Electronic Product Design

I agree with you Stevem. Analog circuitry is not going away soon.

Beside your examples I should add, also, the power supply circuits. Although the PWM controller is -a kind of- digital, the rest of the circuit remains purely analog.

Good point.

Exactly what I was going to say. Digital is great until you have to interface to the real world....

Stevem,

You have it exactly right. The real world is analog. The continuous stream of an electric signal is an analogy of the real condition being measured. So making the interface from the analog real world to the digital surreal world and vice versa will always require some analog circuitry. Also there are several critical things that analog circuitry must do that digital processing cannot, such as the vast fields of signal conditioning and power distribution. For it is only after real world signals are limited to a predictable bandwidth and amplitude by analog signal conditioning that analog to digital converters can actually measure anything with any reliability. As far as power distribution, well I'll leave that to the any researcher here willing to compare the arguments between Edison and Tesla about AC or DC power distribution.

I'm actually startled that this question is being posted at this forum. Normally this kind of question get asked by people who have no concept of analog and digital signals.

I work with electronics as a hobby and for some side business. I see too many devices already with digital control systems that are fussy, finicky and expensive to replace and time consuming to design in the first place. And whats worse is these digital circuits are easily replaced by simple cheap and very reliable analog components.

To me too many engineers put digital controls on devices only because they want to force something to work in a particular way. If they had a better understanding of the physics involved in the actual device they are trying to control they would have quickly seen that what they were trying to accomplish could have been done in a far simpler and often times more effective and robust way with an analog control system.

I have also been around people that do the digital design and programming and often times their reasons for using it are rather frightening. Most often I here things like 'I don't understand analog or physics so that why I use digital'. WTF? If you don't understand the physics of how something works how do you write the proper control software for it?

Digital is only as good as the software that drives it. Analog is only as good as the physical working limits allow it to be!

There are many applications were digital is the best thing but still there are many more that are far better suited to having analog controls systems.

Analog will never get replaced.

The progression is to push more of the rendered signal into the digital domain sooner. Tasks that once were more economical to perform in the analog domain are now more economical to do in the digital domain. Tasks that once could not be done because they were way to cumbersome to do in analog and not yet doable in digital are now easy and practical due to advances in both.

The eternal challenge to digital will be to convert with less noise at lower power levels and higher resolution. For digital it will be to keep up with higher bandwidth processing while maintaining low power demands.

I have a toolset on my bench for a digital hearing aid. In one chip it amplifies, converts, does a 128 bin FFT, a 128 bin IFFT, converts back to analog and amplifies to drive a receiver. It has 5 DSP's doing an equivalent of 40 MIPS in your ear on a hearing aid battery in your ear. Simply amazing!!!!

Economic (to produce) versus convenient (to use)...

The radio in my car has digital controls. I hate it and it's impossible to modify it. Digital controls with less buttons than would be convenient to use drive me crazy.

As Stevem has writen the world is analog so when someone starts to design any digital gadget first they would become familiar with the analog functionality. Then they can decide what the better way would be.

Analog (electronics) will never go away, maybe it will become less, but it will always be around. like those audiophiles that swear by analog amplifiers

A curious thought though, that neurologically, we are largely digital. The synapse is either on or off and muscle fibers are either on or off.

Absolute rubbish ... fist a synapse is a connection to and from neurons as neurons have as many 10 000 synapse connections it makes the neurons practicly analog, secondly your muscles and limbs usually sit happily in many different positions ie you dont always have your arms full extended or retracted .. in most cases you vary it for comfort control or position making it very analog movement ..

Think before your post !!!

Clearly you put a lot of thought into that. Perhaps had you given it more thought you would have gotten my inference. What are you saying, that your muscles are "half way on" when your arm is half way extended? Sounds like you don't know much about how the body works. Another worthless "Guest" post.

I don't want to exaggerate or be Pollyanna about this, but...

I think "analog" is (and will increasingly be) being discovered by a digital "generation". Witness the rediscovery of vinyl music vs. CD. LCD TV's are nice, but the picture still isn't up to CRT standards. Etc., etc., etc.

I deal with analog circuits. I deal with digital circuits. I am not an "analog guy" or a "digital guy". Just an electronics engineer.

What fate awaits analog IC's? Well, for sure, the ratio analog/digital circuitry will become smaller, we want it or not. Mr. Fourier had no idea how his mathematical work will develop a new branch of electronic processing. How about digital filtering? get a much better filtering just by multiplications and additions! (I already mentioned DSP).

A sensor will keep collecting analog signal but the output, in many cases, is a stream of serial data. Analog IC future? They will become more sensitive, less noisy but... they will have as an output a digital signal. As for analog output...class D amplifiers, PWM are they analog?

Thirty years ago nobody would predict the stage of present electronics. Keep this link for another thirty and compare our discussions with the future reality.

Your sincerely,

1001 1110 0100 0101 1100 (ASCII for Indel)

For example, a digitally generated sine wave is very noisy compared to a natural anolog sine wave.

Good clean sinewave instruments can't be made digitally. Controlling them with a Digital User Interface is another story.

Jon

As has obviously been stated, the world and most everything in it is analog. The digital world was conceived to make it easier for humans to manipulate. The ease comes with numerous inefficiencies due to the conversions A-D then D-A etc. It is my belief that sooner or later man will become intelligent enough to design the analog equivalent of digital technology and regain the efficiency that has been lost. Don't think that it's possible? Look under your hat!

Just My Thoughts!

Analog is insufficient to describe or represent the universe and everything in it... but also digital can't do it either. The universe is not entirely analog.. but things are relative or probabalistic. Do you live on the discrete digital unit called planet earth, or do you probably exist on the bell curve represented blob of matter/energy/quanta called earth? Are you alive? Yes or No? True or False? Black or White? Is the sun radiating upon you? Y/N... both analog and digital are true. it is relative to the question asked, or the system being represented, and the level at which you measure.

Neither will go away. What will happen is that quantum computing will interface to both.

Chris

I would contend that there is no Yes or No, no True or False, no Black or White, but rather a gradient of all possibilities, colored by your particular definition and perspective. Your last sentence seems to refute your own argument. If there are infinite variables then how can you come up with a single 2 state conclusion.

Just My Thoughts!

Just round off the numbers!

I see quantum computing as being able to pose and read an infininite number of questions and answers.... therefore, all possible digital states and analog values and spin, color, flavour, smell, taste, and whatever else is out there...

Big task to ask of any system/technology. Guess I will have to wait to see a computing system pose a question, much less an infinite number of questions. With infinite possibilities all answers are possible and no answer is possible. Sounds like a Do loop with no possibility of exit.

Ref: #18 - TCMTECH

Probably as good as we can hope for, GOOD ENOUGH!

Just My Thoughts!

yes, but the DO..LOOP always returns the same answer.. that is why it is stuck usually.

quantum computing will work more like the human mind... you will use the processes to collect possible right answers, and arrange them on the bell curve. you will probably just sample infinity, lol. Who knows but that Dreaming might be the mind sampling infinity...?

Chris

OP :

And with the recent discovery of the fourth basic analog component, the memristor, it may very well be sooner than later.

Anyone claiming that digital techniques will replace analog techniques has little knowlege of analog. I had a more crude way of saying that, but I decided to be good.

The day that they can take a digital signal and send it over the air at 2.4 GHz (say like IEEE 802.11) without using analog technology, I might applaud then. But I will be dead and gone by that time. I'll leave that as an exercise for the students (aka the young kids).

Bill

Let's look at a commonly held view: "As time goes on, everything except the interface can with advantage be done digitally".

Digital systems chip designers should always ave taken the stability of the power supply (droop, ripple etc.) and clock attenuation into account (not that they always did - but sometimes they got away with it - but less and less as margins reduce).
For about the last 15 years, large digital ICs have internal oscillators and PLLs for synchronisation. Analogue in the time domain.
In order to be competitive, flash memory now uses as many bits/memory-cell as leakage and noise will permit. Gate input threshold Voltage variation will be a major limiting factor in the next generation of semiconductor geometries.

Sure, true analogue is shrinking as a proportion of the whole - but is pure digital getting to be an illusion?