11

In a previous question, it was brought to me that LEDs do not obey the Ohm's law. (See Calculate expected voltage around a resistor)

Simply put: how is that?

What makes them behave so differently? How should we treat them in a circuit and calculations?

Are there other components with similar behavior?

Community
  • 1
Antoine_935
  • 1,144
  • 4
  • 12
  • 22
  • 5
    The non-ideal diode model has an exponential term in it. More importantly, Kirchoff's laws are satisfied, and those always apply. – Matt Young Apr 16 '13 at 19:31
  • @MattYoung just for clarity, the ideal diode has an exponential term, and the threshold model is just a very rough simplification – clabacchio Apr 16 '13 at 19:40
  • Try applying a certain variable voltage to water. What you'll find is that the resistance changes with voltage. Air also doesn't obey Ohm's Law - you've got gigantic voltages floating in the air. But there's almost no current until the voltage reaches a certain level. What you observe then is a spark in the form of a lighting. Ohm's law applies only to resistive materials - by definition. What does not obey Ohm's Law is not a resistor. – Christoph B Apr 16 '13 at 19:45
  • 5
    You have just discovered the principle of linear vs non-linear devices. Expect non-linear behavior from all the semiconductors. – gbarry Apr 16 '13 at 20:48

4 Answers4

17

They do - they just do not have a "fixed" resistance. If you look at it from a standpoint of having a fixed forward voltage drop (which they sort of do - depending on operating region) - look at them more as having a fixed voltage across them. Therefore, as different currents go through them, their voltage will stay (relatively) constant, but the resistance will change.

This is a simplistic answer - but I think you're talking at this level.

Brad
  • 407
  • 1
  • 4
  • 12
  • I am indeed talking at a relatively "newbie" level. But I'm willing to get the details of this. What makes their resistance change? – Antoine_935 Apr 16 '13 at 19:36
  • 5
    Well, it is a _diode_, which is a *semiconductor* which inherently means it does not have a fixed _conductance_, like a normal conductor. The properties of this (and other) semiconductors are complex. They do different things in different operating _regions_. It's resistance is more of an *artifact* of it's operation at any specific point - as opposed to a fixed quantity. See "Voltage-Current Characteristic" here: http://en.wikipedia.org/wiki/Diode – Brad Apr 16 '13 at 19:44
  • Waw, I don't really get all this already, guess I need to study a little more :) thanks for this good answer, it helps anyway! – Antoine_935 Apr 16 '13 at 20:13
  • Definitley look at the Wikipedia article on Diodes. Another good reference: http://www.allaboutcircuits.com/vol_3/chpt_3/1.html – Brad Apr 16 '13 at 20:15
  • 3
    This is not a good answer. If a circuit element obeys Ohm's law, the voltage is *proportional* to the current, i.e., the voltage across is a *linear* function of the current through - full stop. Moreover, this answer conflates the notion of resistance, V/I, and *dynamic resistance*, dv/di. See, for example, http://www.youtube.com/watch?v=QF6V74D2hbY – Alfred Centauri Apr 16 '13 at 21:42
  • 3
    I don't agree. Ohm's Law does not assert that resistance cannot be a function. To deny this means that, for instance, a potentiometer or rheostat do not obey Ohm's Law, because someone can turn the knob. – Kaz Apr 16 '13 at 23:16
  • 3
    @Kaz, linearity and time invariance are distinctly different. You're conflating the two. It they were the same, we wouldn't need to separately specify, for example, linear *time-invariant* system. A variable resistor is, at *any* moment of time, a resistor with resistance that is constant with respect to the voltage across it or the current through it. – Alfred Centauri Apr 17 '13 at 00:42
  • 2
    @AlfredCentauri But so is a diode, at any moment in time. It's like a rheostat, except that the daemon inside the diode which turns the knob is looking at the forward voltage rather than time. – Kaz Apr 17 '13 at 00:49
  • 1
    @Kaz, no at any moment in time, the current through the diode is a non-linear function of the voltage across the diode. For a resistor, variable or not, at any moment in time, the voltage across the resistor is a linear function of current. Think of a 3-D space with voltage, current, and time as axes. Take any constant time slice through that space to get the I-V curve. It is either linear or it isn't. – Alfred Centauri Apr 17 '13 at 00:53
  • Kaz's analogy is good. The rheostat obeys ohms law, even though the daemon in the box is acting in a non-linear fashion. The function of the box can both be non-linear, but yet obey ohms' law. There is nothing (except Wikipedia) which says R has to be a constant. – Brad Apr 17 '13 at 01:56
  • 1
    @Brad, from hyperphysics: "The ratio of voltage to current is called the resistance, and if the ratio is **constant** over a wide range of voltages, the material is said to be an "ohmic" material." – Alfred Centauri Apr 17 '13 at 02:09
  • 1
    @Brad, from Britanica: "Ohm’s law, description of the relationship between current, voltage, and resistance. The amount of steady current through a large number of materials is directly proportional to the potential difference, or voltage, across the materials. Thus, if the voltage V (in units of volts) between two ends of a wire made from one of these materials is tripled, the current I (amperes) also triples; **and the quotient V/I remains constant.**". – Alfred Centauri Apr 17 '13 at 02:11
  • @Brad, you wrote: "There is nothing (**except Wikipedia**) which says R has to be a constant." Evidently, that statement is false. – Alfred Centauri Apr 17 '13 at 02:12
  • 1
    I'm with you up to the point where a non-"ohmic" material "violates ohms law". – Brad Apr 17 '13 at 02:31
  • 1
    This answer is outright incorrect, as it dismisses the very basis of Ohmic relationship, perhaps due to limited conceptual understanding. – Anindo Ghosh Apr 17 '13 at 05:30
16

Ohm's law applies to resistance. All resistive aspects of a device will behave according to OHm's law.

If you invert your question you see that every thing that behaves according to Ohm's law must be a resistor. There is only so much that one can do with pure resistance. So logically the anything that doesn't behave according to ohms law isn't a resistor. Or any thing that isn't a resistor won't behave according to ohms law.

I believe that is called a Tautology.

In circuit design we have many different devices all having unique properties to be able to implement different things/functions.

placeholder
  • 29,982
  • 10
  • 63
  • 110
  • 3
    I think your anwswer needs to be highlighted more as it is the only correct one (at the moment I am writing this). Ohm's law is empirical and was originally derived from obserwing the behaviour of wires of different length. Water doesn't obey Ohm's Law, air doesn't - only conductive materials do, and even then not always. – Christoph B Apr 16 '13 at 19:42
  • 2
    If I had a black box, and ran a current through it - then measured the voltage across it - I could calculate it's resistance at that point in time. It' doesn't matter what is in the black box. – Brad Apr 16 '13 at 19:55
  • @Brad exactly, and then how would you use that single data point to then tell you what would happen at say 2X the current? WIth Ohms law and a resistor you'd be able to say 2X the voltage. You are talking about an equivalent resistance at a single operating point. that does not describe a device. – placeholder Apr 16 '13 at 20:15
  • 2
    Exactly - like I said: A diode does not have a "fixed resistance". I could however argue that it does have a known resistance for a given current. The question was about if it obeys ohms law, and it does. It just doesn't have a constant resistance. – Brad Apr 16 '13 at 20:18
  • @Brad I won't reply any further after this. Just because you CAN map the device to be a variable resistor doesn't mean you should. It doesn't reveal any inherent behaviour and is misleading. You could as easily model the device as a variable capacitor with even less explanatory power. The I vs. V characteristics are THE defining aspect and are fundamental to what the device is. And even in this aspect you're wrong because you should define a device in it's small signal resistance NOT large signal as you have done. You must deal with partial differentials to reconstruct the I/V curve. – placeholder Apr 16 '13 at 20:31
  • 3
    @Brad, it *doesn't* obey Ohm's law - period. For Ohm's law, V and I are *proportional*, i.e., the ratio is *constant*. – Alfred Centauri Apr 16 '13 at 22:16
  • 2
    You are describing the fact the the response is *nonlinear* - and I agree it is not. However that was not the question. By this logic a "variable resistor" would also disobey ohms law. Ohms law defines the relationship between resistance, current and voltage as an equation - proportions to each other. It merely states a change in one will require at least a change on one other to remain valid. You are insisting R must maintain constant while only V and I change for a device. This would describe a device which is linear and purely resistive - but not the only one which would apply. – Brad Apr 16 '13 at 22:46
  • 1
    @Brad, don't you see that you're generalizing the concept of Ohm's law into simply "if you take the voltage and divide it by the current, you get a number"? Of course you do. But there's *more* to Ohm's law than that! Further, by introducing *explicit* time variation, you're muddying the waters. The crucial point here is this: if a circuit element obeys Ohm's law, the ratio of voltage to current *does not depend on the particular voltage or current value*. (cont.) – Alfred Centauri Apr 16 '13 at 23:01
  • 1
    @Brad, (cont.). From Wiki: *Ohm's law states that the current through a conductor between two points is **directly proportional** to the potential difference across the two points. Introducing the **constant of proportionality**, the resistance,[1] one arrives at the usual mathematical equation that describes this relationship... **More specifically, Ohm's law states that the R in this relation is constant, independent of the current.*** – Alfred Centauri Apr 16 '13 at 23:12
  • 1
    "Proportional" - yes. "R is a constant" - nope, it doesn't state that. Why does it "state this" for R and not for V and I. "Proportionality" is akin to "linearity", which I do agree. Wikipedia may state that - but ohms law does not. It's call ohms "law" for a reason. It is universal, not just for purely resistive, nonlinear devices. – Brad Apr 16 '13 at 23:23
  • 1
    Then again, maybe Alfred has a point. For instance, do you still call it Hooke's Law, if you have a spring that has a non-linear force with respect to elongation? – Kaz Apr 16 '13 at 23:26
  • It's a "law" - you can't violate it! – Brad Apr 17 '13 at 00:44
  • 1
    But, say, Hooke's law can be violated. Any elastic thing whose elongation/compression is not proportional to the displacing force does not obey Hooke's spring law. – Kaz Apr 17 '13 at 00:52
  • 1
    So it's a matter of word semantics. To me, resistance is an important quantity; it's more than just some number that pops out when you divide voltage by current. It has meaning. – Kaz Apr 17 '13 at 02:54
  • 1
    @Brad Perhaps you need to re-read Ohms law, and see why it doesn't apply to **non-Ohmic** (yes, google that term) devices. That might provide some clearer fundamental understanding of the law, rather than a debate based on applying specific state-variants. – Anindo Ghosh Apr 17 '13 at 05:28
4

Simply put, because the're not resistors but p-n junctions, and because of that their V-I ratio is exponential.

It doesn't mean that you can't calculate their current, just that it's not as simple as for resistors. For instance, you can treat them with a threshold model, with a fixed voltage drop. Then the current will be set by external resistors or active components.

The LEDs are diodes, so that's the obvious similarity. Also the base-emitter junction of a bipolar transistor is a diode, and behaves similarly. The only difference with diodes is that their threshold voltage is higher due to the different materials and doping.

clabacchio
  • 13,481
  • 4
  • 40
  • 80
4

A lightbulb, on first examination may not appear to obey ohms law. Measure its resistance with a multimeter and it might be 5 ohms. Connect it to a power supply capable of illuminating it and measure current and voltage and its resistance will have considerably risen (maybe 20 or 30 ohms). Its still a resistor but its resistance changes with power delivered to it.

A light dependent resistor is another example - its resistance changes with incident light - it's still a resistor and obeys ohms law - but it takes a little bit more than a linear volt-current graph to figure things out.

Andy aka
  • 434,556
  • 28
  • 351
  • 777