Ohmic and Non-Ohmic Resistances

[MikeMl]

It is the thought of threads like this one which bring me out in a cold sweat whenever I type the words "Ohms Law" in reply to some simple question here on ETO.

JimB

Just the thought of being sniped by one who is hopelessly pedantic???

 

[Ratchit]

Hi again,

Yes that doesnt make any sense to me, unless i still dont understand the point you are trying to make. That's like saying in order to call something a liquid or non liquid it "has to be at least a little wet". It still sounds like you are saying that "everything has resistance" which although holds in real life in most cases, in theory we dont always have to deal with that resistance. So theory really tells us that even if everything does have resistance, it can still be separated from the other part(s) and that part or parts can be discussed independently.
To put this another way, if something doesnt have any resistance then it has to be non ohmic. But if something is non ohmic, that doesnt mean that it has to have no resistance at all.

What I am saying is something like in order for a material to have a viscosity, it has to be a liquid. And to be a liquid it has to slosh. I don't agree with your statement that if something does not any resistance, it is nonohmic. "Nonohmic" means nonlinearity of a resistance. If no significant resistance is present in a material for a desired application, it is nonresistive, not nonohmic. That is what I mean when I say that resistance has to be present before nonohmic has any meaning.

Ratch

 

[Ratchit]

Just the thought of being sniped by one who is hopelessly pedantic???

JimB,
Don't you think it is important to define concisely and precisely what terms mean? Otherwise they can be vague and fuzzy.

MikeML,
No matter how pedantic, correctness counts. If I am not correct, then it should be a snap to show with facts where and how I am wrong. No amount of verbal ju-jitsu can abrogate the physical facts.

Ratch

 

[JimB]

Don't you think it is important to define concisely and precisely what terms mean? Otherwise they can be vague and fuzzy.

I do think that it is important to define terms correctly.
However, for most practical purposes, "Ohms Law" whatever that is, is nicely summarised by the linear equation V = I x R. In which case the device being considered is "ohmic".
If there is some offset or inherent non-linearity then the device is "non-ohmic".

We could argue all day whether a tungsten filament lamp was ohmic or not.
Over the range of its working voltage, from 0 to (say) 240volts, there is some non-linearity.
But if we tested that same lamp over a range of (say) 0 to 5 volts, it may appear quite linear within the limitations of our measurements.

JimB

 

[Ratchit]

I do think that it is important to define terms correctly.
However, for most practical purposes, "Ohms Law" whatever that is, is nicely summarised by the linear equation V = I x R. In which case the device being considered is "ohmic".
If there is some offset or inherent non-linearity then the device is "non-ohmic".

The equation V=IR is not a statement of Ohm's law. It is a misnomer to say it is. V=IR is always correct no matter if the material has a straight IV curve or not. A material obeys Ohm's law and has the property of being ohmic if its IV curve is straight, and is nonohmic if its IV curve is not straight. Again, resistive linearity is a property of a material and not defined by V=IR.

We could argue all day whether a tungsten filament lamp was ohmic or not.
Over the range of its working voltage, from 0 to (say) 240volts, there is some non-linearity.
But if we tested that same lamp over a range of (say) 0 to 5 volts, it may appear quite linear within the limitations of our measurements.

Yes, materials appear to be resistively linear or ohmic if the range of interest is small enough.

Ratch

 

[Tony Stewart]

I did not understand how that barrage of facts related to what ohmic and nonohmic mean.
Ratch

in 40 years I have never spoke nor heard any other senior designer use the terms "nonohmic" . It's use seems to mean "Ohm's Law cannot be used here". I interpret it as the author does not know how to apply Ohm's Law to non-linear devices.

I disagree with the use of this term except for newbies, because as I stated before the nonlinear junction response of BJT's used as switches when one only considers the inpututput response.

In fact all saturated diodes and all switches (CMOS, MOSFET, IGBT or BJT's) have a fairly constant equivalent on-resistance for a given bias condition in the datasheet.

For BJT's, most devices are rated at Vce(sat) vs Ic for a fixed rated of Ic/Ib=10, 20 or the best components at 50.
This is approximately 10% of the maximum hFE but for consistent design results, these standard ratios are used in datasheets. Diodes Inc.(nee Zetex) led the industry with low Rce(sat) values are were first to show this in datasheets. Others make low Rce parts but you have to compute this value implied in their VI curve or table of values for Vce,Ic.

This allows a linear approximation of the device when saturated and thus Ohm's Law applies. Thus it is straight forward to predict the Vce for a given switch design using Ohm's Law .
This proves my point that the term non-ohmic means the speaker doesn't know how/when to apply Ohm's Law to non-linear devices.​

e.g.
FMMT619
50V NPN SILICON LOW SATURATION TRANSISTOR IN SOT23
Rce(Sat) = 68mΩ for a low equivalent on-resistance

Capiche?

 

[Ratchit]

Where is your reply?

Ratch

 

[Ratchit]

in 40 years I have never spoke nor heard any other senior designer use the terms "nonohmic" . It's use seems to mean "Ohm's Law cannot be used here". I interpret it as the author does not know how to apply Ohm's Law to non-linear devices.

I disagree with the use of this term except for newbies, because as I stated before the nonlinear junction response of BJT's used as switches when one only considers the inpututput response.

In fact all saturated diodes and all switches (CMOS, MOSFET, IGBT or BJT's) have a fairly constant equivalent on-resistance for a given bias condition in the datasheet.

For BJT's, most devices are rated at Vce(sat) vs Ic for a fixed rated of Ic/Ib=10, 20 or the best components at 50.
This is approximately 10% of the maximum hFE but for consistent design results, these standard ratios are used in datasheets. Diodes Inc.(nee Zetex) led the industry with low Rce(sat) values are were first to show this in datasheets. Others make low Rce parts but you have to compute this value implied in their VI curve or table of values for Vce,Ic.

This allows a linear approximation of the device when saturated and thus Ohm's Law applies. Thus it is straight forward to predict the Vce for a given switch design using Ohm's Law .
This proves my point that the term non-ohmic means the speaker doesn't know how/when to apply Ohm's Law to non-linear devices.​

e.g.
FMMT619
50V NPN SILICON LOW SATURATION TRANSISTOR IN SOT23
Rce(Sat) = 68mΩ for a low equivalent on-resistance

Capiche?

I stated many time before that Ohm's is a property of a material. And ohmic means that a material has that property, which is resistive linearity. If you divorce from your mind that Ohm's law means V=IR, then it should be clear how to talk and think about Ohm's law.

Why are you discoursing about devices? Ohms is about materials not devices.

Ratch

 

[crutschow]

“When I use a word,’ Humpty Dumpty said in rather a scornful tone, ‘it means just what I choose it to mean — neither more nor less.’

― Lewis Carroll, **broken link removed**

 

[Tony Stewart]

I stated many time before that Ohm's is a property of a material. And ohmic means that a material has that property, which is resistive linearity. If you divorce from your mind that Ohm's law means V=IR, then it should be clear how to talk and think about Ohm's law.

Why are you discoursing about devices? Ohms is about materials not devices.

Ratch

Your response indicates you only wish to be argumentative and contradictory and ignorant of the statements I've made.

Ohm's Law defines a linear DC property of conductors BUT ONLY for the restricted range of parameters, which are often ignored for practical or simplistic analysis, such as ;
- fixed temperature, ( pulsed duration that causes no temperature rise)
- limited voltage range,(that does not cause breakdown)
- limited current range, (that does not cause non-linear flux density)
- benign external EM fields, such as RF, gamma radiation, gravitational waves
- benign mechanical and climatic environment. ( such as shock, vibration, acceleration, spin, altitude, humidity, temp,)
- lack static voltage accumulation such as high ohmic resistance of insulators.

In all these above cases, deviations from simple Ohm's Law may be expected for which their may be correction factors for temperature or current.
It is also used for low frequency AC resistance when parasitic reactance contributes negligible error or when factored separately by phase shift or vector notation. Ohm's Law is in fact a reduction of higher Laws of nature such as Maxwell's Laws and Gauss's Law.

Some materials are more linear than others, so we call them linear for practical purposes.
All materials can be made so-called "non-ohmic" or nonlinear, without exception, if you ignore the above constraints and do not know how to deal with them.
This is why the constraints in all materials and components in datasheets.

But with experience, one can make practical use of Ohm's Law by understanding all the constraints, such as the saturation point of semi-conductors and the linearization of semiconductor's as resistive switches with a linear tolerance.

This makes the term non-ohmic to be only used by newbies who do not have not learnt yet that nothing is ideal and experience comes with knowing the tolerances and factors which change it for all materials, yet how and when to apply a fixed resistance.

Essentially" non-ohmic" has been defined as non-fixed resistance, which applies to all stressed materials, but treated as fixed for practical constraints.

~fini

any contradiction to above without proof is ignorance

 

[MrAl]

Hi,

The reason why it is hard to agree on even a simple equation like:
E=I*R

is because that is an analytical equation, and it contains no link to any physical geometry. It is impossible to calculate anything with a purely analytical expression. We need geometry to go with it, and by geometry i mean the physical geometric constructs as found in the application itself.

For a wild example, who can tell me what numerical value Y is in:
Y=3.2*|a|+b*7+c^2*8

If we dont know what this applies too, we cant calculate anything.

Another real life example is the calculation of the average of an AC sinusoidal voltage. If you say, calculate the average voltage of a signal that is 10 volts peak, you can get two different answers depending on the application:
1. 0v
2. 6.366v

which is correct?

In an application involving noise or small AC voltages #1 would be correct, but for power applications #2 would be correct.

So the equation:
Vavg=(1/T)*Integral(v(t)) dt

is simply NOT enough all by itself. Im sure there are many more examples of this kind of thing on the web.
This means that the equation:
E=I*R
can be written as:
E=I*R
or
E=I*r

and would be different because of the upper and lower case, had we written them in the more exact form required for each respective application.

For the ohmic and non ohmic discussion, i base my conclusions on the following equations:
E=I*R (or E=I*r)
P=I^2*R or P=V^2/R (or P=I^2*r and P=V^2/r)
(note R is in all of those)
V=L*di/dt
I=C*dv/dt
(note R is in neither of those)

Some of the elements above might be called ohmic, while others would be called non ohmic or simply "not ohmic" (that's what non ohmic means).
Since the last two equations above do not contain any resistance, i would say they are "non ohmic" or "not ohmic", or better "non resistive", as they dont dissipate any power so we know there cant be any resistance.
And my experience has been different than Tony's as we've used the term "non ohmic" lots of times on the job. What's more is if you read about almost any discussion on diodes, you'll see both terms "ohmic" and "non ohmic" come up in regard to the ohmic contacts and the non ohmic rectifying part. In this context ohmic refers to the actual linear equation E=I*R while non ohmic refers to the non linear equation E=I*r(x) where x is another variable for example current i. These two terms are widely used so i cant see how anyone would never use them, but i wont judge this on a job by job basis as terminology can be restrictive in different working environments.
In other contexts however ohmic might just mean the part that dissipates power, and may be non linear. This terminology is used because another part of the analysis shows a non ohmic part that does not dissipate any power, so we refer to the two parts as "the ohmic part" and the "non ohmic part", which refers more to the power dissipation attributes of each part than the actual VI curves.

 

[steveB]

I do think that it is important to define terms correctly.
However, for most practical purposes, "Ohms Law" whatever that is, is nicely summarised by the linear equation V = I x R. In which case the device being considered is "ohmic".
If there is some offset or inherent non-linearity then the device is "non-ohmic".

We could argue all day whether a tungsten filament lamp was ohmic or not.
Over the range of its working voltage, from 0 to (say) 240volts, there is some non-linearity.
But if we tested that same lamp over a range of (say) 0 to 5 volts, it may appear quite linear within the limitations of our measurements.

JimB

Well said. -concisely and correctly.

Here you say V=IR is a linear equation, which implies R is constant. The equation V=I R(V,I,T), where resistance can be a function of voltage, current and temperature, is sometimes (actually, usually) called Ohm's law, but this is for a non-ohmic material.

The idea that a material is ohmic is an idealization because no material is perfectly linear in V and I. The best examples of nearly ideal resistors (closest to being ohmic) are the modern resistors with very small temperature coefficients. But, they have limits over a range of power, as all resistors do.

I agree that Ohm's law is supposed to mean linearity in V and I, but doesn't it seem strange (even ironic) that the formula that is more universal, and more deserving to be called a law is V=I R(V,I,T), or some variation on that, which works in more general situations?

 

[cowboybob]

May we go back up the rabbit hole now... ?

 

[Ratchit]

“When I use a word,’ Humpty Dumpty said in rather a scornful tone, ‘it means just what I choose it to mean — neither more nor less.’

― Lewis Carroll, **broken link removed**

How does a quote from a fairy tale relate to what we are talking about?

Ratch

 

[Ratchit]

Your response indicates you only wish to be argumentative and contradictory and ignorant of the statements I've made.

Translation: I don't agree with you.

Ohm's Law defines a linear DC property of conductors BUT ONLY for the restricted range of parameters, which are often ignored for practical or simplistic analysis, such as ;
- fixed temperature, ( pulsed duration that causes no temperature rise)
- limited voltage range,(that does not cause breakdown)
- limited current range, (that does not cause non-linear flux density)
- benign external EM fields, such as RF, gamma radiation, gravitational waves
- benign mechanical and climatic environment. ( such as shock, vibration, acceleration, spin, altitude, humidity, temp,)
- lack static voltage accumulation such as high ohmic resistance of insulators.

In all these above cases, deviations from simple Ohm's Law may be expected for which their may be correction factors for temperature or current.
It is also used for low frequency AC resistance when parasitic reactance contributes negligible error or when factored separately by phase shift or vector notation. Ohm's Law is in fact a reduction of higher Laws of nature such as Maxwell's Laws and Gauss's Law.

Some materials are more linear than others, so we call them linear for practical purposes.
All materials can be made so-called "non-ohmic" or nonlinear, without exception, if you ignore the above constraints and do not know how to deal with them.
This is why the constraints in all materials and components in datasheets.

But with experience, one can make practical use of Ohm's Law by understanding all the constraints, such as the saturation point of semi-conductors and the linearization of semiconductor's as resistive switches with a linear tolerance.

From the above discourse, I think you agree with me that Ohm's law pertains to a material property of resistive linearity, and not to V=IR. That is all I wanted to say. Of course material properties change when the materials are subjected to changing environments and extreme conditions. One can write a book on how to ameliorate those effects, but that is not what I am talking about.

This makes the term non-ohmic to be only used by newbies who do not have not learnt yet that nothing is ideal and experience comes with knowing the tolerances and factors which change it for all materials, yet how and when to apply a fixed resistance.

And, also used by physics professors who write physics books such as I quoted in post #18 of this thread.

Essentially" non-ohmic" has been defined as non-fixed resistance, which applies to all stressed materials, but treated as fixed for practical constraints.

~fini

any contradiction to above without proof is ignorance

That is a fairly good alternate definition of nonohmic.

Ratch

 

[Ratchit]

Hi,

The reason why it is hard to agree on even a simple equation like:
E=I*R

is because that is an analytical equation, and it contains no link to any physical geometry. It is impossible to calculate anything with a purely analytical expression. We need geometry to go with it, and by geometry i mean the physical geometric constructs as found in the application itself.

For a wild example, who can tell me what numerical value Y is in:
Y=3.2*|a|+b*7+c^2*8

If we dont know what this applies too, we cant calculate anything.

Another real life example is the calculation of the average of an AC sinusoidal voltage. If you say, calculate the average voltage of a signal that is 10 volts peak, you can get two different answers depending on the application:
1. 0v
2. 6.366v

which is correct?

In an application involving noise or small AC voltages #1 would be correct, but for power applications #2 would be correct.

So the equation:
Vavg=(1/T)*Integral(v(t)) dt

is simply NOT enough all by itself. Im sure there are many more examples of this kind of thing on the web.
This means that the equation:
E=I*R
can be written as:
E=I*R
or
E=I*r

and would be different because of the upper and lower case, had we written them in the more exact form required for each respective application.

For the ohmic and non ohmic discussion, i base my conclusions on the following equations:
E=I*R (or E=I*r)
P=I^2*R or P=V^2/R (or P=I^2*r and P=V^2/r)
(note R is in all of those)
V=L*di/dt
I=C*dv/dt
(note R is in neither of those)

Some of the elements above might be called ohmic, while others would be called non ohmic or simply "not ohmic" (that's what non ohmic means).
Since the last two equations above do not contain any resistance, i would say they are "non ohmic" or "not ohmic", or better "non resistive", as they dont dissipate any power so we know there cant be any resistance.
And my experience has been different than Tony's as we've used the term "non ohmic" lots of times on the job. What's more is if you read about almost any discussion on diodes, you'll see both terms "ohmic" and "non ohmic" come up in regard to the ohmic contacts and the non ohmic rectifying part. In this context ohmic refers to the actual linear equation E=I*R while non ohmic refers to the non linear equation E=I*r(x) where x is another variable for example current i. These two terms are widely used so i cant see how anyone would never use them, but i wont judge this on a job by job basis as terminology can be restrictive in different working environments.
In other contexts however ohmic might just mean the part that dissipates power, and may be non linear. This terminology is used because another part of the analysis shows a non ohmic part that does not dissipate any power, so we refer to the two parts as "the ohmic part" and the "non ohmic part", which refers more to the power dissipation attributes of each part than the actual VI curves.

My position on using ohmic and nonohmic, when you really mean resistance is present or not, is that the words are being used as slang.

Ratch

 

[Ratchit]

Well said. -concisely and correctly.

Here you say V=IR is a linear equation, which implies R is constant. The equation V=I R(V,I,T), where resistance can be a function of voltage, current and temperature, is sometimes (actually, usually) called Ohm's law, but this is for a non-ohmic material.

V=IR is valid for any material under any condition whether its property is ohmic or nonohmic. It is the definition of resistance, and has nothing to do with Ohm's law.

The idea that a material is ohmic is an idealization because no material is perfectly linear in V and I. The best examples of nearly ideal resistors (closest to being ohmic) are the modern resistors with very small temperature coefficients. But, they have limits over a range of power, as all resistors do.

Yes, that has been pointed out countless times. Within a small enough range of interest, just about any material is ohmic.

I agree that Ohm's law is supposed to mean linearity in V and I, but doesn't it seem strange (even ironic) that the formula that is more universal, and more deserving to be called a law is V=I R(V,I,T), or some variation on that, which works in more general situations?

No, Ohm's law defines a property of a material and does not have a formula. V=IR is empirical formula like D=VT is for distance, velocity, and time. Why do folks not puzzle of what to call D=VT?

Ratch

 

[Ratchit]

May we go back up the rabbit hole now... ?

Only if you are a rabbit.

Ratch

 

[JimB]

I feel that I must lead Humpty Dumpty and Alice back out of the rabbit hole and into the real world.

Every couple of years here on ETO, this discussion of the work of Georg Simon Ohm ends up like this.

Grasping at all the metaphores, it is as though an express train has collided with a hand cart at a level crossing, the occupants of the hand cart were reported as being on a journey to the underworld.

With that thought, this thread is locked.
If you disagree with my decision, raise a report to the moderators.

JimB