Analog vs Digital Engineers

[AlphaOmega]

I work with several people who are either an analog engineer (signal chain or power) or a digitial engineer (usually microcontrollers and DSP's) and there seems to be an Us vs Them attitude. Many people that do not deal with analog on the daily basis, which is the case of many digital engineers I talk to, seem to think analog is antiquated and/or useless. I even had a person tell me that they do not deal 'with that analog crap anymore, it's junk.'
What the hell! How else are you going to interface the real world to a computer?
If digital is nothing more than a 1 or a 0 (on/off, high/low) then what is my light switch when I flip it on and off? Digital? Nobody is going to say that my light switch is a digital device.
What about microcontrollers and DSP'? Sure they are 'digital devices' but they are made up of discrete components;i.e., transistors, resistors, caps, etc. Those are not considered digital.
Wouldn't this suggest that digital is nothing more than an implementation of analog components;i.e., petrified analog?
Besides, why would I pay a software guy $45 an hour to code all sorts of calculus functions in C when I can slap a 50 cent op-amp with a resistor and capacitor?

Viva la analog!

[GrandMasterTerwynn]

I work with several people who are either an analog engineer (signal chain or power) or a digitial engineer (usually microcontrollers and DSP's) and there seems to be an Us vs Them attitude. Many people that do not deal with analog on the daily basis, which is the case of many digital engineers I talk to, seem to think analog is antiquated and/or useless. I even had a person tell me that they do not deal 'with that analog crap anymore, it's junk.'
What the hell! How else are you going to interface the real world to a computer?
If digital is nothing more than a 1 or a 0 (on/off, high/low) then what is my light switch when I flip it on and off? Digital? Nobody is going to say that my light switch is a digital device.
What about microcontrollers and DSP'? Sure they are 'digital devices' but they are made up of discrete components;i.e., transistors, resistors, caps, etc. Those are not considered digital.
Wouldn't this suggest that digital is nothing more than an implementation of analog components;i.e., petrified analog?
Besides, why would I pay a software guy $45 an hour to code all sorts of calculus functions in C when I can slap a 50 cent op-amp with a resistor and capacitor?

Viva la analog!

Is there a point to all this, or are you just venting? And if so, what in the world is it doing in SLAM?

And yes, we all know that, at the very heart of a digital processor are a collection of analog components on the nano-scale, transistors, resistors, capacitors and so on. And only the most ignorant digital hardware engineers are ignorant about the analog side of the fence. After all, when you crank up things fast enough, you suddenly have to worry about things like clock jitter, signal noise on inputs or on the power and ground planes, capacitance on inputs, refresh times for memory cells, power consumption and dissipation.

Of course, moving away from the board design/hardware level, a properly designed digital system offers high levels of precision, speed, and noise-tolerance that purely analog implementations lack. Not to mention, these days, you can have a program a single DSP, microcontroller, or FPGA to perform very different tasks, and consume relatively less power while doing it. So while analog has its place, it's not quite the end-all you seem to be making it out to be.

[AlphaOmega]

Is there a point to all this, or are you just venting? And if so, what in the world is it doing in SLAM?

Partial vent, partial observation. There seems to be a rift between the two. Analog expertise seems to be on the decline as many students are pursuing a digital career. Since it is about engineering, wouldn't this be the best place for this post?

And only the most ignorant digital hardware engineers are ignorant about the analog side of the fence. After all, when you crank up things fast enough, you suddenly have to worry about things like clock jitter, signal noise on inputs or on the power and ground planes, capacitance on inputs, refresh times for memory cells, power consumption and dissipation.

I don't know. I have talked to many people on the digital side of the fence and this hardly comes first to their minds. Stray capacitances between planes and traces, inductances due to trace lengths, horrendously chopped up ground planes always seem to cause trouble.

Of course, moving away from the board design/hardware level, a properly designed digital system offers high levels of precision, speed, and noise-tolerance that purely analog implementations lack. Not to mention, these days, you can have a program a single DSP, microcontroller, or FPGA to perform very different tasks, and consume relatively less power while doing it. So while analog has its place, it's not quite the end-all you seem to be making it out to be.

True, there are things that are better served by a digital solution, hell, why do people still use a 555 timer when a simple, inexpensive RTC can be used?
Also, because of typical 3.3V I/O and 1.8V core on modern DSP's (TI and ADI are the two that come to mind) many analog components; especially op-amps and dataconverters are going lower and lower for supply volages.
There is an article about a German company trying to push an analog signal processor rather than the common place DSP. I will have to try to find it.

But without high precision analog how else are you going to interface to the real world? Each has their place but like I said in my original post, many people see it as antiquated and useless which it surely is not.

[Darth Wong]

I work with several people who are either an analog engineer (signal chain or power) or a digitial engineer (usually microcontrollers and DSP's) and there seems to be an Us vs Them attitude. Many people that do not deal with analog on the daily basis, which is the case of many digital engineers I talk to, seem to think analog is antiquated and/or useless. I even had a person tell me that they do not deal 'with that analog crap anymore, it's junk.'

If he's talking about controllers, he's generally right. Even manual high-precision machinery has used various devices to digitize cartesian co-ordinates for positioning for decades. If you want precision, you need to convert to digital at some point because analog is just too prone to loss of calibration. Obviously, if he's talking about power systems he's just blowing hot air.

What the hell! How else are you going to interface the real world to a computer?

The same way you interface a mouse to a computer: by digitizing the analog input immediately. That's the same way CNC machines work.

If digital is nothing more than a 1 or a 0 (on/off, high/low) then what is my light switch when I flip it on and off? Digital? Nobody is going to say that my light switch is a digital device.

Not if it's a dimmer switch, but if it's a simple on/off switch, then it's a digital device by definition. What are you talking about?

What about microcontrollers and DSP'? Sure they are 'digital devices' but they are made up of discrete components;i.e., transistors, resistors, caps, etc. Those are not considered digital.

Yes they are. "Digital" electronics are simply those that deal with signals that have discrete quantized states. The fact that they employ physical transistors doesn't change anything.

Wouldn't this suggest that digital is nothing more than an implementation of analog components;i.e., petrified analog?

Are you a fucking retard? Seriously, I have never seen someone spout such idiocy about the very simple concept of digital vs analog.

Besides, why would I pay a software guy $45 an hour to code all sorts of calculus functions in C when I can slap a 50 cent op-amp with a resistor and capacitor?

Viva la analog!

You honestly don't understand why we started using reprogrammable computers instead of hard-wired circuits?

[Braedley]

Any digital electrical engineer that says they never (or almost never) deal with analouge signals is kidding themselves. DSP sometimes deals with sampling analouge signals, then digitizing that sampled signal.

Right now, I'm working on a software EQ for a large scale analouge system. Please tell me how that is not both analouge and digital. In fact, all my work, save the final implementation, will all be on the computer with fully discrete data to an analouge system. Personally, I would find it easier to finish my project in full analouge, despite my appreciation of what a computer can do for me.

[Darth Wong]

A power amplifier is a good example of a totally analog device that wouldn't benefit at all from digital circuits. On the other hand, anyone who tried to build an analog controller for a precision milling machine would be the mother of all idiots. But that has nothing to do with AlphaOmega's retarded assertion that digital systems aren't really digital if they employ discrete transistors.

[aerius]

A power amplifier is a good example of a totally analog device that wouldn't benefit at all from digital circuits.

Actually there's a new generation of digital power amplifiers from Panasonic (the XR50 for instance) and a few other companies which sound pretty darn good while being far more efficient than analogue transistor amps. It's not going to beat a top end Krell amp but it'll give the entry to mid-level Brystons a good run for the money.

[Darth Wong]

A power amplifier is a good example of a totally analog device that wouldn't benefit at all from digital circuits.

Actually there's a new generation of digital power amplifiers from Panasonic (the XR50 for instance) and a few other companies which sound pretty darn good while being far more efficient than analogue transistor amps. It's not going to beat a top end Krell amp but it'll give the entry to mid-level Brystons a good run for the money.

You know, I had heard about these things a while ago but totally forgot about them. Thanks for reminding me.

[AlphaOmega]

The same way you interface a mouse to a computer: by digitizing the analog input immediately. That's the same way CNC machines work.

True, but the analog signal is coming from 'the real world' in which a mixed-signal device like an ADC would be utilized.

Not if it's a dimmer switch, but if it's a simple on/off switch, then it's a digital device by definition. What are you talking about?

I would assume a dimmer switch would have something like a potentiometer involved. Move the wiper, change the resistance and therefore change the voltage drop.
But a simple on/off switch is not a digital device by definition and its not even dealing with a digital signal. Where do you get this from?
When the switch is opened (off) there is 0V ideally. I flip to switch (on) and an AC voltage gets rectified to a DC level in a lamp etc (15V I am guessing -- I confess I do not know how much voltage a bulb needs).
You are saying that this on/off nature is digital? Why? Because it simply is on or off?
The on/off voltage levels may be discrete, but they are certainly not quantized in any fashion. A digital signal is discrete and quantized. That on/off voltage level may be discrete (on/off, 0V/15V etc.) but it still has infinite precision like an analog signal. A digital signal would have a finite precision...want more? Get more bits.

Yes they are. "Digital" electronics are simply those that deal with signals that have discrete quantized states. The fact that they employ physical transistors doesn't change anything.

Now take what you said here and read above. How is a light switch a digital device? It's not because the on/off signal (voltage level) since the leves do not represent a discrete quantized signal. Discrete...yes. Quantized...no.

As an aside, saying that digital is petrified analog is my POV that I like to put forth to stir up the hornets nest. I know the difference between an analog and digital signal. I should have conveyed it better.

Are you a fucking retard? Seriously, I have never seen someone spout such idiocy about the very simple concept of digital vs analog.

Here's the problem. Your say that a digital signals deal with discrete quantized states AFTER you argue that a simple light switch is a digital device. Tell me what I am missing...discrete voltage levels? Check. Quantized states...uuhhhhh none that I see. Clue me in here, chief.

You honestly don't understand why we started using reprogrammable computers instead of hard-wired circuits? [/quote]

I was being facetious. Part of my post was serious so I can see why someone could take it this way.
But that integrator is going to remember its calculus faster than anyone else can cut and paste! [/b]

[Darth Wong]

True, but the analog signal is coming from 'the real world' in which a mixed-signal device like an ADC would be utilized.

So? How does that change the fact that it's trivially easy to interface digital devices with the real world, despite your implication to the contrary?

I would assume a dimmer switch would have something like a potentiometer involved. Move the wiper, change the resistance and therefore change the voltage drop.

But a simple on/off switch is not a digital device by definition and its not even dealing with a digital signal. Where do you get this from?

The definition of the word "digital", dumb-ass.

When the switch is opened (off) there is 0V ideally. I flip to switch (on) and an AC voltage gets rectified to a DC level in a lamp etc (15V I am guessing -- I confess I do not know how much voltage a bulb needs).
You are saying that this on/off nature is digital? Why? Because it simply is on or off?

Yes. The secret to the concept of the word "digital" is difficult to find, but you may be able to locate it if you search in secret wellsprings of arcane knowledge known as "books".

The on/off voltage levels may be discrete, but they are certainly not quantized in any fashion. A digital signal is discrete and quantized. That on/off voltage level may be discrete (on/off, 0V/15V etc.) but it still has infinite precision like an analog signal. A digital signal would have a finite precision...want more? Get more bits.

You're an idiot; digital signals are logically quantized, not physically quantized. In a 5V digital circuit, we take a signal with a voltage anywhere from 3V to 7V and treat it as if it's equal to 5V. And you do not get "infinite precision" by using analog devices; that is utterly retarded and indicated that you do not understand what "precision" means. You get functionally infinite variability, but that is not the same as infinite precision.

Yes they are. "Digital" electronics are simply those that deal with signals that have discrete quantized states. The fact that they employ physical transistors doesn't change anything.

Now take what you said here and read above. How is a light switch a digital device? It's not because the on/off signal (voltage level) since the leves do not represent a discrete quantized signal. Discrete...yes. Quantized...no.

Do you honestly not understand the concept of logical vs physical quantization? Or are you trying to be dense on purpose?

As an aside, saying that digital is petrified analog is my POV that I like to put forth to stir up the hornets nest. I know the difference between an analog and digital signal. I should have conveyed it better.

In other words, you are saying something stupid in order to troll for reactions. Smarten up, asshole. Either you're trolling (which is a violation of online etiquette and makes you an asshole) or you're an idiot and you're just pretending otherwise in order to backpedal from a dumb-fuck statement.

Here's the problem. Your say that a digital signals deal with discrete quantized states AFTER you argue that a simple light switch is a digital device. Tell me what I am missing...discrete voltage levels? Check. Quantized states...uuhhhhh none that I see. Clue me in here, chief.

Your inability to buy a clue on your own is not my problem.

I was being facetious. Part of my post was serious so I can see why someone could take it this way.

But that integrator is going to remember its calculus faster than anyone else can cut and paste! [/b]

That integrator's calculus is limited to such a pitifully simple hard-wired single function that it can hardly be considered calculus.

[AlphaOmega]

Awww hell! Sorry Michael, I typed up a nice long response but apparently I was not signed in so it asked me for my password. I hit the back button but apparenlty I missed it or it doesn't save (I assume it should be the first option). I will have to come back later and respond.

Off to bed.

[aerius]

I would assume a dimmer switch would have something like a potentiometer involved. Move the wiper, change the resistance and therefore change the voltage drop.

Not anymore. These days they're potentiometer controlled triacs, with the pot setting the conduction curve of the triac. This then sets the amount of time that the triac will conduct on the AC cycle. The dimmer doesn't work on voltage, it works on a primitive form of pulse width modulation based on the 120Hz AC cycle, which is why an AC dimmer won't work on DC. DC dimmers work strictly on PWM, a semi-conductor switch turns the power on & off a few thousand times a second, varying the length of the turn-on time to control the brightness. Completely digital.

[WyrdNyrd]

Yeah, wouldn't a potentiometer-based dimmer switch not only be wasteful of power, but also hazardous? The excess heat would be dissipated in the switch, at the very least burning your finger when you turn the sucker off.

[aerius]

Yeah, wouldn't a potentiometer-based dimmer switch not only be wasteful of power, but also hazardous? The excess heat would be dissipated in the switch, at the very least burning your finger when you turn the sucker off.

Yep, huge, wasteful, and toasty.

Also, my previous post should say 60Hz AC, I was thinking voltage instead of frequency.

[AlphaOmega]

True, but the analog signal is coming from 'the real world' in which a mixed-signal device like an ADC would be utilized.

So? How does that change the fact that it's trivially easy to interface digital devices with the real world, despite your implication to the contrary?

When did I ever imply that interfacing to the real world is difficult? I said: “What the Hell! How else do you interface to the real world?” The point being, many digital engineers that I have spoken to consider analog old and bunk. But high performance analog is a billion dollar a year industry and is quite needed to interface a dumb, deaf, mute DSP to the real world, isn’t it? Bad eisegesis on your part my friend.

I would assume a dimmer switch would have something like a potentiometer involved. Move the wiper, change the resistance and therefore change the voltage drop.

But a simple on/off switch is not a digital device by definition and its not even dealing with a digital signal. Where do you get this from?

The definition of the word "digital", dumb-ass.

Before we go any further, there are two parallel arguments developing here. Basically, one is “What is a digital device?” and the other is “What is a digital signal?”

When the switch is opened (off) there is 0V ideally. I flip to switch (on) and an AC voltage gets rectified to a DC level in a lamp etc (15V I am guessing -- I confess I do not know how much voltage a bulb needs).
You are saying that this on/off nature is digital? Why? Because it simply is on or off?

Yes. The secret to the concept of the word "digital" is difficult to find, but you may be able to locate it if you search in secret wellsprings of arcane knowledge known as "books".

In context of the light switch, you are balls out incorrect. The light switch is indeed an analog device. Never mind the nature of its output signal; I am talking about the device itself. If you are going to be ever-so-brazen and say that a simple light switch in my house is a digital device because I can turn it on and off (make it a high or a low), then why is a comparator considered to be an analog device when it can do the exact same thing (on/off, high/low)? Why is an op-amp considered to be an analog device when it too can serve the exact same function of my light switch? Put the damn thing in open loop and what happens with a given input of 0V and then any other input voltage level. The output will be 0V and then saturate to the rail even with a slight input. Hell, why is a god damn transistor considered an analog device (component) when one of its basic functions can be a SWITCH?

Check your wellsprings of arcane knowledge lately? You know, “books”?

The on/off voltage levels may be discrete, but they are certainly not quantized in any fashion. A digital signal is discrete and quantized. That on/off voltage level may be discrete (on/off, 0V/15V etc.) but it still has infinite precision like an analog signal. A digital signal would have a finite precision...want more? Get more bits.

You're an idiot; digital signals are logically quantized, not physically quantized. In a 5V digital circuit, we take a signal with a voltage anywhere from 3V to 7V and treat it as if it's equal to 5V. And you do not get "infinite precision" by using analog devices; that is utterly retarded and indicated that you do not understand what "precision" means. You get functionally infinite variability, but that is not the same as infinite precision.

Keep your mind grounded in signal processing and try not to let it venture off in some other area of engineering for the time being. Two things here:

One, this is where we get into the second argument. What is a digital signal? Honestly, I have never heard of “logical” or “physical” quantization when it comes to sampling an analog signal. I assume that when you say “physical quantization” you are speaking in terms of a physical level of a voltage, current, etc; i.e., 1V, 2V, 5V and “logical quantization” is a term to mean a HIGH or a LOW similar to that of threshold voltages for a given logic family (given your example 3V to 7V meaning 5V). If not, please clarify.

I will clarify the terminology as I see it. A discrete signal is sampled from a continuous signal and those individual samples still have an infinite **precision. You then quantize those samples so each sample is represented by a certain N number of bits. **Precision being (again, from a signal processing perspective) the number of digit or bits a number can be represented. That is exactly why I said “need more precision – get more bits.” Too bad you didn’t catch it. To cure your dogmatic approach to definitions, try your secret stash of arcane knowledge. You know, “books.”

So, if an output of our light switch is 15V when it is on and then 0V when it is off what is that? It certainly is a continuous signal on the output. Is it sampled to a discrete signal? No. But maybe it can be considered something like a 1-bit ADC and ‘already sampled’ so it can be viewed as ‘automatically’ discrete. One ‘sampled’ value of infinite precision. This is the case for delta-sigma ADC’s which have a comparator inside which is said to be a 1-bit ADC. Others argue that a digital signal is a set of 1’s and 0’s that represent an analog signal; i.e., the digital codes from the output of an ADC But hey, 1 bit is enough, right? Not everyone agrees and this is why the line between analog and digital can be blurry.

I will concede that the output of our light switch can be viewed as a type of ‘digital transmission’ but the light switch certainly is not a digital device as you specifically stated.

As an aside, saying that digital is petrified analog is my POV that I like to put forth to stir up the hornets nest. I know the difference between an analog and digital signal. I should have conveyed it better.

In other words, you are saying something stupid in order to troll for reactions. Smarten up, asshole. Either you're trolling (which is a violation of online etiquette and makes you an asshole) or you're an idiot and you're just pretending otherwise in order to backpedal from a dumb-fuck statement.

No, no, and I don’t appreciate the accusation. It is a fun statement to get peoples attention, nothing more. Besides, there is nothing wrong with taking a step back to make a clarification and that certainly is not ‘backpedaling.’

But that integrator is going to remember its calculus faster than anyone else can cut and paste! [/b]

That integrator's calculus is limited to such a pitifully simple hard-wired single function that it can hardly be considered calculus.

Nice nitpick. -1/RC (integral) Vin(t) dt for a simple inverting integrator.

[Darth Wong]

When did I ever imply that interfacing to the real world is difficult? I said: “What the Hell! How else do you interface to the real world?” The point being, many digital engineers that I have spoken to consider analog old and bunk. But high performance analog is a billion dollar a year industry and is quite needed to interface a dumb, deaf, mute DSP to the real world, isn’t it? Bad eisegesis on your part my friend.

And again you ignore the point, moron. A digital controller is designed to interface with analog inputs; this does not make it any less digital, nor does it eliminate the argument that a digital controller has advantages over an analog one.

Before we go any further, there are two parallel arguments developing here. Basically, one is “What is a digital device?” and the other is “What is a digital signal?

That is your hair-splitting, not mine. I'm talking about digital CONTROLLERS, ie- entire systems. This is a distinction I made in my first post, which you apparently didn't read. Analog vs digital is technically a distinction of how components are used, not what their ordering numbers are in a parts catalogue. The fact that some components are considered "analog" is not due to the fact that they are intrinsically analog, but the fact that they are traditionally used in analog systems. You are confusing convention with definition.

Yes. The secret to the concept of the word "digital" is difficult to find, but you may be able to locate it if you search in secret wellsprings of arcane knowledge known as "books".

In context of the light switch, you are balls out incorrect.

In the context of exact terminology, YOU are balls-out incorrect.

The light switch is indeed an analog device.

Because you say so? Your argument is that it could not possibly be considered a digital device, when in fact it does fit the definition and you have failed to show otherwise. That's because you don't even seem to think in terms of proper definitions, and instead confuse convention with precise terminology.

Never mind the nature of its output signal; I am talking about the device itself.

A device which has only two operating modes: open and closed. Yet again you reveal that you are an imbecile who doesn't distinguish between convention and technical accuracy. A light switch is technically a digital device; people just normally don't give a shit about its precise classification because it's such a mundane item. Need I remind you that it is still common practice among electricians to treat electricity as if it flows from positive to negative, even though the reverse is actually true?

If you are going to be ever-so-brazen and say that a simple light switch in my house is a digital device because I can turn it on and off (make it a high or a low), then why is a comparator considered to be an analog device when it can do the exact same thing (on/off, high/low)? Why is an op-amp considered to be an analog device when it too can serve the exact same function of my light switch? Put the damn thing in open loop and what happens with a given input of 0V and then any other input voltage level. The output will be 0V and then saturate to the rail even with a slight input. Hell, why is a god damn transistor considered an analog device (component) when one of its basic functions can be a SWITCH?

Hey moron, do you realize that a computer CPU is nothing more than a huge pile of transistors? Does that mean that you think computer CPUs are analog devices too? It's the way it's used, you fucking half-witted hemorrhoid. A computer CPU's transistors switch from 0 to 1, whereas the transistor in a power amplifier cycles continuously to produce an analog signal.

GROW A FUCKING BRAIN; a device is not intrinsically analog or digital; it depends on the nature of its output signal and how it's used. Transistors are NOT necessarily analog devices; they can be either analog or digital. Do you honestly think that the definition of analog vs digital has something to do with whether the device is discrete or built into an IC? Are you truly that fucking ignorant? What IS your definition of analog vs digital? Do you even have one?

Check your wellsprings of arcane knowledge lately? You know, “books”?

Yes I have. Unlike you, I even know how to read them.

Keep your mind grounded in signal processing and try not to let it venture off in some other area of engineering for the time being. Two things here:

One, this is where we get into the second argument. What is a digital signal? Honestly, I have never heard of “logical” or “physical” quantization when it comes to sampling an analog signal.

Obviously because your physics knowledge is pitiful. Electronic systems essentially DECIDE that a signal exists in discrete states even though it doesn't. In a physically quantized system such as the quantized packets of energy in electromagnetic physics, the energy actually exists in separate, discrete physical packets, rather than just being interpreted that way by a device. Get it now? Or would you like some sort of crayon depiction?

I assume that when you say “physical quantization” you are speaking in terms of a physical level of a voltage, current, etc; i.e., 1V, 2V, 5V and “logical quantization” is a term to mean a HIGH or a LOW similar to that of threshold voltages for a given logic family (given your example 3V to 7V meaning 5V). If not, please clarify.

I will clarify the terminology as I see it. A discrete signal is sampled from a continuous signal and those individual samples still have an infinite **precision.

No, they don't. The precision of any sampling is limited by the device which performs the sampling; the very idea of infinite measuring precision of any kind is idiotic and indicates that you have ZERO knowledge of the subject. It is not even theoretically possible.

You then quantize those samples so each sample is represented by a certain N number of bits. **Precision being (again, from a signal processing perspective) the number of digit or bits a number can be represented. That is exactly why I said “need more precision – get more bits.” Too bad you didn’t catch it. To cure your dogmatic approach to definitions, try your secret stash of arcane knowledge. You know, “books.

Next time you try to hurl rhetoric, make sure you're actually not talking out of your ass first, dumb-fuck. The concept of analog to digital conversion does NOT make the original analog signal an example of "infinite precision". This is like saying that all analog microphones are perfect and that the only conceivable inaccuracies come from analog to digital conversion.

So, if an output of our light switch is 15V when it is on and then 0V when it is off what is that? It certainly is a continuous signal on the output. Is it sampled to a discrete signal? No. But maybe it can be considered something like a 1-bit ADC and ‘already sampled’ so it can be viewed as ‘automatically’ discrete. One ‘sampled’ value of infinite precision. This is the case for delta-sigma ADC’s which have a comparator inside which is said to be a 1-bit ADC. Others argue that a digital signal is a set of 1’s and 0’s that represent an analog signal; i.e., the digital codes from the output of an ADC But hey, 1 bit is enough, right? Not everyone agrees and this is why the line between analog and digital can be blurry.

You have obviously generated your definitions of "digital" vs "analog" from the section headings on electronic supply house component ordering books rather than examining the actual proper definition of what it means to be digital. As I said, confusing industry convention for proper definitions.

I will concede that the output of our light switch can be viewed as a type of ‘digital transmission’ but the light switch certainly is not a digital device as you specifically stated.

Yes it is. It fits the definition of a digital device, because it only has two distinct states. You have presented no contradiction whatsoever to this definition; you simply ignore it and state that you must be right anyway, because you know that it's not found in the "digital systems" section of a parts supply catalogue.

No, no, and I don't appreciate the accusation. It is a fun statement to get peoples attention, nothing more. Besides, there is nothing wrong with taking a step back to make a clarification and that certainly is not backpedaling.

I don't give a flying fuck whether you appreciate the accusation.

But that integrator is going to remember its calculus faster than anyone else can cut and paste! [/b]

That integrator's calculus is limited to such a pitifully simple hard-wired single function that it can hardly be considered calculus.

Nice nitpick. -1/RC (integral) Vin(t) dt for a simple inverting integrator.

How the fuck is it a "nitpick" to point out that a simple op-amp integrator has as much to do with real calculus as a four-function calculator has to do with a fucking supercomputer?

[R. U. Serious]

So, you are trying to make a point that no matter how much things get digitzed, there will always be a need for analog stuff in some ways, for example when interfacing with the "real world". Fair enough, I don't think anybody really would disagree with that statement.

But then you go and wrap it in rhetoric and trolling to get people riled up (yes, in the online world that is trolling; in a verbal discussion with a handful of people, the same behaviour is often referred to as "fun statements" - but a face-to-face discussion with a handful of people, is a different setting that an online-discussion). In the end you come across just like the "digital-loonies" that you meant to criticize.

I mean, really, what is your objective when babbling things like "Wouldn't this suggest that digital is nothing more than an implementation of analog components;i.e., petrified analog?"

You complain about the "us vs them" attitude of the other "tribe", but then go on to defend your "tribe" and somehow make them turn out to be superior. Please... Light a few candles at your analog appreciation altar and get over your inferiority complexes...

[AniThyng]

As a computer engineer in training, I feel compelled to point out that one does not get by knowing only digital systems or only analog systems, we need both, and I don't feel much of a us versus them mentality, because really, if what I need for my fancy DSP microcontroller system is a switching power supply, then by god there needs to be a switching power supply or we're not going anywhere.

[Darth Wong]

I wonder if AlphaOmega realizes that you could make a digital circuit with op amps, or that conversely, you could make an analog IC chip.

The idea of trying to see which one is universally "better" is retarded in the extreme; it's a question of which one is appropriate for any particular task.

[drachefly]

Uh, I think you just restated AlphaOmega's main point.

[Darth Wong]

Uh, I think you just restated AlphaOmega's main point.

How so? He thinks that certain components such as discrete transistors are intrinsically analog, and that analog devices have "infinite precision", in order to exaggerate the glories of analog.

He seems to say this out of a reaction to digital circuit designers who are contemptuous of analog circuits, but that's largely new thinking vs old thinking; the fact is that digital circuits are superseding older analog ones for a shitload of functions for a reason, specifically their superior flexibility and the absence of signal loss once the information has entered the digital domain. There are still areas where an all-analog system works so you can't say one is universally better than the other, but there are reasons for the general shift toward digital, especially for controllers.

[drachefly]

Looking at the OP, it seems that his main point was that analog and digital systems complement each other.

Later, he definitely allowed that there are many jobs that digital devices are best for -- a position you denied he held only a few posts ago.

I agree with your position, and the position I think he was trying (badly) to get across.

[Darth Wong]

Looking at the OP, it seems that his main point was that analog and digital systems complement each other.

Later, he definitely allowed that there are many jobs that digital devices are best for -- a position you denied he held only a few posts ago.

I agree with your position, and the position I think he was trying (badly) to get across.

Actually, I think he doesn't even know what the difference between "digital" and "analog" is. If you look at his OP it seems to be more of a rant against certain individuals he knows who express contempt for analog systems, but in many industries that contempt would be justified.

[Durandal]

Besides, why would I pay a software guy $45 an hour to code all sorts of calculus functions in C when I can slap a 50 cent op-amp with a resistor and capacitor?

Yeah, programming languages are totally useless next to glorious analog circuits. Tell you what, I'll give you a program I wrote in C, and you can convert it into an analog circuit for me. It's only a few hundred lines of source code; I'm sure it'll be no problem for genius such as yours.

[aerius]

That integrator's calculus is limited to such a pitifully simple hard-wired single function that it can hardly be considered calculus.

Nice nitpick. -1/RC (integral) Vin(t) dt for a simple inverting integrator.

Do you even know what such an integrator does? It's a voltage sampler used in servo circuits for correcting DC offset on circuit outputs. If I punch in "4 Pi R squared" it's not going to give me "4/3 Pi R cubed".