What's so cool about "analog"?

[TimFox]

No. Amplifiers (which may not be connected to digital systems), oscillators, etc. are examples of truly analog systems.

[tggzzz]

so... all sensors are of analog, and only CPU's are digital devices?

CPUs are analogue devices with individual transiators operating in saturated mode, that process digital signals.

[AlfBaz]

Analog is trending toward cool as now "non-binary" is becoming a thing

[Doctorandus_P]

Some people are working on analog computers to simulate neurons and build ai networks.
It makes sense, as each "unit" does not have to be very accurate and can be very simple, so a lot can work in parallel.

[SiliconWizard]

The world is inherently analog, ...

Nah, the world is quantised, the relevant quantums being photons and electrons. Hence nature is digital

But the wave functions must be continuous, at least as far as I understand QM (that is, only a little bit.) So you tell me if that means the universe is analog or digital. Considering wave functions, I'd be tempted to lean towards the former. That's probably material for a 100-page long thread, and counting.

[chris_leyson]

Would anyone please explain to me why they love "analog"? I always thought digital IC's are better. Right?!

All digital ICs are based on analogue design principles and that is why analogue design is cool. If you look into the history of digital logic families like DTL, TTL, ECL and MOS you will learn that they are all based on analogue design principles. For example ECL or emitter coupled logic is a family that doesn't saturate to get high speed. Nice lecture here
https://youtu.be/MqidrO1TPvQ

You can't make sensitive low noise amplifiers with digital ICs, they're all analogue.

On the other hand you can do some things where digital is better, for example DSP or digital signal processing where you can mathematically model analogue filters. You still need analogue electronics like A to D and D to A converters to swap between the two.

Perhaps if you explained your reasoning as to why you think digital ICs are better than analogue ICs that would help.

[tggzzz]

The world is inherently analog, ...

Nah, the world is quantised, the relevant quantums being photons and electrons. Hence nature is digital

But the wave functions must be continuous, at least as far as I understand QM (that is, only a little bit.) So you tell me if that means the universe is analog or digital. Considering wave functions, I'd be tempted to lean towards the former. That's probably material for a 100-page long thread, and counting.

I think the best answer is "mu", as in https://csarven.ca/mu

[TimFox]

It's worse than that:  wave-function solutions to Schrödinger's equation must be complex functions.  Note the "i" in that equation.
This is not the same as "j" in electrical engineering, but a physical "i".

[aneevuser]

The world is inherently analog, ...

Nah, the world is quantised, the relevant quantums being photons and electrons. Hence nature is digital

But the wave functions must be continuous, at least as far as I understand QM (that is, only a little bit.) So you tell me if that means the universe is analog or digital. Considering wave functions, I'd be tempted to lean towards the former. That's probably material for a 100-page long thread, and counting.

I suspect that the original comment was tongue-in-cheek, but if not, it's certainly not true to say that the world is "digital" because of quantisation - for example, the energy spectrum of an unbound state (e.g. that of a free electron) is continuous. So we could equally well argue from QM that the world is continuous.

[Rick Law]

All those "nature is digital" and/or "nature is analog" are tongue in cheek.  We do rely on certain quantum effects for our nature to be the way it is, but the nature we know is far too macro for quantum effects to be dominant, at least for the most part.  We not only live in a macro world, we live in a statistical world.  Statistical effects are the dominant ones in our day-to-day life.  Even with people: A person can be very unpredictable where as a group of people will follows very predictable behavioral paths.

What make analog more interesting is that you can see what is going on.  From the days of vacuum tubes to today's ICs, you can trace this input to that vacuum tube to do whatever, then to the next component-configuration to do whatever, follow the line all the way to to output.

Digital is more abstract.  Worst is of course MCU and the likes where you can have the same component doing entirely different things depending on the software that you can't see.  They are "black boxes" that do their magic and you don't exactly know how it was done, or what else might have been done.  So there are plenty of unsolved mysteries looking at MCU based circuits and some you can never solve unless you are given the software to look at.  Personally, I don't like unsolved mysteries.  I like to understand things.  But that is personal choice and yours may well be different, or not.

But, digital and analog are often mixed making a much more interesting world.  Look at Pianos.  The old pure analog ones, you can see the key press linking to the striking hammer against the piano-wire of specific tune.  The digital ones?  You really cant be sure how the sound is made.  And then there are the mix:  real analog Concert Grands with real piano-wires, sound wood boards plus digital record and edit capability using the real piano hammer striking real piano wires.  Actually, I would not mind one of those...   But may be not.  If I do have one of those, I would be likely be editing most of the time rather than playing it most of the time.

[Nusa]

Any discussion of music has to recognize that human ears are analog devices.

[tggzzz]

Any discussion of music has to recognize that human ears are analog devices.

Functionally you can't separate the ear from the brain. Nerve impulses are inherently digital.

[SiliconWizard]

42.

[tggzzz]

42.

"Mu" is better.

[SiliconWizard]

42.

"Mu" is better.

But 42 is even.

[fourierpwn]

This thread is dumb, as is my response.

[TimFox]

Any discussion of music has to recognize that human ears are analog devices.

Functionally you can't separate the ear from the brain. Nerve impulses are inherently digital.

Actually, the ear is a Fourier analyzer, and the nerve bundle to the brain is frequency encoded.

[tggzzz]

Any discussion of music has to recognize that human ears are analog devices.

Functionally you can't separate the ear from the brain. Nerve impulses are inherently digital.

Actually, the ear is a Fourier analyzer, and the nerve bundle to the brain is frequency encoded.

The amplitude of each frequency is digitally encoded as a pulse density modulated bitstream signal. Yes, I know that's a bit simplistic, but it is not a bad description.

[coppice]

In the 70s it was cool to be digital, as it was really hard to do things digitally. Gradually it became easier to do a wide range of things digitally, the stuff most engineers found really hard (like DSP) became reusable packages, and there was something of a revitalisation of analogue as the clever stuff only a niche group of engineers was good at. Its surprising how few semiconductor vendors really excel at analogue any more, and its usually for places digital can't go, like super low noise signal conditioning. Still, they are enough to serve the market.

[tggzzz]

In the 70s it was cool to be digital, as it was really hard to do things digitally. Gradually it became easier to do a wide range of things digitally, the stuff most engineers found really hard (like DSP) became reusable packages, and there was something of a revitalisation of analogue as the clever stuff only a niche group of engineers was good at. Its surprising how few semiconductor vendors really excel at analogue any more, and its usually for places digital can't go, like super low noise signal conditioning. Still, they are enough to serve the market.

Try finding a new high frequency PNP transistor; the old ones were better. Even JFETs are becoming rarer.

[Capernicus]

digital computers put log complexity on what analogue does.

Analogue works in a single bus, to do its arithmetic,  you can do multiplies with voltage dividers,  takes up alot less space.

It makes way more sense to make an analogue computer than a digital one when it comes to space takeup.

Analogue pong is 10 times smaller than digital pong.

[m k]

It's worse than that:  wave-function solutions to Schrödinger's equation must be complex functions.  Note the "i" in that equation.
This is not the same as "j" in electrical engineering, but a physical "i".

So the wave function is not real.

Can two low energy photons that can't release an electron become one high energy photon that can?

[TimFox]

That would not conserve charge.

[m k]

What is a free electron then, and how close is closest of really close?

[TimFox]

What is a free electron then, and how close is closest of really close?

Could you rephrase that?
The question sounds like a Monty Python sketch.