I am an Engineer. No, not all the Universities teach that R cannot vary. It's typical to teach a simple version of a physical law in the introductory courses, something that my Physics professor would rage over. What do you mean I won't listen to anyone else? Lots of people agree with my I've written on both of the threads that deal with this matter. Just because I know someting doesnt' mean I don't listen. You have a way of making statements about things you don't know about.

__________________
You don't need a quadraphonic Blaupunkt -- you need a curve
ball.

 

Look I'm just happy that no-one is objecting to the statement:
1. in most homogeneous materials (i.e. no junctions), for all practical purposes V = IR where R is a function of temperature only

This is what Ohm was first to discover and what we use all the time, every time we use a resistor in a potential divider, to sense current, to calculate I^2R losses etc...

You won't change the world. If the only exposure many folk have had to Ohm's law is in circuit theory then I can see why they think that it is a fundamental relationship between V, I and R and is not related to physics (even though they assert it is a physical law). I think that they then get themselves into a little trouble when they state things like an IN914 diode with V=0.5V has a resistance of 5kΩ and then at 0.7V has a resistance of 140Ω (computed from the datasheet), when if they said that anywhere else no-one would have any idea what they were talking about.

 

You cant possibly be an EE and not know what an ohmic device is
vs a non ohmic device.
So name me one simple ohmic device that exists and name me one simple
non ohmic device. Dont cop out this time, name two devices in real
life that exist and can be discussed.
I can see you just like arguing so if you decide not to name two devices
this time i wont continue talking to you.

 

Yes, apparently some people just dont really care what Ohm's Law
really is. They want to make it cover everything.
Yes, the diode is funny. Dont ask BrownOut for an ohmic device though,
he might bring you a 1N4148 diode! ha ha

 

I honestly don't have any idea what is going on in this thread, but if you are trying to tell us that resistance can never change, I have a surprise for you.

Frankly, this is an absolutely pathetic discussion and you seem to have some compelling need to split hairs. At best this is a pedantic argument of who can be the most asinine about their definitions.

Let me ask you this, if R is always a constant, why is ohms law sometimes written as R = V/I? Have you ever seen dR/dt in an equation? I sure have.


In some cases, one would be forgiven for making the generalization that current is directly proportional to voltage; however, in reality that will never be true.

Current will never be directly proportional to voltage. Ever. In a real circuit resistance will never be constant.

 

Hello there,


First of all, welcome to the discussion. It's nice to hear from other people too
about this issue.

To answer your questions, many people dont think this is splitting hairs.
In fact, the discussion about Ohm's Law is a very good example of good
math gone bad.

To make this a little clearer, the tendency for some people to take the
math before the physics is a good example of the misapplication of
mathematics as applied to a physical phenomenon. I will explain.

Part of engineering knowledge is gained by learning various maths, including
calculus and differential equations, but even the simple algebra like
"C=A*x and solve for x".
But that's only *part* of the total education. The other part is learning
how to apply that math to a given physical phenomenon.

The more natural way to find out the mysteries of the universe is to first
observe the physical phenomenon, then figure out the math, not the
other way around. We dont usually start with an equation and then go looking
around the universe searching for a physical phenomenon to fit our equation.
In other words, chronologically the physical comes first, then the math.

What good would it do for me to write:
z=x^9.3+y*3.24+8.1234
and then go looking around for some physical phenomenon that happens to fit
this formula? On the other hand, it makes sense that if i found a previously
unknown phenomenon and then figured out the math to that phenomenon,
i would have gained a tool to understand it and convey that information to
other people.

Here is a good geometrical example...

Take the equation:
r^2=x^2+y^2

Now looking at that equation (note we are looking at an equation first, not
a physical phenomenon first) can anyone tell me what it describes?

Most people will exclaim in earnest, "It's a circle of course!". But those people
would not be correct. They would have misapplied the equation:
r^2=x^2+y^2
to a geometrical object known as a circle.

How could this happen?
It's because we looked at the math first, and then jumped to the conclusion
that it covered another familiar equation for an object known as a circle.
If we had looked at a circle first, and then tried to come up with an equation
to describe that object, we would have instead came up with this:
R^2=x^2+y^2
which is not the same as the original equation
r^2=x^2+y^2

Why arent they the same?
Because in one equation R is a constant, and in the other r (lower case) is
a variable.

What is the difference, what is the big deal?
The difference is that one equation describes a very unique object (a circle)
while the other describes many other objects that can have some pretty weird
shapes such as a very strange looking star fish, just to name one of the many.

What is the basic mistake we would make if we applied
r^2=x^2+y^2
to a circle?
The basic mistake is that we would have misapplied the math to a geometric
phenomenon.

This happens quite a bit i might add, but in the case of Ohm's Law the
misapplication of the equation
v=i*r
to Ohm's Law:
v=i*R
makes the very same mistake, except that it is a physical phenomenon this time
rather than a geometric object.

We can however apply the equation:
v=i*r
to Ohm's Law, but in doing so we have to remember to make the little 'r'
a constant:
v=i*R
or else we do not accurately portray the rule of Ohm's Law. In other words,
"It no longer looks like Ohm's Law".

It's as simple as that.

We might also apply the equation
v=i*r
to objects that do not obey Ohm's Law, in which case r would stay a
variable, but in many cases will take on a new definition which we would
learn through experiment, and then r might become r(...). This class of
object would then be known as 'non ohmic', one which does not obey
Ohm's Law.

We can also think of it this way...
if we were to allow
v=i*r
to be the exact same as
v=i*R
then how would we distinguish 'ohmic' devices from 'non ohmic' devices?

This is why i asked for two devices, one ohmic and one non ohmic, because
in naming these two objects it becomes very apparent that there are in fact
those devices that follow Ohm's Law and those that dont.

Sadly, i am starting to think that this is not the best forum to be
discussing this kind of issue about nature. I am starting to think that
this is a more relaxed forum where people simply want a place to talk about
electronics and not have to be too rigorous about the principles that
govern the science behind that electronics...which is still "all good" BTW.

 

This is frustrating. I think if one has learned something useful, the best way to convince others--at least in technical circles--is if one deeply and thoroughly understand the topic at hand, can explain same, and can point to established definitions and knowledge to support what one knows.

By contrast, so far all I have seen is terminology and equations being bandied about without much in the way of true, widely accepted definitions provided (e.g., as I did when I posted definitions for coefficients ), and with nebulous references to physics and Ohm's experiments, without any truly cogent explanation of the real physics that lies behind Ohm's law.

Wikipedia has done a hundred times better job of this in a dozen paragraphs than this thread's managed to do so far in all these posts.

I suggest that we start pointing to established definitions, such as "ohmic" and "non-ohmic" if that hasn't already been done. Likewise, definitions of parameter/variable, and constant or any other relevant mathematical terms.

I also think we'd get a lot farther discussing the physics and the origins of Ohm's law (what little I know of it comes from Wikipedia; but I of course want to verify against other sources).

Such as the physics behind:

J = σE -- "where J is the current density at a given location in a resistive material, E is the electric field at that location, and σ is a material dependent parameter called the conductivity."

Or getting closer to the physics, we could get into the basis of:

p=-eEτ -- "Since both the momentum and the current density are proportional to the drift velocity, the current density becomes proportional to the applied electric field; this leads to Ohm's law."

Michael

__________________
Michael Shimniok
http://bot-thoughts.blogspot.com/
Microcontrollers can solve world hunger, too!

 

"Ohmic Devices

If a device follows Ohm's law at all voltages that are placed across it, the device is called an Ohmic device and the resistance is said to have static resistance. In this case, the plot of V(I) is just a straight line with the slope =R.

"Non-Ohmic Devices

If a device behaves in a way that is NOT described by Ohm's law, (i.e. The resistance is not constant, but changes in a way that depends on the voltage across it.) the device is said to be non-Ohmic. In that case V(I) is not a straight line, but has some curvy shape. In this case, the resistance is called dynamic resistance because it's constantly changing."

Source: Response to circuit elements

__________________
Michael Shimniok
http://bot-thoughts.blogspot.com/
Microcontrollers can solve world hunger, too!

 

Hi again,


I, myself, tried long ago to point to established definitions and concepts but
there are people who seem to reject every written word on Ohm's Law when
it contradicts their own 'personal' concept of what it is.

The thing is, the underlying physics never came into question because that
was never the problem that anyone really had. The only problem that came
up over and over again was the one that i explained above...the misapplication
of an equation that is more far reaching than the law itself. Also, i assumed
that we all already knew that Ohm's Law was a statistical phenomenon that
involved a huge number of electrons, and in many applications the electrons
will behave the same way and will lead to the same conclusion as Ohm's Law,
but, that's not going to make Ohm's Law understood when it is the application
of math that is the problem in the first place.

Anyway, thanks for your links and comments, it's just too bad that i tried
that approach several times without any luck with some people :-)
Every time you mention "constant" resistance they want to through the equation
v=i*r at you, not realizing the huge difference between r and R.
As i said before, with R we have a physical law, but with r we ONLY have an
equation that is not yet applied to anything.

I am also glad you mentioned the concepts of ohmic and non ohmic, because
in finding out exactly what these two terms mean one eventually comes to the
right conclusion about Ohm's Law.

Are you perhaps saying that we have to resort to the total underlying physics
to explain the difference between r and R?
See the problem we are facing here in this thread is not the underlying physics,
but the correct application of math to a physical problem that has already been
completely outlined.
Case in point: the circle 'paradox' where r and R are again different. Note that
there is no real paradox here, but because the math can get misapplied it appears
that there is one.

Larry and John go fishing every week for 10 years. After a few weeks however they
notice that John ALWAYS catches twice as many fish as Larry. The continue to
joke about this over the years, until one day Larry goes deaf and blind, but still
manages to go fishing on the same day as John as usual. Larry now can not hear
John or see his fish, but he can count his own fish by feeling around and knowing
how many he dropped in the bucket, and today that happened to be five. The
question is, how does Larry figure out that John caught 10 fish that day?
Simple, he multiplies the number he has caught by 2.
Could he have done that if they always caught a varied number of fish each day
over the years? NO WAY.
The reason he could calculate the number of fish (for any day) is because the
ratio never changed, ie, it was a constant!
Without that constant nothing could have been calculated.
We can call this "John and Larry's Rule" he he.

 

I am saying that looking at the underlying physics will make it more clear why Ohm's law requires a static R. It is my belief that if the underlying physics is more deeply understood, then R, used as a physical property / coefficient / constant / static value -- for a given circuit, mind you -- becomes more apparent. I think if we go beyond the idyllic, restrictive setting afforded by Ohm's law, we can see just what that law is all about.

Ohm's law derives from the physical behavior of current density in a conductor as a result of electromagnetic field -- a conductor with a static resistance; one that does not vary with voltage or current. One for which all voltages applied across it yield a directly proportional current, so says the definition I posted.

From Wikipedia: "In a true ohmic device, the same value of resistance will be calculated from R = V/I regardless of the value of the applied voltage V. That is, the ratio of V/I is constant, and when current is plotted as a function of voltage the curve is linear (a straight line)."

According to the Wikipedia article, Ohm used a thermocouple as a stable voltage source, and varying lengths of wire. He measured current with a galvanometer and voltage was determined based on the thermocouple's temperature. His equation, x = a / (b + l') shows a proportionality relationship between, essentially, current (the galvanometer reading, x) and voltage (the thermocouple temperature, a), given a coefficient (comprised of the wire length, l' and "a constant of the entire setup," b). So from this knowledge it isn't terribly surprising to see R specified as a static value for each resistor to which the formula is applied.

I say "static" at least for the moment, as I think it is more intuitive. When I heard the word constant I mistakenly thought of a mathematical constant like π or e, and of a physical constant like the speed of light, Planck's constant or the Boltzmann constant. Until I looked up the definition of a constant.

Some confusion may arise when one objects that V=IR applies to all kinds of different resistors. Sure, it does. R can take on any value at all -- but the point is that for a given resistor to which Ohm's law is applied, that particular resistor has a static (constant) value. So I prefer to think of R as a coefficient.

To me this is analogous to any equation describing the deceleration of a car as a function of friction and drag coefficient. Such an equation can and should be generalized to enable us to apply the equation to many, many cars. But that doesn't mean the equation applies to a car that is constantly changing it's frontal area, or changing the size or stickiness of its tires. The coefficients of friction and drag are static for a particular car. But the equation is generalized to acknowledge that each car has a different coefficient of friction and of drag.

The same is so of Ohm's law. Resistance for a particular circuit to which Ohm's law holds, is static--it is a constant value. According to Wikipedia, the simplest example of generalizations by Ohm is J = σE, describing the relationship between J, the current density, and E, the electric field, given a "material dependent parameter called the conductivity," σ.

Digging further into conductivity, we discover that it is dependent on temperature -- so it does vary, and not linearly, though over a set range it can be approximated as a linear relationship. That suggests to me that J = σE can be evaluated at specific temperature points where σ is static (ie, it acts as a coefficient). That J = σE is a generalization by Ohm, this suggests to me that V=IR is actually a more specific case.

The point is that even in the general case, it seems to me that the law holds, so long as resistance (or conductivity/resistivity) is considered static, or constant. In other words, we pick a particular resistor at a particular temperature, and Ohm's law holds; voltage and current are directly proportional. Even if you have a thermistor or a CdS cell, Ohm's law will hold at a particular temperature or amount of light, respectively.

The subtlety here, that I have at last come to grasp, is that variables V and I can vary over time (AC, DC, etc) within the resistor itself, while Ohm's law holds -- provided that the coefficient, Resistance, is static over time -- a constant -- for the resistor in question -- though the equation is generalized to accommodate many different resistors with different values for conductivity/resistance/resistivity.

Primarily, Ohm's law does not hold when resistance is a function of voltage or current. That is to say, unless you can predict the same R for all values of V and I applied, Ohm's law doesn't hold. This gets tricky when resistance depends on temperature which depends in turn on the combination of time and current... but it seems that if one simplifies and fixes temperature (or time), one can use Ohm's law to predict V given any I or vice versa.

Anyway, I feel like I understand the underlying concepts behind Ohm's law much better now. It's been most interesting. And with that, adieu.

Michael

__________________
Michael Shimniok
http://bot-thoughts.blogspot.com/
Microcontrollers can solve world hunger, too!

 

Hello again,


First, thanks for the interesting and detailed post. It's nice to hear about some
of the underlying principles involved. I think that is important too.

The reason i started this thread is because i knew that there were still
people out there that couldnt come to grips with why R should be constant,
or why it should be held constant. Since even the simplest definition of
Ohm's Law clearly states 'proportionality', i felt that the main problem that
had been encountered by some people was that there was a mistake being
made in the application of the mathematics to the definition. The definition
i am talking about is the one that says that voltage is proportional to current.
If someone can not understand what proportionality really is, then unfortunately
they will never be able to understand why R is constant is Ohm's Law. Thus,
i decided to attach the problem from the root cause, which i can now label as:

The difference between a "proportion" and a simple "ratio".


When we use a ratio, we compare two *numbers* such as:
a/b

A proportion, however, is a statement that compares two *ratios*:
a/b=c/d

and of course since each ratio has two numbers, a proportion must
have at least four numbers.

Now see the definition of Ohm's Law should be easy to understand from
here because v must be *proportional* to i, not JUST the *ratio* of v to i.

In other words, writing:
R=V/I
or simply:
V/I
isnt enough. What is implied by the definition is
V1/I1=V2/I2
(a proportional relationship that requires 4 values).


When i started this thread i wasnt intending to rewrite Ohm's Law either.
I was assuming that anyone reading would believe the almost countless
definitions (on the web and in books) that say that voltage must be
*proportional* to current. Thus, i set out to clarify what proportionality
really means. If someone has a question about the origins of Ohm's Law,
then maybe that should be the subject of another thread.

I might add that in electrical engineering there are certain principles
adopted that are used as a measure for comparing other things. Ohm's Law
is taken to be perfect, as a straight line, to be used to measure other things
that come up in engineering. The temperature is forced to be constant,
and the electron activity is taken to be perfectly statistical.
This leads to the definition of an 'ohmic' device. The contrary, something
that does not follow this strict definition, is then defined as 'non ohmic'.
Note however that the ohmic device doesnt have to be super perfectly
ohmic, it only has to be approximately ohmic, but the definition itself is
perfectly ohmic just like any other definition is perfect.
An interesting name for this might be "Ohm's Ruler", because like a straight
edge, we hold that up against an new unknown device we encounter and
decide if the new device measures up or fails.

One of the advantages of adopting this kind of definition is that we can
quickly classify certain devices that appear in electronics and gain a quick
understanding of how we might have to deal with them. For a good example
take the LED vs the resistor. The resistor is ohmic and so it responds to
current and voltage in a very predictable way, while the LED is non ohmic
so it requires a lot more detail in order to be used effectively in a new circuit.

Ok, just to recap a little...

The difference between Ohm's Law R=V/I and a simple ratio R=V/I is that
Ohm's Law requires that V and I be proportional, and this requires two
sets of values:
V1/I1=V2/I2
while a simple ratio V/I only requires two values alone:
V1/I1


Thus, understanding the difference between a proportionality and a ratio
will lead to a better understanding of Ohm's Law for many people.

Also, here is a link which helps to clarify the difference between a ratio and a
proportionality:
http://www.math.com/school/subject1/.../S1U2L1GL.html

 

I am an EE, with a degree and everything. You make too many ASS-umptions about what I do and don't know. This discussion was not about my knowledge of what an ohmic device is or isn't; it was about whether or not R is a constant. You made as ASSumption about the extent of my knowledge, and BTW you got it WRONG! Just as pretty much every thing else you got wrong.
Also, I'm not your effing student, so you don't get to test my knowledge. Besides, I've already answered more of your irrelavnat and trivial questions than I ever care to answer. I don't feel like wasting any more of my time on someone who will just pretend I've refused to answer questions that I've clearly given my answer to. It's clear to me by now that you have nothing else of value to add to the discussion, and you're now just using obfuscation and distortion with all you're spurrious charges about questions, question, questions... You're not going to be satisfied no matter how many of your questions get answered, and will always have more qustions, so you can pretend that someone isn't responding correctly. Well, nobody cares what you think a correct response it. I know I know what I know, so you can take your dime-store quzzes and stuff then somewhere, I care not where.

__________________
You don't need a quadraphonic Blaupunkt -- you need a curve
ball.

 

I agree with alot of what you've said, but I think if we only consider "ohmic" materials in ohm's law, that would be too restrictive. We can find, for example, cases where we can define a resistance at a particular point in the V-I-Temperature domain. Once defined, that resistance can be used to further analyze the circuit. Only those who focus on trival matters would make a big deal if we defined this a an application of ohm's law.

__________________
You don't need a quadraphonic Blaupunkt -- you need a curve
ball.

 

BrownOut has better things to do than to bring people materials. You evidently have nothing better to do than write insulting things about people you don't know.

__________________
You don't need a quadraphonic Blaupunkt -- you need a curve
ball.

 

Hello again,


My main goal was to show how Ohm's Law and proportionality played hand in hand, and i am sure
i have shown that now.

Ok, so good luck to you then in your future projects...