Speaking of clipping, I see a lot of effort to be sure amps clip nicely and recovery well. Sounds smart. I understand how feedback can be larger than one may expect. But, I have never seen any clamp on the input to keep stupidity from just overloading it in the first place. A good preamp can swing 8 to 10 volts, yet most (consumer) amps reach rated power with one to one and a half. I have no doubt there is a good reason, just have not seen it covered in the texts.
Another one I like, in the same category as "this too shall pass":
"It's more complicated than that".
A maths friend of mine (Oxford PhD, studied with the late great dynamical systems theorist Vladimir Arnold) is quite amusing when he attempts to explain something. He once addressed a room of (mostly) engineers and some of us started to count how many times he said "basically". The final count for his short report on his group's activities was about 200.
You can never have too much math. But the "danger" is, it will become an end in itself and you will become a mathematician. And it's dangerous enough to start writing code, as one often becomes a programmer.
We take circuit theory often for granted, and suppose that it's wholly complete. Have you seen Paul Slepian's monograph? It takes the whole book to put resistive circuits on a rigorous mathematical foundation. No reactive elements let alone controlled sources, let alone anything nonlinear. I presume legions of grad students have extended the work since.
Anyway, somebody needs to design circuits.
I use optical limiters in my Pyramid amps. They sense current peaks of screen grids of output tubes and limit input signal level accordingly as soon as the amp approaches clipping.
Desigining a clean clipper is nontrivial. Although, I agree with you, the right place for it is in the signal domain (as opposed to further along when more power is involved). One recent article about phono preamps insisted that a preamp with a 36dB overload margin was needed. Well. Nice work if you can get it (if you make the gain low enough it gets easier, but that's cheating), but if you suppose that at normal decent listening levels you are running around 250mW average power and some healthy crest factor, where in the wide wide world of sports are you going to find a power amplifier that will cleanly reproduce a 36dB overload??
Brad
No one could ever accuse you of wasting chassis material ......
It is called optimization.
GOOTEE,
Yes, like ground but for power. I was thinking of splitting the DH-120 power as the input and VAS are fed after an additional RC filter on the rail, and the grounds are already separated on the board. Probably not the highest priority, but a thought.
Yea, subscribe. Did not know where they hid it.
Did not see Slepian. They have made it way to hard to get to the university library here. I can get a pass, but there is no parking. If he delved into what is actually happening at the quantum level, it could take several volumes and the final one would still be blank because we just don't really know. What I want to know, in use in a practical circuit, it is true that a half watt CF has about the same noise as a eighth watt MF but half the price? Seems to me this was suggested by none other than the Hafler company. Marketing or engineering?
"For any given product, there comes a time to shoot the engineers and start production"
Yes, like ground but for power. I was thinking of splitting the DH-120 power as the input and VAS are fed after an additional RC filter on the rail, and the grounds are already separated on the board. Probably not the highest priority, but a thought.
Yea, subscribe. Did not know where they hid it.
Did not see Slepian. They have made it way to hard to get to the university library here. I can get a pass, but there is no parking. If he delved into what is actually happening at the quantum level, it could take several volumes and the final one would still be blank because we just don't really know. What I want to know, in use in a practical circuit, it is true that a half watt CF has about the same noise as a eighth watt MF but half the price? Seems to me this was suggested by none other than the Hafler company. Marketing or engineering?
"For any given product, there comes a time to shoot the engineers and start production"
A good mathematician is good physicist and a good engineer. Sadly it doesn't go the other way around. If you can't see it you're probably an engineer.
I agree that mathematicians seem so clever that you might assume that they could easily do mere engineering without breaking sweat. I love the maths that can apparently model the real world, like reducing the bitrate of an audio signal inaudibly - suggesting that it really is some sort of model of the way nature, our brains and ears work.
But most people who profess to know maths... Don't they just apply and regurgitate standard formulae and techniques without necessarily knowing whether they are truly valid and applicable? I might like to think I know how to use maths to build a motor controller (or an audio amplifier...), but the maths ceases to be useful in the face of backlash, stiction and all the other nasties in the real world. A brilliant mathematician might blunder in, naively thinking he can solve such a simple problem in an afternoon, but the engineer would know it wasn't going to be so simple.
Engineering is about the messiness of the real world, sometimes on the level of gut feel, or dogged persistence and experience, where maths can't help you.
As an aside, I'm fascinated by the idea of self-learning systems that can simply bypass the need to learn certain types of maths - and then be disappointed to find the maths was so 'naive' it didn't solve the problem anyway! (You always end up twiddling the 'pots' manually on your PID controller or phase locked loop, seeking speed of response with stability, even if you went by the book when designing it).
Here's an example where a mathematician couldn't have helped you to solve a real world problem, but where there was an empirical solution (captured by use of a self-learning 'black box'):
IEEE Xplore - The truck backer-upper: an example of self-learning in neural networks
This might be true when they are infants. But after working in one field or the other and gaining experience and "the mindset", they are usually no longer able to do each other's job, very well. It doesn't go EITHER way, easily.
But, if need be, an engineer could probably fill the others' shoes better than they could fill his, because while many engineers are extremely good at math and physics, the engineer's job (in many cases) also has non-math and non-science components that are absolutely critical to success, mainly consisting of juggling cost, schedule, and (product) performance, and the many risks that affect them all. Any mathematician or physicist who also does that, or is good at it, IS an engineer, also.
(Edit: ) The more I think about it, the more I think that the OP had it exactly backwards.
Last edited:
Interesting article - thanks for the link. He says:
That is why I now focus on re-creating the original artistic intent
but nowhere does he say how he learns what this actually is. Nowhere in that article (I haven't read part 2 yet) does he say what an 'artistic intent' is. How is this engineering? Sounds more like religion to me - where something non-falsifiable ('artistic intent') is used to justify some activity.
Hi,
I have. Often. Mark Levinson and Cello Amplifiers come to mind, as do many pro-amp's.
My Pro Amp's have usually used optical limiters with clipping metering... The ones that did not have this ex factory had it added...
Ciao T
I have. Often. Mark Levinson and Cello Amplifiers come to mind, as do many pro-amp's.
My Pro Amp's have usually used optical limiters with clipping metering... The ones that did not have this ex factory had it added...
Ciao T
My first amplifier was NAD3020 and this has a soft clipping switch on the rear. I haven't looked at the circuitry inside to see if its an input limiter.
I'll give you some of that. Not quite all. I only made it through second semester calc, which is not enough for engineering. Higher math is only one of many tools an engineer needs. My argument is that math is the language of nature, the language of science. It can be developed, studied, improved and quite elegant. "We" invent new maths to describe ever more complex observations of science. Geometry, Calculus, Chaos. It is not the science. "We" excludes me and the math I invent. As a dyslexic, is not very useful.
BTW, my boss is a mathematician, my sister a mathematician, about half of my customers are mathematicians. None of them are engineers. Most need supervision around sharp objects.
I know just two mathematicians, and judging by their example, you are oh so right! They live in a different space-time continuum, and occasionally drop in to count us, to see of we're all still here.
Well, my first degree being in Applied Mathematics, it seems you guys have worked with the "wrong" type of mathematicians (i.e. theoretical).
Last edited:
- Status
- Not open for further replies.