Now what

Discussion in 'General Electronics Chat' started by J0ker, Apr 18, 2012.

  1. J0ker

    Thread Starter New Member

    Apr 18, 2012
    17
    0
    Hi!

    I'm learning electronics as I hobbie. Maybe in the future I'd like to go into some kind of robotics (i'm a programmer) or IA...

    Long story short, I've studied and understood the basic laws of electronics, same with basic components (capacitors, transistors and so on). Now that i know all of this, what should I do know? I find myself not able to understand the purpose of a circuit when I see a schema.

    How can I start from simple circuits to more complicated ones to start grasping the real application of components? I have the multisim software to try and investigate things by myself, just need some guidance I guess.

    Thank you!
     
  2. panic mode

    Senior Member

    Oct 10, 2011
    1,320
    304
    you can sim, double sim, multi sim... but until you make real circuit, you didn't do anything.
     
  3. WBahn

    Moderator

    Mar 31, 2012
    17,751
    4,797
    I wholeheartedly agree with panic mode. The theory and sims are nice and valuable and all that, but until you get your hands dirty and start letting the smoke out of some devices and burn your fingers a few (dozen) times, you haven't really learned much at all.

    I don't know what is available today, but when I was a kid perhaps the single most influential gift I ever received was the Radio Shack 65-in-1 electronics kit. The thing, I think, was intended for high school age kids and I was only about eight and I couldn't grasp most of it, but I built a lot of circuits that I was able to follow and learned how to read schematics (from the standpoint of implementing them) and learned a lot by making changes to the schematics in the book and seeing what happened. Over the course of a few years, I slowly investigated the thing into oblivion as my experiments sent various components to never-work-again land.

    The ideal thing for you (and I don't have any specific suggestions) would be to find a set of schematics for a project box that had a bunch of little circuits that could be wired together to make interesting things and start by wiring up the little circuits and getting them to work.
     
  4. @android

    Member

    Dec 15, 2011
    178
    9
    Simply start building circuits yourself. Search for small electronic projects on google.
    Look around, in your home you'll find there is need of some circuits...say water levels indicator for your water tank, may be a door bell with multiple tones, burglar alarm, headphone stereo amplifier etc.
     
  5. MrChips

    Moderator

    Oct 2, 2009
    12,446
    3,362
    Reading a schematic diagram is like learning a new language. It is a skill in communication that can only be developed through lots of practice. There is no short cut.

    Start off by doing simple things.

    1. Connect a resistor, LED and battery together and make the LED light without blowing it. Vary the LED and observe the brightness vs current. Do the calculations first.

    2. Make the LED flash using a PNP and NPN transistor and a few resistors and capacitor. Do the calculations.

    3. Make the LED flash using a 555 circuit. Do the calculations.

    4. Build a crystal radio.

    ... just for starters.
     
  6. russ_hensel

    Well-Known Member

    Jan 11, 2009
    818
    47
    If you are at all serious get some test and prototyping equipment. A scope is, to me, the most important peice of test equipment. You can use you computer ( with a sound card program ), but it is very limited and can damage your computer. I have a nice 100 mhz analog scope that I got locally on ebay for under 50. bucks. You can find a list of test equipment at: http://opencircuits.com/Test_Equipment_and_Other_Equipment and on this forum,
     
  7. J0ker

    Thread Starter New Member

    Apr 18, 2012
    17
    0
    Thanks to all the responses. Actually I bought a few items (a little training board, resistors, leds and so on to mess a little bit). As a kid i learnt to solder for a couple tech projects for school, I did it pretty good and i'd say i still have the touch :p

    I definately need to practice from simple to complex, anywhere to find simple circuits to start understanding? I'll try them and try to understand what they do and how.
    I found this site: http://www.aaroncake.net/circuits and this http://gregsbasicelectronics.com/
    More suggestions?

    And also, I found this super interesting link that has helped me a lot understanding what really goes on: http://www.falstad.com/circuit/

    Also, why is it so different trying things on Multisim than on real life? I think it's saving me a lot of time so in the end I get to try and know more things, wouldn't you say? (specially because i'm moving in a few months but right now i don't have space for really work well with scopes and so on...i'll have to wait for that probably)

    Thank you so much!!
     
  8. Wendy

    Moderator

    Mar 24, 2008
    20,766
    2,536
    Because life is much more complicated than simulations. Many simulations have to use some shortcuts, it shows up. There are always negligible factors that aren't. They show up.

    Unless you wrap your head around the real thing in practice, you will be lacking some basic skills. Least of which is tracing wiring around. This sounds minor, but it isn't.

    I have never used a simulator, but I am very good at designing and drawing electronics. Only 40 years of practice got me here.
     
  9. WBahn

    Moderator

    Mar 31, 2012
    17,751
    4,797
    Like Bill Marsden said, its because the pretty ideal world of a simulator and the real world we live in are miles apart in complexity. Yes, a simulator is a very useful tool, but it is most properly used when you are using it in conjunction with real world experiences (or with circuits that you already have some level of experience with).

    I taught an Electronics II (simple differential and multistage transistor amplifiers, for the most part) course about fifteen years ago and was asked to teach it two weeks before classes started, so I pretty much had to just adopt the prior instructor's syllabus. It had three design projects, each of which asked the students to design a circuit, simulate it, and write a report. It was obvious that the students had put in a lot of time and effort, but it was even more obvious that they had no clue about what the simulations were telling them or if/how the circuit might actually work. So the next time I taught the class, I developed a slightly different, but quite similar, set of projects but required that the students design a circuit, simulate it, build it, test it, demonstrate it, and then write a report. The first project was very simply (took me two hours to dream it up, design it, build it, and get it working) while they had six weeks. It was a blood letting and no one came particularly close to getting something working. But when I showed them in class how simple the solution was, from scratch, it really sank home to them that they clearly weren't learning the material like they needed to. By the end of the third project, which was considerably more difficult, every team got it working and was excited and trying out different ideas from class and using real engineering thinking.
     
  10. crutschow

    Expert

    Mar 14, 2008
    13,018
    3,235
    I've also had a lot of experience (about 47 years) in designing electronic circuits, mostly analog but some digital, and the difference between simulators and real life generally isn't as large as some have implied. I would never build a circuit, even a simple one, without simulating it first (and I haven't since simulators became available). Usually the simulation is quite close to the actual circuit operation, and in many cases revealed errors in my design before I built it. The exceptions are when the models are inaccurate or the circuit parasitics have a significant effect on circuit operation, such as in high frequency circuits.

    (One important place to add a few tens of pF of stray capacitance during simulations is at the summing node of an op amp. This can significantly affect the frequency response and peaking (ringing) of an inverting op amp configuration, especially if the circuit resistances are high).

    So I believe simulation is a good and relatively easy way to become familiar with how different circuits operate. You can easily probe the various circuit voltages and currents to help understand their operation, something not always readily done with a real circuit (especially currents).

    But you do need to have some understanding of the circuit to know when the simulation is giving you reasonable results and when it is not (such as showing a node voltage of 1E99 volts for example). Often when that occurs it's because there is some error in the circuit or your simulation conditions.

    Of course at some point, when you get the circuit operating properly in the simulation, you need to build it to verify its operation and get some experience with a real circuit.
     
  11. @android

    Member

    Dec 15, 2011
    178
    9
    That is really nice experience. When you taught something to someone....the least you expect is the enthusiasm & true efforts.
    Yes making them think like real engineer was really great. Beacause personally I think....'Teaching is an art that can only be mastered by a STUDENT'.
    And of course, unless you dirty your hands...you can't get insight of resistor bands!
     
  12. WBahn

    Moderator

    Mar 31, 2012
    17,751
    4,797
    That was certainly what I had expected when I first starting teaching as a graduate lab TA twenty years ago. But moving from the physics department (where students have no realistic ambition of ever making lots of money and are in it for the love of learning how the world works) to the engineering department was a true culture shock for me. I was dumbfounded when I quickly discovered that there really do exist engineering students who actively loath engineering and everything about it, particularly problem solving. They are there solely to check off the minimum number of boxes and collect a degree under the (mis)impression that with an engineering degree automatically goes all kinds of large salary offers. When I first encountered it, I wrote it off as just a fluke specific to the two in question, but I quickly discovered that attitudes like that are closer to the norm than the exception, a trend that only seems to be getting worse. So when I encounter students (and there are a handful in each class, usually) that do want to learn and understand, I happily bend over backwards to shepherd them along. But what really turns my crank is when I see a group of students that are very disengaded and disinterested become interested and engaged. I've seen it happen a dozen or so times to individual students, but I can only really say that I've seen it happen to a significant portion of an entire section of a course only two, maybe three times.
     
  13. panic mode

    Senior Member

    Oct 10, 2011
    1,320
    304
    with all due respect - i beg to differ.... a lot... ;)

    if simulation WAS giving results that are way off, there would be no point to using simulators.

    but that is not the point. in real life, real circuits are used. nobody relies on simulator to start their car or watch their home when they are away or defrost meal or surf internet.

    to put it differently: i don't care how many western movies somone watched, i still don't believe they can ride a horse and shoot the rope around someone else's neck with a single bullet.

    moral:
    simulators are designed to be forgiving. real life is not. thinking you know it is one thing, KNOWING you know - is completely different.

    if you do bad circuit design in simulator, and you KNOW what to look for, you are good. but if you do the same and DON'T know what to look for, you may never find out what was wrong.

    simple example: simulators, even if they can calculate dissipated power of a device, are hardly going to jump into your face with a message "R10 is overheating", because unless you specify resistor power and check for it, simulators are likely to ignore it. i have yet to see simulator that INSISTS that ALL information is provided (including power rating of resistors and voltage of capacitors). without EVERY bit of information, simulation is incomplete and can be deceiving - specially for inexperienced who does not know what to expect.

    but when you go to build a circuit and you power it and it SMOKES... now THAT is real experience. i can say with confidence that NOBODY using simulator will feel NEARLY as bad or guilty about picking wrong component as someone who actually fried (possibly one of the last few) real components at hand.

    Even if the cost of the smoked component is insignificant (say $0.01 resistor) and one DOES have hundreds of the same ones at hand at the moment, that one burned resistor teaches you to be more careful in a way that no simulation ever can.

    i see 100s of students graduate as EE. some of them did some real projects (they used soldering iron, they know how to recognize and read components, they know what PCB is etc.), most didn't. to put all of them in the same basket and say they are equally ready of executing project is wrong.
     
  14. WBahn

    Moderator

    Mar 31, 2012
    17,751
    4,797
    The accuracy of a simulator is dictated, primarily, by the quality of the models used. Also, most simulators are not charge conserving. A decade or so ago, the computational requirements for having sufficiently complex models was too high and so compromises had to be made. I remember getting a model from National Semiconductor for an opamp and putting it in a circuit. The results, including input bias currents, were quite reasonable. I then wanted to measure how much total current was being consumed from my 9V power source and got something like 84A. All of that was being consumed by the op amp. After checking that I hadn't done anything wrong, I contacted National and, after a bit of phone chasing, spoke with someone that help develop the models. He stated, unapologetically, that the models were only intended to provide reasonably good results at the I/O pins and nothing more. When I asked him how a designer was supposed to determine power consumption, his answer was to build the circuit and measure it.

    I frankly have no idea if this has changed or not, but I suspect it has given both the natural tendency over time to improve models and the drastically increased computational power available to the simulator.

    An area where the models are exceptionally good, usually, are the device models for IC designers, and they have been for a long time. When I first started working as an ASIC designer I was amazed at how close the real circuit performance was to the simulation results. But when doing an IC, you can't prototype it and spinning a design, even twenty years ago, was many thousands of dollars (and today it can easily be many hundreds of thousands of dollars). But the complexity of the device models has exploded. On the last chip design I worked on (about three years ago), the model for a transistor was a subcircuit with over three hundred components in it -- and the device library has different models for the typical, slow, and fast process corners and different models for room temperature and cryogenic regions.

    Even then, we frequently had to struggle with significantly different results for certain simulations, such as noise sims, depending on the type of integration method was specified for the simulator and other parameters.

    As for the rest of your post, we are in strong agreement. In the second project in the course I talked about earlier, they had to design a power amplifier that could deliver 5W to an 8ohm load and the only supplies the lab had available was a 30V/1A fixed supply. Despite the fact that I carefully went through all the thermal considerations and calculations in class and told them how applicable this was to their project, it was obvious that few paid much attention. A transistor is a transistor, after all. So most of them used 2N3904/2N3906 small signal transistors for their output stages instead of the 2N3055 (and can't remember the other one I got them) power transistors and even those that did didn't put on the heatsinks I provided. Similarly, they used little 1/4 W resistors for their collector resistors instead of the power resistors I had purchased. They let the magic smoke out of a lot of transistors and resistors, let me tell you! But one night a student came to me and complained that his output transistor was getting really hot. I asked him what device he was using (expecting him to say the part number for one of the small signal transistors) and he looked at his thumb and said, "2N3055". He had touched the top of the transistor with his thumb and branded the part number into it!

    On the next exam, the class as a whole did amazingly well on the thermal management questions, whereas the prior year's students (who, academically, were far better students) did rather poorly. So the burned hand really does teach best.
     
  15. MrChips

    Moderator

    Oct 2, 2009
    12,446
    3,362
    I had a chat with National Semiconductor's analog guru the late Bob Pease and I gathered he wouldn't go anywhere near a simulator.
     
  16. MrChips

    Moderator

    Oct 2, 2009
    12,446
    3,362
    I see it is the same where ever you are. The aptitude and attitude have been deteriorating each year. The best we can do is to motivate students by having them work through and design real world meaningful applications. When they can actually get their hands wet (or burnt as you say) they seem to be more motivated.
     
  17. panic mode

    Senior Member

    Oct 10, 2011
    1,320
    304
    @WBahn

    what i meant was:

    "if simulation WAS giving results that are way off, all the time, there would be no point to using simulators"
     
  18. WBahn

    Moderator

    Mar 31, 2012
    17,751
    4,797
    @ panic mode

    Oh, I agree. As I stated in my first post on the issue, simulators are a valuable and important tool and I certainly wouldn't have said that if I believed they were way off, even a significant fraction of the time. My point is that they do not capture all of the complexity of the real life circuit (not without a lot of serious modelling on everyone's part) and therefore only give you a rather narrow, albeit highly informative, glimpse into the circuit's behavior.

    Bringing everything together, I would say that simulators are pretty good at what they are designed to be pretty good at, but that they are not designed to be pretty good at everything that would ideally be included in the simulation. So the designer is left with the onus of keeping in mind the limitations of the simulation and not expecting it to provide information beyond what it is reasonably capable of. That awareness and, even harder, knowledge of where those boundaries are is very difficult to acquire except through a lot of hands on experience seeing where the simulations stop producing decent approximations to the real world.
     
  19. crutschow

    Expert

    Mar 14, 2008
    13,018
    3,235
    To each his own. He also drove an ancient VW bug and you know what happened to that.
     
  20. panic mode

    Senior Member

    Oct 10, 2011
    1,320
    304
Loading...