Ohms Law.

Straight_ManStraight_Man Geeky, in my own wayNaples, FL Icrontian
edited February 2004 in Hardware
Would any of you folks like to discuss how Ohms Law is used to evaluate PSUs and trace faults???? Understanding that is key to isolation of problems that are not software. Discriminating between soft and hard errors\defects\faults is the one hardest thing to do.

I suspect we have engineer level folks here, in the hands-on variety, and a discussion like this is something that can be summarized in simple ways. But getting the grokking of Ohm's Law is a learning stepping stone to really help enable finding out intellectually how to parse and then discriminate between hard and soft faults. Also, understanding Ohm's Law and how computers bus things is a key to know what hardware parts may be failing with soft "failures" that keep reoccuring on only one or a few boxes if they are not traceable to malware.

I know, I lacked that deep grokking understanding for half my career.

John.

Comments

  • mmonninmmonnin Centreville, VA
    edited February 2004
    V=ir
    I=v/r
    R=v/i

    P=iv
    P=v^2/r
    P=i^2r
  • Straight_ManStraight_Man Geeky, in my own way Naples, FL Icrontian
    edited February 2004
    Right, but how do they interrelate??? And why bother to derive resistance from Amperage and Voltage?? And when do you want to do this???

    See here for something you might find fun, and others might be able to understand interactions better with:

    http://www.the12volt.com/ohm/ohmslaw.asp

    Basicly, folks who understand Ohm's Law need to know when to use it, how to get the measurements to use it right (resistance is not easily measurable HOT directly(base logic says NOT to measure it directly), but is logic-derived mostly if you have a multimeter and not an ocilloscope plus analysis software that can calc this and display this).

    But increased resistance beyond wanted resistance is usually core of problem (other major fault cause is too much or two little power to part that has faulted, but that can be directly measured if you know where to look starting from sysmptoms and how to drill down with minimum time and effort), and lack or poor ground is increase of resistance, and wrong ground route is an implicit short. So, understanding Ohm's law in the way non-engineering folks think can be valuable for younger folks and folks who are not hardware engineers.

    Lets talk about how to apply it also, from gross generality to howto by computer function....

    How is power eval like a river ecosystem??? How can circuit branching be like trees of deciduous kind,and why are there so many detailed diffs as there are between a random forest of all species of tree would have, from tree to tree???

    How do you know when the problem is software and when it is more likely to be hardware and which hardware part to swap out or stick in another box??? How can knowing the gross relationships (larger rules explicated by Ohm's Law) help decide these things????

    John D.
  • mmonninmmonnin Centreville, VA
    edited February 2004
    Something sometimes more useful if you dont have all the measurements is CDR and VDR - Voltage/Current Divider Rule.

    Ix=(IR)/Rx
    Vx=Rx(V/R)

    Very helpful if you dont know the total current, its in parallel, or you dont know the total Voltage in parallel.

    I am getting towards the end of my AC/DC class so I know how to calculate OHMs Law, CDR, VDR, Power, resistance given the resistor properties, Capacitance given its properties, inductance given its properties, voltage drop across a resistor (VDR), current thru a resistor (CDR), impedance of Capacitors and inductors, RMS and peak values in AC circuits, phasor(complex form) and time domain forms, and current, voltage, resistance, capacitance, and inductance in parallel and series circuits. Man I learned ****ing ****load 11 in weeks already. DAMN I didnt realize until I wrote it all out.

    So what would you like to know.:)

    Nice link.:)
  • qparadoxqparadox Vancouver, BC
    edited February 2004
    Just wait mmonnin, the real fun stuff starts when you get into the frequency domain and get to use various transforms. 3 Phase power is also really interesting. The best part about transforms won't ever have to solve the constant coefficient second order linear ODE the long way ever again :P.
  • mmonninmmonnin Centreville, VA
    edited February 2004
    Yeah I know, the book we are using is called "Contemporary Electronic Circuits" so I can tell there is much more than whats in that book. I dont really need to know all that for what I want to do. If I end or knowing it then looks like more knowledge for me.:)
  • TheBaronTheBaron Austin, TX
    edited February 2004
    modeling capacitance and inductance over time is kind of interesting, really everything just gets more complicated from where you are. the thing i've always found interesting is that no matter how much more sophisticated the concepts get, the thing about electronics and circuitry is that it always makes sense. meh, dont get the EE in me started (wait til im done with the degree for that)
  • botheredbothered Manchester UK
    edited February 2004
    Don't forget resonance. Just look what it's done to Mr Cosby there.
  • TheBaronTheBaron Austin, TX
    edited February 2004
    no i think cosby is just stuck in an infinite loop. he has a CS condition, not something easily fixed by a EE
  • edcentricedcentric near Milwaukee, Wisconsin Icrontian
    edited February 2004
    One problem with testing PSUs and other devices is that they are not simple DC devices. They have voltage regulators and filters and as a result often have very 'non-linear' behavior.
    I see PSU testing done under load, this is a good start. But people use a pure restive load, maybe not too realistic. The talk about voltage drop, this is good too. But they never mention the waveform or harmonic noise involved. I would rather that the filters on my mobo didn't have to work too hard.
    The real (and imaginary) world can get so messy.
  • mmonninmmonnin Centreville, VA
    edited February 2004
    What do you mean by waveform? DC has no waveform, its a constant.
Sign In or Register to comment.