Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

I hit the limit of my capability of understanding electronics

Status
Not open for further replies.

Fluffyboii

Active Member
I reached the peak of my mental capability few years ago. I was doing great at first year of college but as soon as mathematical topics started piling up I started being unable to understand the way of how things work and had to memorize most of the stuff for being able to keep going. But I run out of memory at some point and I started forgetting stuff from the back like a short tape loop. Because I need to keep building on concepts I learn even though past topics are keep being forgotten, the effort I need to put into classes get to a point of being impossible and fruitless.

For example I can remember the gain of CS amplifier with source resistance or without source resistance because it is recently learned. But I always struggled with writing KVL and KCL equations and having the voltage controlled current source of the mosfet small signal model and body effect and ro makes it even harder for me to write a equation with my logic without memorizing. I keep trying to memorize and forgetting and the things I remember are not aligning with what I found myself so I always have to make a %50 bet on which one is right and for some reason it is always not the one I pick. Every source, book solves the questions slightly different and one thing have multiple right and wrong ways to solve it. With the approximations that are going on analog questions answers aren't always what I expect. When things like frequency response adds into topics and capacitors and Miller approximation are involved I just can not progress. The explanation makes sense but applications doesn't for me.

Even if I put hours into calculating stuff according to book correctly, simulations doesn't give me the expected results because there are too many imperfections in place that rigorous calculation becomes meaningless. Op Amps are near perfect and mostly predictable but BJT and Mosfet is just to much for me to understand and design circuits with outside of the textbook examples.

I wanna know how very knowledgeable people here overcome the complexity of these types of electronic circuits. Currently it is hard for me to even understand when I see the answer, can't even think about solving some kind of questions. I boil with hatred while watching my GPA to continuously drop since I keep getting C, B- grades from these EE courses with unrealistic expectations. I accepted the reality of the situation and started playing on passing the classes instead of performing good at them like most people do but one of my electronics teachers have high expectations of me and I don't want to let him down but I am also unable to not do so. The fact that I created the impression of being the electronics guy that builds weird stuff with electronics components put me in this situation. I mean I like painting model planes but that doesn't make me a plane specialist. Same with electronics, I can solder precisely and have some intuition of how stuff works but I can not make rigorous proofs of why they do so. It is like looking at a programming language I don't know like Verilog and somewhat understanding what does what but I can not write a code because I don't exactly know the rules to do so. In a sense I am useless as an engineer. I never wanted to be one but there was/is no other choice.

I just add the first lab report of Analog circuit design that almost caused me to go insane in the progress. The fact that getting 10 gain from a single stage mosfet amplifier being impossible without having a ridiculous high Rd and driving the mosfet into cut off region instead of saturation is something else.
 
My memory is also somewhat limited, so I only remember what I use frequently and look up the rest.
I do have the advantage of a rather intuitive feel for how electronic circuits work, but I think that is something innate rather than learned.
But some of the intuition developed as I worked.
For example, I studied negative feedback in college, but didn't really understand it until I started working with op amps for a design task at work.
There there was a sort of a eureka moment when it all became much clearer.

I'm also not particularly good at math even though I got fairly good grades in the classes by cramming to learn what I needed to pass the classes, but then forgetting a lot of it afterwards.
But for the analog circuit design work I did, I seldom used more than simple Algebra, and never needed to do KVL or KCL analysis or deal with a matrix, or use hybrid transistor parameters.
Calculators and simulators did most of the grunt work for me.

Most of engineering schooling is to expose you to a lot of theoretical knowledge above many different subjects, with only a small amount of that likely being used in the job you will be doing.
And even that knowledge does not usually include all the practical information you will need to actually do your job (so what type of transformer or transistor or diode does this circuit actually need to perform the desired function).

So if you need to stay in engineering, than I suggest you do the best you can in the courses to get good grades, and don't worry about whether you retain it for future use.
When you graduate and get a job (hopefully doing something you like) then you will determine what you need to know to do you job (and hopefully remember where you originally learned it) so you can look up the details to refresh your knowledge.
And of course, these days the internet is a great source of knowledge for just about anything.
 
You break the circuit (or machine or whatever) down to component level & basic functional blocks, the basic parts that you do understand.

Then figure out how they have been combined to make the circuit or machine function.

If you know the basic principles and how different transistor configurations work etc., you can build just about anything up from that.

The same applies with mechanical systems, software etc.
It's like the old joke, "How do you eat an Elephant?" - "One bite at a time".
Don't waste time trying to grasp an over-complex problem in one go - look at all the parts you can follow.

For learning basic circuits, you cannot beat actual components and a multimeter to see voltages and (from voltage across resistors) currents.


When you get stuck, read/study the problem over again then do something else for a while & let your brain work on it. A lot of the time you will come up with an answer after a while anyway.

Same thing with exams; don't waste time on studying any one question more than briefly, just read and go to the next, unless you 100% know the correct answer immediately. After you have read everything, go back and answer the relatively easy ones, then go over again and do the next harder ones etc.
Quite often there are clues to methods or even answers in other questions.

And get used to doing approximate calculations in your head to check ideas out, before doing detailed calculations or messing with simulations - I've never used simulators except for doing stuff on here!
 
On simulators. Not perfect, not worthless.

I love sims (too much maybe) but they are subject to the models the manufacturers
give us, the completeness thereof. A complete model can reveal design and process
secrets, so guess what, we are rarely given complete models.

Also actual circuits have parasitics that we may not incorporate into sim, more
incompleteness.

And sims for RF work are costly, like the Agilent simulator, to manage circuits
best described by wave propagation characteristics. Spice can be improperly used
to try to push the frequency boundary and produce poor results.

But overall I like banging out a quick sim, especially monte carlo analysis of T and V
and component variation to get a rough idea of circuit sensitivities.

Sims have some nuances, its easy to ignore or compromise time step (to get a
fast output) and compromise analysis of non linear behaviors.

Knowing no analysis is perfect since assumed models of devices and actual circuits
not perfect I find they do a pretty good job of predicting overall behavior and values.

With respect to math. I have the same fasciation with math as I do sims. So still
use, occasionally, LaPlace on simple circuits, like OpAmp filters as example. Not
all the time, mostly use tools to get things done. But I was fascinated back in 70's
with doing a 1 pole roll off of gain analysis on input and output Z analysis on
OpAmps, to see how Zout rose as the OpAmp G was dropping. Losing its V
source characteristics. Or stability analysis. Capacitive loading effects. All provided
insight for future work. So in short most EEs do not use a lot of math throughout their
career, unless you are pushing the boundaries of measurement and precision.
Such as in test equipment design.

But the above is not generally used in day to day job. Unless you are working with'
bleeding edge designs, like 28 bit level based A/D measurements. Actually when
one does complete error analysis on 16 bit systems one finds just how difficult
that is over T and V and.....

My one complaint is I think a lot of mixed signal designs EEs are specing precision
and accuracy based on one component, like A/D, and not doing it over entire signal
path and all component errors. Like PSRR, Noise, Offset, crossover distortion, AC
performance, all ADC errors. Here is an ap note I read early on that made me realize
how extensive these problems can be for hi res/accuracy designs, attached. And this
is an old ap note, more considerations spread thru out literature have been found
and discussed.

So to close the long winded diatribe school will show you a touch of many ways
of looking at problems, will result in increased self confidence/willingness to tackle
problem solving, and lessen your fright at being an engineer. Lastly plan on reading
your entire career, and read that which you do not know. Even if only 5 minutes
every night while watching Leave it to Beaver on TV. It will build that database of
knowledge in ones Gerbil capable human mind.

Regards, Dana.
 

Attachments

  • AN280 Mixed Signal Circuit Techniques.pdf
    2.1 MB · Views: 378
Last edited:
Regarding Spice simulations -- I know there are a number of naysayers here that don't particularly like or use them, and it certainly has limitations, such as from poor models, unknown parasitics, etc, but they have helped me greatly in my analog designs, to find design errors, generate a reasonable approximation of the real circuit results in most cases, allow an easy check of what component tolerances do to circuit operation, and allow the view of currents and voltages in a circuit that are essentially impossible to view in a real circuit.
But of course you always need to build the real circuit in the end to verify the circuit operation.
 
One last comment, I found myself at end of Junior year at University
exhausted, struggling with grades (I did not do homework on many
requirements, mostly went into my lab to apply some of what I learned,
which hurt grades). I had gone thru Navy, 3 years of schooling. After
leaving another 3 years with a year to go in EE, and just simply exhausted.
Told wife I was going to quit. 1 week later still could not pull the trigger
on quitting so continued on.

Subsequently leaving school fell in love with learning and have been doing
it entire career. I figure by now I know .001% of the entire worlds engineering
knowledge, and am striving for .002%.

Great career to be a tech or engineer in my humble opinion.

Do not despair, Valhalla waits for your arrival

1668250882090.png



Regards, Dana.
 
God bless you. I faced this type of problem in my university life. Thankfully, now there are many forums like this one, unlike that time.
Having these forums seriously really nice. I took way to much help from here. I want to give something back and help others questions but I am not in that level yet and honestly idk if I ever will be there.
 
On simulators. Not perfect, not worthless.

I love sims (too much maybe) but they are subject to the models the manufacturers
give us, the completeness thereof. A complete model can reveal design and process
secrets, so guess what, we are rarely given complete models.

Also actual circuits have parasitics that we may not incorporate into sim, more
incompleteness.

And sims for RF work are costly, like the Agilent simulator, to manage circuits
best described by wave propagation characteristics. Spice can be improperly used
to try to push the frequency boundary and produce poor results.

But overall I like banging out a quick sim, especially monte carlo analysis of T and V
and component variation to get a rough idea of circuit sensitivities.

Sims have some nuances, its easy to ignore or compromise time step (to get a
fast output) and compromise analysis of non linear behaviors.

Knowing no analysis is perfect since assumed models of devices and actual circuits
not perfect I find they do a pretty good job of predicting overall behavior and values.

With respect to math. I have the same fasciation with math as I do sims. So still
use, occasionally, LaPlace on simple circuits, like OpAmp filters as example. Not
all the time, mostly use tools to get things done. But I was fascinated back in 70's
with doing a 1 pole roll off of gain analysis on input and output Z analysis on
OpAmps, to see how Zout rose as the OpAmp G was dropping. Losing its V
source characteristics. Or stability analysis. Capacitive loading effects. All provided
insight for future work. So in short most EEs do not use a lot of math throughout their
career, unless you are pushing the boundaries of measurement and precision.
Such as in test equipment design.

But the above is not generally used in day to day job. Unless you are working with'
bleeding edge designs, like 28 bit level based A/D measurements. Actually when
one does complete error analysis on 16 bit systems one finds just how difficult
that is over T and V and.....

My one complaint is I think a lot of mixed signal designs EEs are specing precision
and accuracy based on one component, like A/D, and not doing it over entire signal
path and all component errors. Like PSRR, Noise, Offset, crossover distortion, AC
performance, all ADC errors. Here is an ap note I read early on that made me realize
how extensive these problems can be for hi res/accuracy designs, attached. And this
is an old ap note, more considerations spread thru out literature have been found
and discussed.

So to close the long winded diatribe school will show you a touch of many ways
of looking at problems, will result in increased self confidence/willingness to tackle
problem solving, and lessen your fright at being an engineer. Lastly plan on reading
your entire career, and read that which you do not know. Even if only 5 minutes
every night while watching Leave it to Beaver on TV. It will build that database of
knowledge in ones Gerbil capable human mind.

Regards, Dana.
I love that mixed signal circuit techniques start with the explanation of Murphy's Law.
1668355333318.png

This resonates within me.
 
A corollary to that is:
  • A failure will always occur at the worst time, such as just before a demonstration of a system to a customer.
 
On that subject:
The earliest know written equivalent of Murphy's Law, in the "If anything can go wrong, it will" sense, dates from around 500 BC:

"No plan survives contact with the enemy" - Sun Tzu, The Art of War
 
Don't make any large decisions now. You're in the wrong frame of mind. Complete your schooling, no matter the grade.

Unfortunately, understanding is not taught. It can't be, for no one knows the true fundamentals.

But with math, the ability to design and preform a function can be done. Without any understanding. And it works so well, some think that math is a part of nature. Shortly if not already, software can design a function and even manufacture the function. All with math.......and complete lack of understanding.

You will not reach your thinking limit until many years in the future. You are just stressed and worried.

You will feel better after school. THEN......you may search for understanding. Searching for understanding now, only adds to the stress.
 
All good points...

For the most part I am self taught and have been fortunate to have my foot in the right place at the right time on many occasions throughout the years. Now, that said, the ability to teach yourself is really what it is all about. Sure, you "learn" fundamentals in school, but do you really learn it or is it just covered material that meets the curriculum requirement. Think about that for a moment. Education is or should be designed to empower an individual to asses and manage the available resources to work towards a desired goal or solution.

When I was in school, we barely had computers, and there was no internet at least not as we know it today to lookup an answer... we had the library and encyclopedias. ... So what did I do? ... I wrote programs to do my homework for me and present the answer in long hand format. I justified this by having to know and understand the basic fundamentals of the problem in order to write the program. Remember you couldn't go on-line and download a program to do it for you. This worked well in High school and college for both Math and Science applications. I was sure to ask permission from my teachers and professors and in some cases sit down with them and show them exactly what I was doing. I just hated the busy work involved with handwriting the answer, when I felt I could be utilizing my time more efficiently with something else. In some cases it took more than 3 sheets of paper to solve one problem.

So I guess my point and advice is ... hang in there and focus on the fundamentals.
 
Also more or less self taught I started attempting to solve methods / problems with just mechanical mechanisms , if you ever looked inside a complex vintage adding machine ....
gears.jpg

you may appreciate the learning curve required. Move on 10 years electronics takes over, same but different hill to climb, but experience is the best instructor ( not forgetting data sheets ). To the OP experiment, build, make mistakes but just hang on in there .
 
Having these forums seriously really nice. I took way to much help from here. I want to give something back and help others questions but I am not in that level yet and honestly idk if I ever will be there.
Knowledge is built one step at a time. Welcome to the club, you are already a member
with your experiences and pursuit of knowledge.


Regards, one who knows just enough to get his pants on in the morning, Dana.
 
I probably used 1% of what I learnt and forgotten 99.9%. The secret is that you are learning how to learn anything and some will be very useful and the rest you will know where to look if you need it.

Learn to read specs and then design by writing a list of specs and editing as required. A "perfect job" is on-time, on budget and meets all the must-have specs and > 90% of the nice-to-haves.
That's in theory, in practice you need lots of practice to do things fast.

Don't try to reinvent the wheel. If parts of a design exist already. At least you aren't afraid to ask for help. Have fun. Don't sweat the small stuff.

I had a good mentor in my 1st job. Even though I could design a VLF Nav. receiver as my 1st project, I knew nothing about it.
 
I love that mixed signal circuit techniques start with the explanation of Murphy's Law.
View attachment 139267
This resonates within me.
If anything can be reversed or inverted, it will be. ;)

All effects of nature may be included simultaneously, but if you learn how to measure margin to failure, and apply superposition if linear, this allows you to budget for each then test to budget and beyond.

But the critical skill to learn is how to measure margin, non-destructively for electro-magnetic, climatic, mechanical and chemical reactions. ( the origin of the "Schmoo test", I recall was to deviate the clock frequency and supply voltage x% simultaneously in 4 corners. This is in essence the fundamentals of statistical evaluation with multivariable analysis and in Industrial Engineering for solder, there can be a dozen variables each varied +/- 10% to isolate why solder is the biggest defect in manufacturing (reflow or wave) and is solved using Taguchi Methods. with "Design of Experiments" (DoE). This might assume the design of the product is defect-free and was done with DFT, DFM, DFC, DFX, but often something is overlooked.

The time pressures of school are great and work starts with a relaxed pace then builds up with greater experience. But learning how to learn is key to your learning curve.
 
Last edited:
Status
Not open for further replies.

New Articles From Microcontroller Tips

Back
Top