Don't agree. You can't design a simulator if you dont know Thevenin or Norton, just to mention a few. Also, for convergence, you better know about parasitic resistance in capacitors.
And who will build your models? An area where electronics and device knowledge is required at the deepest level.
And an engineer with a sim but no math gets nowhere. How do you size R and C for a roll off to make an amp stable without math?? Trial and error? That doesn't seem too intelligent.
jan
A simulator is just solving a set of simultaneous differential equations surely? Its generic - it could be solving some equations from a completely different domain, other than electronics. Where does Thevenin or Norton fit in?
I agree modelling is where electronics knowledge is necessary, but models are not part of the simulator per se. Of course they're provided in software packages to make them useful to engineers.
Genetic algorithms?
Surely you need to write the equations before you can solve them?
I have been in many software projects where the programmer was given a 'neutral' set of specs without training him on the underlying reality that he should code for.
These projects fail.
jan
"Johnson noise is the noise one would theoretically get from a "perfect" resistor, and depends only on resistance and temperature. Real resistors have excess noise as well which depends on material, applied voltage etc. A good resistor will have noise only slightly higher than the Johnson noise, whereas a lousy resistor will have much worse noise."
That is an explanation that makes sense.
Without understanding the impedance of the CM, how on earth could one estimate the gain of the diff pair? The books say adding degeneration improves their action but lowers stage gain. They stop at that which does not help. I will go looking for better textbooks. There are a lot online, it is just finding it. MIT has everything, buy if you don't know what you are looking for, it is hard to find.
That is an explanation that makes sense.
Without understanding the impedance of the CM, how on earth could one estimate the gain of the diff pair? The books say adding degeneration improves their action but lowers stage gain. They stop at that which does not help. I will go looking for better textbooks. There are a lot online, it is just finding it. MIT has everything, buy if you don't know what you are looking for, it is hard to find.
Hi,
Beautiful, the Matrix, innit?
I would call that a pure realist.
All animals are pure realists.
That which defines Men as being on a different level than an animal is the fact the Men can transcend reality and both conceive a different reality and take actions to substitute this more desirable reality for that which is (e.g. light a fire to get warm, rather than continuing to freeze).
A pure realist considers not what things should, could or may be like but gets on with one specific job he is focused upon. Such approaches gave us poison gas, atomic bombs, Chernobyl and Fukushima and about any other undesirable aspect in human life and society.
Such realists in Trinity where having bets if setting off the first nuke would trigger a kind of chain reaction in the whole earth atmosphere or if it would be limited to a state or a smaller area (thankfully they where all wrong) but pressed the button anyway.
I think a true realist sees both that which could or should be and that which is and adjusts his or her actions accordingly for the best outcome.
Ciao T
Beautiful, the Matrix, innit?
I would call that a pure realist.
All animals are pure realists.
That which defines Men as being on a different level than an animal is the fact the Men can transcend reality and both conceive a different reality and take actions to substitute this more desirable reality for that which is (e.g. light a fire to get warm, rather than continuing to freeze).
A pure realist considers not what things should, could or may be like but gets on with one specific job he is focused upon. Such approaches gave us poison gas, atomic bombs, Chernobyl and Fukushima and about any other undesirable aspect in human life and society.
Such realists in Trinity where having bets if setting off the first nuke would trigger a kind of chain reaction in the whole earth atmosphere or if it would be limited to a state or a smaller area (thankfully they where all wrong) but pressed the button anyway.
I think a true realist sees both that which could or should be and that which is and adjusts his or her actions accordingly for the best outcome.
Ciao T
Sublime, yep.
Oh look, one of those generalizations. All animals would include humans would it not?
How would it be possible to know that animals (other than homo sapiens) do not conceive of different realities?
Guilt by association? In my earlier remark I was hinting that a 'real (or pure, no matter) realist' would not bother (i.e. waste time) announcing the fact. How would he/she know that s/he was a 'pure realist' anyway? S/he'd be keenly aware that everyone has their own meaning of what 'real' and 'pure' is. If you'd like a reference from the Bible, I'm fairly sure I could find one to back this up
With an idiosyncratic meaning of 'best' perhaps?
I'm quite unclear what you're referring to here. The equations that run in the sim are customer supplied.
Well coding to specs, rather than for customers, does tend to have this result yep. The underlying reality is the customer not the spec.
Hi,
Some humans perhaps. I guess it depends on the definition of human.
One definition can emphasise that Humans are just animals with an extra trick, where in turn one may note that animals are just plants with extra tricks and even that plants are really only minerals rearranged, at which point we have de-constructed all division and arrive at the point that all is matter, maya and divisions are artificial.
If we accept the artificial divisions, then one can say that animals include the nature of the plant, but add new qualities, that make them "not plants". In the same way Humans add a new quality which makes humans (most of them anyway anyway) "not animal".
Human is one kind of "not animal", there are others that seem to have a high indication of possessing this new quality, however they are inaccurately still listed as animals.
Egoistical Men will define best purely by reference to oneself, Altruistic Men will define best purely by reference to others. Either are abstract idealisations, as it stands Men is somewhat Egoistical and somewhat Altruistic, the mix varies...
Ciao T
Some humans perhaps. I guess it depends on the definition of human.
One definition can emphasise that Humans are just animals with an extra trick, where in turn one may note that animals are just plants with extra tricks and even that plants are really only minerals rearranged, at which point we have de-constructed all division and arrive at the point that all is matter, maya and divisions are artificial.
If we accept the artificial divisions, then one can say that animals include the nature of the plant, but add new qualities, that make them "not plants". In the same way Humans add a new quality which makes humans (most of them anyway anyway) "not animal".
Human is one kind of "not animal", there are others that seem to have a high indication of possessing this new quality, however they are inaccurately still listed as animals.
Egoistical Men will define best purely by reference to oneself, Altruistic Men will define best purely by reference to others. Either are abstract idealisations, as it stands Men is somewhat Egoistical and somewhat Altruistic, the mix varies...
Ciao T
I'm quite unclear what you're referring to here. The equations that run in the sim are customer supplied.
[snip]QUOTE]
OK, I understand what you mean. Another project that will fail
jan
You can argue all you want, but what with all that math and those simulators, in the end, it's the designer's hearing that will voice his product. In the end, it always comes down to the man behind it all.
Don't tell me none of you ever lived through a situation where maths tell you it should be a 4k7 resistor, and your ears tell you a 5k6 resistors sounds a hell of a lot better?
Or when theory tells you your input stage degeneration is becoming inordinately large, you're down to a gain of say 6:1, you should lower the gain of the second stage and raise the gain of the first, but it comes out better in the "wrong" way?
Anyone saying "yes" to the questions above would in effect be saying we can make a great amp with maths alone.
Don't tell me none of you ever lived through a situation where maths tell you it should be a 4k7 resistor, and your ears tell you a 5k6 resistors sounds a hell of a lot better?
Or when theory tells you your input stage degeneration is becoming inordinately large, you're down to a gain of say 6:1, you should lower the gain of the second stage and raise the gain of the first, but it comes out better in the "wrong" way?
Anyone saying "yes" to the questions above would in effect be saying we can make a great amp with maths alone.
I am not sure, I THINK somebody mentioned that Prof Leach does not like transistor based CCS? Or was I just imagining things?
As far as I am aware, the better the CCS, the better the rejection of supply funnies - or is that wrong?
How is your favorite CCS for the input pair(s) constructed? I love the two transistor arrangement because it's constructionally ideal, you can put the two trannies back to back and improve their thermal behavior. Some thermal compound between them, some method of keeping them together and you're all right.
As far as I am aware, the better the CCS, the better the rejection of supply funnies - or is that wrong?
How is your favorite CCS for the input pair(s) constructed? I love the two transistor arrangement because it's constructionally ideal, you can put the two trannies back to back and improve their thermal behavior. Some thermal compound between them, some method of keeping them together and you're all right.
Except for those who want an amp with as little "voice" as possible.dvv said:
I guess it all depends on how well-founded your maths is, and what you mean by "better". People often say "the maths is wrong" when what they should say is "I have misapplied or oversimplified the theory". If the maths told me that 4k7 gave a flat response, and 5k6 gave a peak/dip, then I would use 4k7 even if 5k6 sounded "better" unless I could be convinced that I was misapplying the maths. If I wanted "voicing" then I would add tone controls or a Wavebourn "niceness" knob.
I think that is called erecting a false dichotomy. A useful debating tool for those with a weak argument.
I prefer 'selfish' (aka egotistical) and 'Selfish' (non-egotistical) - the altruistic people are Selfish which is the natural result of the absence of self. The Selfish man/woman gets by without definitions at all. Dualisms always strike me as lacking in parsimony and hence a waste of valuable effort.
Hi,
I prefer not to eat S(h)el-Fish. Too much Heavy Metal... ;-)
So I'll stick to the other definition, but we are guning semantics here any number of cunning linguists could make our life very hard indeed...
Ciao T
I prefer not to eat S(h)el-Fish. Too much Heavy Metal... ;-)
So I'll stick to the other definition, but we are guning semantics here any number of cunning linguists could make our life very hard indeed...
Ciao T
Hi,
Let me put it this way.
Math is required to get a working product, to have relevant safety margins etc. Products with too little math tend to be unreliable.
Experience is needed to co-relate topology, various design choices and parts selection with subjective percieved sound.
Taste and an innate "knack" is needed to bridge the gap and get what we want from what we have.
Self honesty and additional sets of unbiased ears attached to experience listeners are needed, because often we hear what expect, not what actually happens.
So we must span the disciplines of Circuit Theory and Math across to musical Theory and Art on theoretical side and we must have the listening experience and building experience to underpin the theory, which in turn informs our practice, which in turn forces us to adjust our theories and so on.
Ciao T
Let me put it this way.
Math is required to get a working product, to have relevant safety margins etc. Products with too little math tend to be unreliable.
Experience is needed to co-relate topology, various design choices and parts selection with subjective percieved sound.
Taste and an innate "knack" is needed to bridge the gap and get what we want from what we have.
Self honesty and additional sets of unbiased ears attached to experience listeners are needed, because often we hear what expect, not what actually happens.
So we must span the disciplines of Circuit Theory and Math across to musical Theory and Art on theoretical side and we must have the listening experience and building experience to underpin the theory, which in turn informs our practice, which in turn forces us to adjust our theories and so on.
Ciao T
You mean like those who make everything relative, in the "yes, perhaps , no, not necessarily" mould?
A great example above. Thank you.
Last edited:
I find making things relative to the facts is usually a helpful thing to do. Some things are black-and-white, other things are not. Wisdom is knowing the difference. In the realm of audio electronics, circuit theory is a fact. If your circuit does not do what you expect, then either your expectation is not based on sound maths derived from your circuit or your circuit is not what you think it is. Either is possible. Component parasitics, unwanted EM coupling etc. can complicate things. Circuit theory remains true, even when our approximations to it are inadequate.dvv said:
- Status
- Not open for further replies.