeing stupid people like me go crazy on a scheme that doesn't run on simulation A sadistic AI
But yes I was staring at the schematic of a LM317 If we look at it it is just a small black piece of plastic with some tiny legs ... a tech beatle
But looking at its schematic one can getan idea of its complexity
To be honest in an ignorant way i wonder if all this complexity is really needed Is it really needed ?
Maybe a chip could be made simpler and having more or less same performance ? or is complexity unavoidable ?
@ginetto61
If I may, even though I'm not an expert, in my personal view I think I can say that you will discover soon, when you start to put into practice what you are learning in theory and build your own experience by learning from your mistakes, is that that complexity is nothing more than a set of simple things put together.
That inevitable complexity seems such to your still not too expert eyes, but it is not at all as science fiction as it seems to you now.
Simplification may also be ingenious, but perhaps it costs less for a manufacturer to add another brick to what already exists rather than re-design the whole thing to "simplify" it.
A stroke of genius can always happen, however.
Just my 2 cents...
If I may, even though I'm not an expert, in my personal view I think I can say that you will discover soon, when you start to put into practice what you are learning in theory and build your own experience by learning from your mistakes, is that that complexity is nothing more than a set of simple things put together.
That inevitable complexity seems such to your still not too expert eyes, but it is not at all as science fiction as it seems to you now.
Simplification may also be ingenious, but perhaps it costs less for a manufacturer to add another brick to what already exists rather than re-design the whole thing to "simplify" it.
A stroke of genius can always happen, however.
Just my 2 cents...
The LLM / AI stuff reminds me of my miscellaneous Ubuntu "work flow" a few years ago. Whenever I wanted to solve some Linux issue, I'd go on google, then onto some website, and I'd follow the "thread" on what somebody else did who'd had the same issue. A couple of years later, I reinstalled, found that I had the same issue but couldn't for the life of me remember what I did to solve it.
So I started off at the same point as before, but noticed the process was a big long "tunnel". The recommended solution was filled with errors, but inspite of that, the search engine was just blindly feeding the exact same results, step by step, with no apparent awareness of the end goal. It's kind-of what I see with the AI.
If AI could do such a thing for you, I wonder if that'd sell?
I have a and other similar situations with my Windows machines. Example, if I put files in an arbitrary subdirectory Windows recognizes as music it changes how the file explorer operates, flat out assuming I'd want what they think I'd want. I want it the same as any other files; I want to sort by date and dont give a flap about "artist" or "genre".
There's a fix and I have to go digging through the internet to find the steps, as I never remember. One would think these days you could just say "Hey file manager - display the content of this subdirectory the same as ordinary files" "But they're music files!?" "Dont care - do it anyway". With my luck, all the files on the machine would end up being indexed by artists and genre, with OS reinstall being the only option to fix...
AI. I must be dreaming it could do something like that, or the powers that be would set it up to assist a user in that way. Instead, they want to track and train it on every single thing you do.
Ultimately "their" logic is simple, indeed brutal (in the sense of "in the brute state").
Do they think that if they offer you (actually, force you) a "free" service then you really want to give them "nothing" in return?
And here they take what they want, and much more, without even listing it.
And the majority of people who buy cell phones (just as an example) with built-in AI will use it for the usual BS to "brag" about on social media, they will be amazed with their mouths wide open when they see the "sci-fi" results they think they have obtained for free and the other parts of the world will continue to wage war or have to get the water they need every day from kilometers away, on foot.
If you saw the above happening on another planet through a very powerful telescope, what would you think?
Do they think that if they offer you (actually, force you) a "free" service then you really want to give them "nothing" in return?
And here they take what they want, and much more, without even listing it.
And the majority of people who buy cell phones (just as an example) with built-in AI will use it for the usual BS to "brag" about on social media, they will be amazed with their mouths wide open when they see the "sci-fi" results they think they have obtained for free and the other parts of the world will continue to wage war or have to get the water they need every day from kilometers away, on foot.
If you saw the above happening on another planet through a very powerful telescope, what would you think?
At the moment, for example, ChatGPT is acting as a propagandist for the UN Agenda 2030, with all the nice words and phrases, and it seems that this AI is currently only capable of repeating things that have been predetermined and framed by some people. A critical analysis of this complex issue is neither to be heard nor to expect.
If an AI were to build an AUDIO amplifier suitable for humans, it would currently fail because AI does not have the hearing of humans and cannot carry out comprehensive tests with complex systems e.g.
If an AI were to build an AUDIO amplifier suitable for humans, it would currently fail because AI does not have the hearing of humans and cannot carry out comprehensive tests with complex systems e.g.
good morning everyone I don't know if the various algorithms that analyze the connections of various users to the web and social networks can be considered AI
Maybe is Artificial Psychology ?
but lately I have a rather surprising and worrying feeling
Until recently, when I did a search, I was offered sites more or less in line with what I was looking for
Now when I do a search they offer me sites that are relatively related to what I'm looking for but which in the end interest me more than what I was actually looking for
As if the algorithm basically knows me much better than I know myself And it also does it in a kind way. Like, I know you were looking for something else but isn't this better instead? and I'm sure this is better
I believe that the objective is ultimately rather than influencing the subject to exaggerate by giving him what he wants. A kind of drug for the addict
Why ? because from what I understand you can really make a drug addict do anything and if he refuses you take away the drugs and he goes crazy
An example is taking away the cigarette of a heavy smoker. Try doing it and you will see the impressive transformation
When should one worry? when the algorithm refuses to give you satisfaction
Because you want too much Even for the algortihm
Maybe is Artificial Psychology ?
but lately I have a rather surprising and worrying feeling
Until recently, when I did a search, I was offered sites more or less in line with what I was looking for
Now when I do a search they offer me sites that are relatively related to what I'm looking for but which in the end interest me more than what I was actually looking for
As if the algorithm basically knows me much better than I know myself And it also does it in a kind way. Like, I know you were looking for something else but isn't this better instead? and I'm sure this is better
I believe that the objective is ultimately rather than influencing the subject to exaggerate by giving him what he wants. A kind of drug for the addict
Why ? because from what I understand you can really make a drug addict do anything and if he refuses you take away the drugs and he goes crazy
An example is taking away the cigarette of a heavy smoker. Try doing it and you will see the impressive transformation
When should one worry? when the algorithm refuses to give you satisfaction
Because you want too much Even for the algortihm