Designing Analog Circuit

Thread Starter

N|kolαTeslα

Joined Sep 17, 2017
19
I dont know if its just only me. When I graduated college I realize that I have skills in designing digital circuits but not in analog. Often, analog circuit designs for our projects came from the internet and we only modify the circuit to output what we what but our professor dont taught us to actually design a analog circuit without references although we know the basics. All we do is to modify and analyze not like in digital that you have steps in designing. I want to be familiar in designing analog circuits and even I know the basics I cant totally imagine how to design. I search in google but all pop ups are all about CMOS. Where do I start ?
 

Thread Starter

N|kolαTeslα

Joined Sep 17, 2017
19
hi,
When you say analog circuits, that covers a wide range of topics, could you be more specific.?
ie: Basic theory, transistors, OPA's etc.??
E
Well, I mean in books like Boylestad / ffloyd its all about analysis. What I mean is, Is there any pattern how they design these circuits ?
 

Sensacell

Joined Jun 19, 2012
3,432
Designing Analog is like learning a foreign language.

Analog components are the alphabet of this new language- you learn the alphabet first.
Study all the basic components, learn what they do.

Circuit topologies are the words- study books containing circuits, after a while you will start to recognize certain function blocks, by form alone.
These circuits are all made up from the basic components, as words are made up of letters.

Sentences are combinations of words, conveying meaning.
When you master the function blocks, you then learn how to connect them together to achieve the desired objective- words into sentences.

After a while, you will be able to read a complex circuit diagram the same as you would read a book. When you get really good, you can write your own books, but it takes years to get really good.
 

Thread Starter

N|kolαTeslα

Joined Sep 17, 2017
19
Designing Analog is like learning a foreign language.

Analog components are the alphabet of this new language- you learn the alphabet first.
Study all the basic components, learn what they do.

Circuit topologies are the words- study books containing circuits, after a while you will start to recognize certain function blocks, by form alone.
These circuits are all made up from the basic components, as words are made up of letters.

Sentences are combinations of words, conveying meaning.
When you master the function blocks, you then learn how to connect them together to achieve the desired objective- words into sentences.

After a while, you will be able to read a complex circuit diagram the same as you would read a book. When you get really good, you can write your own books, but it takes years to get really good.
Thank you, so I need really to know a lot of blocks. haha.
 

DickCappels

Joined Aug 21, 2008
10,153
Interesting. One of my life-long friends was very good with digital electronics but could not understand, as he said "how to get from 0 to 1". He was always much better with digital than I. Maybe it is a predisposition to one kind of thinking or another.
 

OBW0549

Joined Mar 2, 2015
3,566
I want to be familiar in designing analog circuits and even I know the basics I cant totally imagine how to design. I search in google but all pop ups are all about CMOS. Where do I start ?
Start with circuit theory: the basics of DC and AC circuit analysis. Without a solid foundation in circuit theory, you'll get nowhere because you'll be forever limited to merely copying designs created by other people, and at the most making minor modifications to them.

@Sensacell raised a good point, in that much of analog design involves adapting and combining some number of "functional circuit blocks" into a complete system tailored to achieve a specific goal. Such circuits as inverting amplifiers, non-inverting amplifiers, integrators, differentiators, differential amplifiers, instrumentation amplifiers, transimpedance amplifiers, summers, diode breakpoint limiters and shapers, high-pass, low-pass, band-pass and band-reject filters, sample-and-hold circuits, limiters, precision rectifiers and many, many more, appear again and again in most circuit designs. Master those functional block circuits by learning how they work and how to modify them for specific purposes, and you'll be well along in learning analog design.

Study well-designed circuits, especially those that appear in the various semiconductor manufacturers' application notes. Texas Instruments, Linear Technology, Analog Devices, Maxim, Microchip and others have excellent collections of application notes, many of which touch on general circuit design with only minimal discussion of specific chips. Read them. Study them. Study, study, study. (Beware of hobby stuff you find on the Internet; too much of it is amateurish junk suitable only for show how NOT to design circuits. Be careful.)

There's one other element that's a bit hard to define: I call it "vision" for lack of a better term-- the innate ability to synthesize something new from scratch. Over the course of my 50+ years in electronics I've come to the conclusion that, for whatever reason, some people just have it, and some people don't-- and it doesn't seem to have anything to do with intelligence. You're either a designer, or you're not.
 

#12

Joined Nov 30, 2010
18,224
Designing Analog is like learning a foreign language.
Best description I've ever seen!
the innate ability to synthesize something new from scratch. some people just have it, and some people don't--
Also a critical condition.

50 years ago, digital was still in diapers, but analog designs were all around me, so I chose analog. It took me 4 years at 40 hour jobs (in electronics) to say, "I can read a schematic faster than I can read English." Analog really is another language, and all the rules of learning a second language apply. Synthesizing a unique analog solution (a new word) is like standing on a mountain top screaming, "I am a creator!"

Yes, it's that difficult and it's that rewarding. Are you going to be an analog wordsmith?
 

ebeowulf17

Joined Aug 12, 2014
3,307
Texas Instruments, Linear Technology, Analog Devices, Maxim, Microchip and others have excellent collections of application notes, many of which touch on general circuit design with only minimal discussion of specific chips. Read them. Study them. Study, study, study. (Beware of hobby stuff you find on the Internet; too much of it is amateurish junk suitable only for show how NOT to design circuits. Be careful.)
I feel like every website that discusses electronics in any way should have this quote (or something very similar) as a banner or header.

So many misleading, distracting, confusing bad designs are out on instructables, in the arduino forums, on YouTube, etc. What's especially bad about all these terrible circuits is that they're rarely trying to accomplish anything novel. More often than not, whatever chip or component they use in their defective project has datasheets and/or application notes that already show how to do the exact same thing properly.
 

xox

Joined Sep 8, 2017
838
Designing Analog is like learning a foreign language.

Analog components are the alphabet of this new language- you learn the alphabet first.
Study all the basic components, learn what they do.

Circuit topologies are the words- study books containing circuits, after a while you will start to recognize certain function blocks, by form alone.
These circuits are all made up from the basic components, as words are made up of letters.

Sentences are combinations of words, conveying meaning.
When you master the function blocks, you then learn how to connect them together to achieve the desired objective- words into sentences.

After a while, you will be able to read a complex circuit diagram the same as you would read a book. When you get really good, you can write your own books, but it takes years to get really good.
Yes, as a beginner I'm just starting to realize the truth in the above. It's frustrating, but it's just going to take some time to master this art!
 

OBW0549

Joined Mar 2, 2015
3,566
More often than not, whatever chip or component they use in their defective project has datasheets and/or application notes that already show how to do the exact same thing properly.
Not only do the datasheets and appnotes show how to do it right, they're well written and readily available on the 'net. Why people don't use them, is beyond me.
 

Thread Starter

N|kolαTeslα

Joined Sep 17, 2017
19
There's one other element that's a bit hard to define: I call it "vision" for lack of a better term-- the innate ability to synthesize something new from scratch. Over the course of my 50+ years in electronics I've come to the conclusion that, for whatever reason, some people just have it, and some people don't-- and it doesn't seem to have anything to do with intelligence. You're either a designer, or you're not.
Well, I dont care if I was a born a designer or not. What I only know is I want to be designer.
 

Thread Starter

N|kolαTeslα

Joined Sep 17, 2017
19
Best description I've ever seen!
Also a critical condition.

50 years ago, digital was still in diapers, but analog designs were all around me, so I chose analog. It took me 4 years at 40 hour jobs (in electronics) to say, "I can read a schematic faster than I can read English." Analog really is another language, and all the rules of learning a second language apply. Synthesizing a unique analog solution (a new word) is like standing on a mountain top screaming, "I am a creator!"

Yes, it's that difficult and it's that rewarding. Are you going to be an analog wordsmith?
Well, I think I can be one but not yet there. I dont think about the reward, its just my hobby.
 

Sensacell

Joined Jun 19, 2012
3,432
Another important concept that comes to mind is "Rough Order of Magnitude" thinking.

While it's essential to master the mathematics of electronics, be careful to avoid "calculator thinking"- the idea that there is an equation for everything and all you need to do is find the right numbers and plug them in. I see this as a fatal, inescapable trap for beginners.

Remember- in analog, there is never an exact anything, it's all a continuum. The more you can learn to think like this, the better designer you will be. Understand the concepts at a high level, but don't fixate on exact numbers, this is why the slide rule calculator is (was) such a great tool for design, it inherently reinforces this concept.

All things analog have error - tolerances, drifts, tempco's, offsets and many other gremlins to contend with.

For example, my favorite question is this one: "how do I calculate the correct value for a pull-up resistor for my I2C bus?"

Well, you can't, there is no exact value, it's pointless to think this way.
Instead, you have to understand that the final resistor value encompasses a large set of compromises, none of which can can be exactly satisfied without considering the impact to the others. Get a handle on what the parameters are, the inherent limits, work from there toward your solution. Understand that some of these things are very 'meta', like choosing a resistor value because it's already on your BOM, a very rational and logical choice.

Get comfortable thinking about the ROM of things, this is the "big picture" that helps you optimize and find creative solutions in the real world.
Align your thought process to the problems at hand, think like an analog computer.

Every great analog designer I have ever heard of thinks this way.
 

ericgibbs

Joined Jan 29, 2010
18,766
Remember- in analog, there is never an exact anything, it's all a continuum. The more you can learn to think like this, the better designer you will be. Understand the concepts at a high level, but don't fixate on exact numbers, this is why the slide rule calculator is (was) such a great tool for design, it inherently reinforces this concept.
hi,
I would agree with Sensacell and say also that 'everything' we do in our lives is only an approximation, sometimes a very rough one, but good enough to get us by.

Most of the world's designers and engineers make decisions based on an approximate answer, sometimes it bites us in the butt.

E
 

Attachments

OBW0549

Joined Mar 2, 2015
3,566
For example, my favorite question is this one: "how do I calculate the correct value for a pull-up resistor for my I2C bus?"
Yeah, that's a good one. Knowing when getting the "correct" value for a component is important, and when an arbitrary choice within reasonable bounds is perfectly acceptable, is an important part of being a designer as it keeps you from wasting time and effort.

My "favorite" questions are the ones that could be answered by a simple Google search-- especially the ones where the search key is blatantly obvious and the most direct answer appears at the very top of the search results. I'm continually amazed at the number of people who will go to the trouble of setting up an account here, logging in, and posting a question that could be answered in less than a minute with a simple Internet search.
 
Top