Is a transistor basically a variable resistor?

[jafo]

I'm probably wrong about this, but it seems almost as if in a common [emitter|source] circuit, the transistor is acting like a variable resistor in the bottom part of a voltage divider. Can that be so? I can more or less see that for FETs, but BJTs are supposed to be exclusively and only about current... maybe "V=I×R" comes in to play somehow? Or am I just making yet another ID10t error?

[fowl]

Can you amplify with a variable resistor?

[Digital Larry]

Any device through which current flows has a dynamic resistance which is just V/I (I really had to stop to think about that LOL).  The question is how does the current change when the voltage changes.  Supposing you have a voltage follower, you know, base is connected to the wiper of a pot between +12 and ground and the emitter goes to a 100 ohm resistor and the collector goes to the power supply.  Does the collector current double when the collector voltage doubles?  What happens when the base voltage goes below 0.5 volts?

The simple answer to your question is "no", but even though I studied electronics pretty thoroughly, my best recollection of how a bipolar transistor works is Dr. Ralph Smith holding a little contraption made of of small boards with ping pong balls rolling around on it.  I'm not sure I got it then - I just learned all the rules of thumb and was able to apply them to the extent required to pass the tests, and later in life I really did make some LEDs blink using some. 

I think you're correct that a FET with a voltage that modulates the number of charge carriers in a channel is much easier to visualize as a variable resistor.

[Rob Strand]

The way you would disprove the variable resistor theory is to change the voltage on the collector then measure the change in current (without the collector resistor present as that would also be measured).   What you would find is the current is very much independent  of the change in the collector voltage.  That would imply a current output, since the current is constant, and not a resistor, where a voltage increase would be imply a current increase.

In reality you will find an increase in the collector voltage does produce a small increase in collector current.  These are secondary effects and the resistance might be 100k or so.  You can see the BJT's VI curve is slightly sloped instead being flat (constant current),



The simple mental model for a transistor operating at DC is a current amplifier (current in current out) but with the addition of the 0.65V DC voltage across the base and emitter.   For AC signals the transistor is a transconductance amplifier (voltage in current out).   That's where the concept of signal gain can be seen more easily.   The addition for the AC model is the transistor input impedance is finite, normally say 10k between the base and emitter - it's not constant but depends on the DC collector current.

Some of this stuff might seem a little abstract.   You might read up on transistor amplifiers on the web.   The simple answers are usually given as  formulas.   However to understand what's actually going on you need quite a bit of maths which might not be your thing.

Oh, at low collector voltages you can treat the BJT as a variable resistor.   That's an even more obscure story.
You can see how that might be true from the above graph where increasing collector voltage has increasing current (like a resistor).

[Keppy]

Yes, for certain definitions of "basically." That's actually how any simple amplifier works. You create a power supply, then use a small signal to control the flow of that power (in other words, varying resistance within a voltage divider on that power supply), creating a larger signal. Any single-device amplifier (BJT, JFET, MOSFET) works in that same basic way, despite all the differences in the details. Your mental model of a voltage divider is correct. I respectfully think some of these responses are missing the forest for the trees. Yes, there are myriad details of biasing and response, but none of that changes the fact that for the circuit in question, a transistor is indeed creating a variable divider.

[Steben]

An automated, non linear, finicky and to be biased potentiometer....

[antonis]

maybe "V=I×R" comes in to play somehow?

Which results into P = V_CE X I = I² X "R" = V_CE² / "R", where "R" stands for Collector - Emitter "imaginary" resistance..

More close to reality comes FET linear (ohmic/triode) region with R_DS (channel resistance) entering into the game..

[Steben]

That generic graph sure looks like it is showing a very smooth transfer curve. Makes me wonder.... To what extend are BJT's to be divided into smooth and hard clippers?
A question that completely shows a transistor is not a resistor ...

[marcelomd]

Well... there MUST BE a variable resistor inside transistors. How else would transistor men work?

[ElectricDruid]

Simple answer:

No, because a resistor has a linear relationship between voltage and current (that's what V=IR says), and a transistor has a more complex curved relationship (that's what the "transfer curve" is).

It looks a bit like a voltage-controlled variable resistor if you only think about a single value, but as soon as you start varying it, that goes out of the window.

[Rob Strand]

That generic graph sure looks like it is showing a very smooth transfer curve. Makes me wonder.... To what extend are BJT's to be divided into smooth and hard clippers?
A question that completely shows a transistor is not a resistor ...

The way it's shown is the collector VI characteristic (actually collect to emitter as it is implied the emitter is grounded).     If you used the collector as a clipper it would show less clipping for large positive signal amplitude.  Instead of clipping it "lets go" at high collector voltages.   That's what happens when you use a transistor as a controlled resistor and signal is large (for example in a tremolo).   For negative swings it depends on the base circuit if the base circuit is low impedance the BC junction will clip like a diode.

When you wire the base to the collector the VI characteristic completely changes.   It acts as a diode with the same exponential characteristic.    The clipping characteristic tends to be harder than a diode because the non-ideality factor for a transistor is around 1.0 whereas a silicon diode tends to be in the 1.5 to 2 zone.

[R.G.]

It has always helped me to look at bipolars as a reverse biased diode (the base-collector junction) that has its leakage increased hugely by being "poisoned" by current/charge carriers inserted through the base. It leaks more the more current into the base.

FETs are inherently resistors which are squeezed closed/open by the gate-to-channel/source electrical field. This is very, very similar to the Garden Hose Effect.

[jafo]

Hunh, turns out that "variable resistor" is a thing on its own; I just meant a resistor whose value can be varied, a noun + adjective. Sometimes terminology bites you in the ass. Sorry.

[GibsonGM]

Don't be sorry.  A transistor does have some properties of a variable resistance, and we surely have used it for that purpose...the guys who made it understood that, and debated about how to describe it.  It is far more complex than just a "RESISTANCE", because the base current is causing amplification - a divider can't do that.  So, there is another 'variable' in play, basically.  Another dimension, if you will. 

But your instinct to see it as a variable resistance is correct, from a simplistic starting point, as you get more familiar with how these things work.  I prefer to view this as a 'variable relay'...base current controlling how much current can flow thru my 'relay from collector to emitter'...we all have our reliance on easy to understand basic principles, and that's OK      Now, I need to get another beer, he he.

HTH... 

[PRR]

Where does the word "transistor" come from? Look it up! (There are at least two stories.)

Granted, the men who named it were still relatively clueless.

[Rob Strand]

IMHO, it's technically wrong to think of a BJT or a JFET as a variable resistor when there is significant voltage at the collector/drain.  The output is a current  - as per experiment in first paragraph of Reply #3.   I'm not going to stop people thinking there's little green men inside 
    .
+ + + + +
^ ^ ^ ^ ^

[Steben]

I'm not going to stop people thinking there's little green men inside 

Thanks

[amptramp]

Where does the word "transistor" come from? Look it up! (There are at least two stories.)

Granted, the men who named it were still relatively clueless.

My understanding is that it was a shortened form of "transfer resistor".

[duck_arse]

Where does the word "transistor" come from? Look it up! (There are at least two stories.)

Granted, the men who named it were still relatively clueless.

My understanding is that it was a shortened form of "transfer resistor".

that story rings a Bell.

[antonis]

that story rings a Bell.

Could you pretend you didn't hear it..??