1
\$\begingroup\$

I'm currently in a first year circuit analysis course. While the concepts are relatively simple they don't seem like something you could use to design circuits.

I mean it's hard enough to analyze circuits that you already have. But I feel like it would be impossible to work backwards to figure out the circuit you need given the values you want.

So, as an electrical engineer working on designing circuits, how much of the circuit analysis techniques taught in a first year course (Nodal/Loop Analysis, Thevenin/Norton Theorems, Steady State Analyis of RC and RL circuits, AC analysis etc.) are actually used in the design of circuits?

\$\endgroup\$

closed as primarily opinion-based by Scott Seidman, The Photon, placeholder, Andy aka, Joe Hass Apr 16 '14 at 10:39

Many good questions generate some degree of opinion based on expert experience, but answers to this question will tend to be almost entirely based on opinions, rather than facts, references, or specific expertise. If this question can be reworded to fit the rules in the help center, please edit the question.

  • \$\begingroup\$ Depends on the application really. Usually just (having learned all this stuff) just make educated guesses or instantly by looking at it understand how it should work, and order of magnitudes etc. \$\endgroup\$ – KyranF Apr 16 '14 at 2:09
  • 3
    \$\begingroup\$ Circuit analysis is part of EE fundamentals; it is a part of the foundation upon which you erect your knowledge and skills to practice circuit design. Practicing circuit analysis helps develop your EE intuition. Your design skills come later and generally with the help of a mentor. \$\endgroup\$ – Alfred Centauri Apr 16 '14 at 2:23
  • 1
    \$\begingroup\$ Do you use the alphabet and a working knowledge of grammatical construction to design sentences? Do you think about it when you do it? Same here. \$\endgroup\$ – Scott Seidman Apr 16 '14 at 12:24
  • \$\begingroup\$ Nah, that's all stuff just to keep EE undergrads busy. When you get to the real world and are designing the control system for a nuclear power plant, you just pick whatever values come to mind and hope it all works later. Just make sure you're not downwind when they first fire up the reactor, and you'll be OK. \$\endgroup\$ – Olin Lathrop Apr 16 '14 at 14:12
3
\$\begingroup\$

Actually, there are usages to circuit analysis, even beyond pure EE fundamentals:

  1. Review a schematic from your colleague in a final design review;
  2. Reverse engineer an older circuit produced by a retired engineer which was sloppy at documenting its designs (True story during an internship...);
  3. Reverse engineer open-hardware design to learn new patterns/circuits;
  4. Understanding side-effects experienced by your design when you have parasitic resistors/capacitors affecting your circuit;
  5. Troubleshooting mistakes in digital circuits: a colleague had an internal pull-down that he inadvertently activated. Understanding circuits explained the cause of the remnant ~0.4V signal that we saw on oscilloscope without wasting hour debugging code and PCB;
  6. Understanding what a circuit that you designed really does. It is not always possible to implement a transfer function literally to components, but with experience, you know what will happen if you add a resistor/capacitor at a given location and if not, you can analyze it quite fast (to guess behavior) prior performing extensive calculus.
\$\endgroup\$
2
\$\begingroup\$

I find myself using filter design stuff I learned in college quite a bit. The other thing that getting good at circuit analysis does is give you intuition. After you work enough problems, you can look at a real circuit and spot potential problems. I think that's more true in analog design than digital, but it still helps in being able to point to a spot in the schematic and tell if the current is going to be really high or really low and if that's bad or not, etc. RC stuff also comes in handy when figuring out how much decoupling capacitance you need on a chip. None of the stuff you'll see in real life is as nicely set up as textbook problems of course, but first year EE circuit analysis is still worthwhile to master!

\$\endgroup\$
2
\$\begingroup\$

Analysis is a fundamental skill. You need analysis to understand and to hone and tweak circuit designs.

Circuit design is synthesis which is pretty much the opposite of analysis.

Design (except in the simplest cases where you're simply finding parameters for a known configuration) is inexact, and consists of constituting a workable solution from parts. Analysis is (literally, lysis means "to separate") breaking down a circuit into constituent parts.

You can test analysis by comparing with reality, whereas I don't think it's possible to prove a non-trivial design is optimal, only that it's sub-optimal if a better one is presented.

When you can break down previously encountered circuits into parts, and reassemble the parts to make new designs, you can apply other reasoning methods such as analogy. This SQUID control circuit you need to design this month may resemble in some way an LF receiver you designed years ago, and you can apply tricks or caveats learned from that experience. A purely mechanical technique might have some analog at GHz frequencies. Plenty of innovation has come from re-applying old ideas in a new way, and in fact that is considered novel and patentable in many cases. But if you don't really understand the old idea (which requires analysis), you're not going to be able to abstract it and re-use it in a new domain.

\$\endgroup\$
0
\$\begingroup\$

Every analysis tool is like a Chess move you can master. The "best tools" are the ones you know how to use. Once the theory is solid, shortcut tools can be trusted. Designs start with awareness of requirements and how to convert them into functional specs,. These need no design implementation at first, Just specs for inputs and outputs , environmental stress factors , cost, time and go to it. Researching prior methods comes quickly when you can recognize what works, needs changes and then improve to exceed requirements.

Analog defines the world of all activity between a one and a zero, the latency, transmission line effects, and all physical effects of insulators, conductors, static charges, DC power, AC transmission, RF communication, Medical sensors. Then you can add smarts with digital processing power, sensors based on physics, Systems on a Chip and create better tools.

Commodity products will end up in China or India, but there is a huge world of advanced technology emerging in many fields, from more efficient electric transport , more advanced medical imaging of live cancer cells, safer smart grid technology, more sophisticated communication devices, smart detectors for predicting earthquakes, Carrington Effects, measuring condition based monitoring to prevent transformer arc failures ..from 10kV to 1.5 gigavolt.

Each of these requires the foundations of experiment and Physics discovered by Faraday, Ampere, Gauss, Weber, Ohm, Siemens, and Maxwell. From there you can either go into, Research, Applied Research, (design), Test Engineering, Software Development, System Design, or become a Business Developer, raise millions and build teams form a company to do the same. ( as many of my colleagues did.. BNR, Symbol, ImRIS, Cubresa, ... )

\$\endgroup\$
-1
\$\begingroup\$

No, circuit analysis is never really used directly, at least in my experience. Basic circuit analysis concepts have serve two purposes, first, they give hiring managers good questions for interviews, and second, form a basis of understanding for much of EE. Accordingly, ignore them at your own considerable peril.

In truth, these concepts are applied, but usually one of two things happens. Either the design in question is so complex that it requires vastly more than basic circuit theorems to understand, or the design is so simple that the use of the theorem is practically the only analysis needed.

\$\endgroup\$
  • \$\begingroup\$ Just curious, why do you disagree? \$\endgroup\$ – dataBus Apr 17 '14 at 0:14
  • \$\begingroup\$ Because there were already answer from other people stating that there were usages to circuit analysis, so adding an answer that explicitly tells that it is useless looks like you didn't read other people's answers first. \$\endgroup\$ – Mishyoshi Apr 18 '14 at 16:18

Not the answer you're looking for? Browse other questions tagged or ask your own question.