-
Current controlled device vs voltage controlled? What is the difference? I saw in MrCarlson's lab that he said a transistor was a current controlled device and a FET? was a voltage controlled device? I thought potential was potential. I know the difference in current and volts. But wouldn't this mean that a PNP was current controlled (Say it was used where you had the device then transistor then ground, so the pnp acts as a path to ground) and NPN was voltage controlled (you had the NPN then your load then ground so it feeding the load). You send voltage to the base of the NPN to turn on while current draw on the PNP turned it on.
-
Bipolar, NPN or PNP, basic bipolar transistor operation is codfied,
low frequency, IC = beta x Ib, a simplified approximation, that is
base current controls current flow in the collector. Beta a characteristic
of the process, geometry, of the device. Varies a lot device to device.
And there are many other tertiary effects, but in simple form the above
suffices.
In the case of a JFET (saturation region) -
Here Idrain controled by Vgs (device has an intrinsic pinch off voltage)
Regards, Dana. -
In general, we distinguish devices according to the easiest description of the input.
For a resistor, current-driven and voltage-driven are equally easy.
For a vacuum tube or FET, the input impedance is very high so the input current is not well-determined. They are best described as voltage controlled.
For a diode or BJT, the input voltage changes little as the input current changes a large amount, therefore they are best described as current controlled.
There is an entire field of non-linear component theory that defines non-linear resistors, etc. according to one variable (e.g. voltage) being a non-linear function of the control variable (e.g. current). -
Both the phrase "current controlled" and "voltage controlled" are simplifications. But they are useful simplifications.
In the case of a current controlled device, the voltage is almost independent of the current, and the amount of current is what matters for control.
In the case of a voltage controlled device, the current is almost independent of the voltage, and the amount of voltage is what matters for control.
The phrase "almost independent" isn't the same as "completely independent". But it's close enough to make it very difficult to reliably use a constant regulated voltage to control a current-controlled device, and likewise difficult to use a constant regulated current to reliably control a voltage-controlled device. -
Both the phrase "current controlled" and "voltage controlled" are simplifications. But they are useful simplifications.
In the case of a current controlled device, the voltage is almost independent of the current, and the amount of current is what matters for control.
In the case of a voltage controlled device, the current is almost independent of the voltage, and the amount of voltage is what matters for control.
The phrase "almost independent" isn't the same as "completely independent". But it's close enough to make it very difficult to reliably use a constant regulated voltage to control a current-controlled device, and likewise difficult to use a constant regulated current to reliably control a voltage-controlled device.
I like this response as it answers the fundamental question.
I did wince a little bit with the phrase "almost independent" - but for the sake of getting the important points across, I would have to endorse it's use.
In regard to PNP or NPN bipolar transistors - both are current controlled. It's just that the difference in polarity requires a little more thought in understanding things - especially if you've been working a lot with one type. I can bias an NPN without blinking, but I have to stop and think if it's a PNP. -
Both the phrase "current controlled" and "voltage controlled" are simplifications. But they are useful simplifications.
In the case of a current controlled device, the voltage is almost independent of the current, and the amount of current is what matters for control.
In the case of a voltage controlled device, the current is almost independent of the voltage, and the amount of voltage is what matters for control.
The phrase "almost independent" isn't the same as "completely independent". But it's close enough to make it very difficult to reliably use a constant regulated voltage to control a current-controlled device, and likewise difficult to use a constant regulated current to reliably control a voltage-controlled device.
I like this response as it answers the fundamental question.
I did wince a little bit with the phrase "almost independent" - but for the sake of getting the important points across, I would have to endorse it's use.
In regard to PNP or NPN bipolar transistors - both are current controlled. It's just that the difference in polarity requires a little more thought in understanding things - especially if you've been working a lot with one type. I can bias an NPN without blinking, but I have to stop and think if it's a PNP.
So what does that mean bias? I see where people have to "rebias" old radios when they put in silicon diodes from germanium. I always thought the NPN is more useful because it goes before the load; simplification:battery-> transistor ->load ->then ground, where as the PNP is the last step before ground so its less useful. -
Current controlled device vs voltage controlled? What is the difference? I saw in MrCarlson's lab that he said a transistor was a current controlled device and a FET? was a voltage controlled device?
For my opinion, this is a clear question - and there can be only one correct answer.
Physical laws do not accept different explanations.
Therefore, I cannot understand vague statements like:
"IC = beta x Ib, a simplified approximation, that is base current controls current flow in the collector. .... but in simple form the above suffices."
"For a diode or BJT, the input voltage changes little as the input current changes a large amount, therefore they are best described as current controlled."
"...therefore we can control collector current by changing base current"
"Both the phrase "current controlled" and "voltage controlled" are simplifications. But they are useful simplifications."
No - we do not need any "simplifications" or "best descriptions".
The answer to raspberrypi`s question is: FET`s as well as BJT`s both are voltage-controlled!
I am aware that in some textbooks and many (too many!) forum contributions the BJT is described as current-controlled.
But this is simply WRONG!
What about the relation Ic=B*IB?
This is just a correlation (and not a chain of causation) - which should better be read as IB=IC/B.
In words: IB is a (small) part of IC. Yes, there is a base current IB which - in most design cases - can/must be regarded as a finite value.
So what? Does this imply any control function?
There is only one fundamental control function - W. Shockley`s equation which describes the voltage-current relationships at a pn junction:
Ic=Is*exp[(VBE/VT)-1] .
All reliable knowledge sources (Universities of Berkeley and Stanford, MIT, ....) describe the BJT as voltage-controlled.
This applies also to high-quality books like "Art of Electronics" (Horowitz/Hill).
To me, this is the only logical description because a small current NEVER can directly control a larger current (even the well-known "water-flow" model cannot function this way from the energy point of view).
The working principle of many circuits (current mirror, differential amplifier...) and many observable effects (EARLY-effect, tempco of -2mV/K,...) can be explained only based on voltage-control.
Therefore, it is really a phenomenon to me that current-control is still under discussion. This view cannot be supported by any explanation - neither physical nor practical.
Two simple examples:
1) The gain expression of a common emitter stage contains the transconductance gm=IC/VT. This parameter gm is nothing else than the SLOPE of the mentioned exponential function (Shockleys equation): gm=d(IC)/d(VBE). How can somebody deny that VBE controls IC?
2.) Question: Can somebody (who still is believing that IB would control IC) explain the function of a stabilizing resistor RE in the emitter path?
We do not need complicated physical considerations (behaviour of charged carrier etc.) - only simple observations of electrical circuits (and explanations of the observed effects) are sufficient to understand that the BJT is, of course, a voltage-controlled device.
-
A bipolar transistor can be controlled either as Vbe or Ib control.Quote
The collector–emitter current can be viewed as being controlled by the base–emitter current (current control), or by the base–emitter voltage (voltage control). These views are related by the current–voltage relation of the base–emitter junction, which is just the usual exponential current–voltage curve of a p-n junction (diode).[1]
https://en.wikipedia.org/wiki/Bipolar_junction_transistor#Voltage.2C_current.2C_and_charge_control
"Normally" one designs to base control current. This is because of the non linear T dependent Vbe
behavior, not the least device to device variations issues.
The Ebers-Moll approximation for transistors -
http://inderjitsingh87.weebly.com/uploads/2/1/1/4/21144104/the__ebers-moll_bjt_model.pdf
This model shows how approximations can be used for simplified calculations, as well as more
exact.
Regards, Dana. -
Approximations for calculations is one thing, conceptual understanding is another...
I was just as much confused by "BJTs are current controlled" as the op of this thread.
The same with op amps, I've been royally confused over the things until I read Darren Ashby's book "Electrical Engineering 101" - where he breaks down the op amp into what it is: a difference operator and a huge gain in series. Now probably this only gave me my "eureka" moment because I have endured four semesters of being hit over my head with control loops, signals and linear systems, but it did for me what endless reiterations of "tries to keep the inputs at the same voltage", and similar rules of thumb didn't.
Of course I do use the "tries to keep the inputs at the same voltage" for calculating feedback loops. But the rule of thumb model for calculations and the definition that made me able to understand what the component does have been quite different.
So addressing BJTs and FETs, the way I understand it, "current controlled" just stands in for "has a low (and nonlinear) input impedance", and "voltage controlled" for "has an almost infinite input impedance". And since it has a low and nonlinear input impedance, it follows that you'll have to primarily manipulate current to get it to do exactly what you want it to do, while with an almost infininte input impedance, you can just put the input voltage on and you're set. -
It is helpful to consider a BJT as voltage controlled when non-linear effects matter,
and current controlled when the conditions warrant. (And charge controlled when
that's useful.)
Folks who do this for a living find all three models helpful at times. Orthodoxy and
iconoclastic adherence may provide some comfort, but the inflexibility makes for
inefficient design.Current controlled device vs voltage controlled? What is the difference? I saw in MrCarlson's lab that he said a transistor was a current controlled device and a FET? was a voltage controlled device?
For my opinion, this is a clear question - and there can be only one correct answer.
--- SNIP ---
The answer to raspberrypi`s question is: FET`s as well as BJT`s both are voltage-controlled!
I am aware that in some textbooks and many (too many!) forum contributions the BJT is described as current-controlled.
But this is simply WRONG!
What about the relation Ic=B*IB?
This is just a correlation (and not a chain of causation) - which should better be read as IB=IC/B.
----- SNIP ----
There is only one fundamental control function - W. Shockley`s equation which describes the voltage-current relationships at a pn junction:
Ic=Is*exp[(VBE/VT)-1] .
----------------- SNIP ------------------
Two simple examples:
1) The gain expression of a common emitter stage contains the transconductance gm=IC/VT. This parameter gm is nothing else than the SLOPE of the mentioned exponential function (Shockleys equation): gm=d(IC)/d(VBE). How can somebody deny that VBE controls IC?
2.) Question: Can somebody (who still is believing that IB would control IC) explain the function of a stabilizing resistor RE in the emitter path?
OK... In answer to the last question:
Consider a voltage source driving a resistor of Rb like this
GND -- Vs ---- Rb ---- B --- E --- Re --- GND
Now using the h-parameter model (that fixes the voltage between B and E
at a "nominal" diode drop), the voltage around the loop would be
Vs - (Rb*Ib + 0.7 + Re*Ic) = 0
If we take as our model a current controlled source Ic = hfe * Ib, then
Vs - 0.7 = Ib * (Rb + hfe * Re)
But for suitably large beta, any incremental change in Vs (call it "vs) causes an
additional current in Re: ie = vs/Re which is more-or-less equal to the
incremental collector current. If the collector load is Rc, then the
incremental change in collector voltage will be Rc*vs/Re -- giving us
a gain of Rc/Re.
And I did it all with a current controlled model. One could do the same
with a voltage controlled model, but you'd end up with a really unusable
analytic model.
----------------------------------
So that's question 2.
Now to the other part that looked interesting:QuoteAll reliable knowledge sources (Universities of Berkeley and Stanford, MIT, ....) describe the BJT as voltage-controlled.
This applies also to high-quality books like "Art of Electronics" (Horowitz/Hill).
To me, this is the only logical description because a small current NEVER can directly control a larger current (even the well-known "water-flow" model cannot function this way from the energy point of view).
The working principle of many circuits (current mirror, differential amplifier...) and many observable effects (EARLY-effect, tempco of -2mV/K,...) can be explained only based on voltage-control.
Therefore, it is really a phenomenon to me that current-control is still under discussion. This view cannot be supported by any explanation - neither physical nor practical.
1. All the reliable engineers I've ever worked with have used both models.
All the textbooks I've ever read employ several models.
2. The "the water flow model"? I'd agree, nor does the "little elves" model, or the
"magic smoke" model. I don't know anyone who uses either to describe the action of
a BJT.
However, please note that, as powerful as the voltage control model is, it does not
explain the behavior of an optically excited BJT. Nor does it explain the triggering
of a parasitic BJT in CMOS circuits by injection of charges from nearby sources.
(I have seen both happen in real products.) These are both charge controlled
phenomena.
Finally, the assertion that "a small current NEVER can directly control a larger current"
is offered as a support for the argument that "a small current does not control a larger current
in a BJT" which is a tautology -- a rhetorical trap for listeners.
In fact, a small current can cause the concentration of minority carriers in the base region
to become large enough to make it possible for current to flow in between the collector
and base, despite the presence of a reverse bias condition. The number of carriers need
only be sufficient to reduce the barrier to further injection of carriers across the C-B barrier.
Once this is achieved, it is only necessary to maintain sufficient base current to support the
minority carrier concentration because of the few injected carriers that are swept into the
BE junction.
As long as we're appealing to the citadel of Berkely, see an EE105 lecture
https://inst.eecs.berkeley.edu/~ee105/sp04/handouts/lectures/Lecture22.pdf
It rationally presents several models, each of which have some use in real
engineering and circuit design.
I don't particularly care what orthodoxy folks embrace, but I will point out that
adherence to one so rigid is neither efficient nor helpful for folks who want to
develop an understanding of how to design and explain the behavior of circuits.
respectfully,
-
Causality in Electromagnetism strikes again!
I just wrote about my opinion about causality: https://www.eevblog.com/forum/chat/how-does-the-electron-make-a-photon-in-an-antenna/msg1135583/#msg1135583
I won't start arguing about it, but I am curious about how this thread will continue.
To be on-topic, even if this contradicts my own opinion that causality in electromagnetism is undefined, I'll say that
- bipolar transistors are current controlled
- field effect transistors are voltage controlled
Again, this voltage/current control classification I am voting for is just for practical and engineering reasons. Otherwise, a debate about the causality in electromagnetism will be interesting, but I'm not sure if causality makes any sense in the first place. -
"Normally" one designs to base control current. This is because of the non linear T dependent Vbe
behavior, not the least device to device variations issues.
@danadak:
Can you, please, explain this statement in more detail? I rather think, that it is not VBE but the currents IE and IC which depend on the temperature.
Furthermore, I can confirm that the Ebers-Moll BJT-model (and the more advanced Gummel-Poon BJT-model) is based on voltage control.
More than that, they are the kernel of the BJT models used in simulation programs. -
It is helpful to consider a BJT as voltage controlled when non-linear effects matter,
and current controlled when the conditions warrant. (And charge controlled when
that's useful.)
.............
Folks who do this for a living find all three models helpful at times.
Well - it was not my intention to discuss several models which might exist.
And I did not want to discuss which view might be helpful or not.
Each student/beginner or designer can decide which approach he is willing to follow.
My only intention was to answer the main question: Is the collector current of the BJT determined by the voltage VBE or by the current IB.
It was my understanding that the question was related to the BJT as a part - and not to a complete circuit where we sometimes use a large base resistor to realize something which we call "current injection".
And, in this context, only one single answer is possible - in spite of different existing small-signal models.But for suitably large beta, any incremental change in Vs (call it "vs) causes an
I am sorry, but I cannot consider this as an answer to my question which was related to the stabilizing effects of RE (starting with an unwanted Ic increase due to tolerances or temperature effects).
additional current in Re.
More than that - where is the logic behind your statement that any change in the supply voltage Vs will cause an "additional current in Re" ?
Is Re directly connected to Vs? No - of course not. Why do you left out the most important part - the base-emitter path between Vs and Re?1. All the reliable engineers I've ever worked with have used both models.
All the textbooks I've ever read employ several models.
Again - I do not speak about models. I speak about the BJT`s physical properties and the mechanism how IC can be varied.
Respectfully
LvW -
Vbe is essentially a forward biased diode, and can be derived from Schockley's
equation (diode driven by a constant current).
Vd = n x Vt x ln( idiode / Isatdiode ). and Vt = kT/q
Regards, Dana.
-
Vbe is essentially a forward biased diode, and can be derived from Schockley's
T
equation (diode driven by a constant current).
Vd = n x Vt x ln( idiode / Isatdiode ). and Vt = kT/q
Regards, Dana.
Basic law: No current without a driving voltage.
Physics: It is the VOLTAGE across a pn junction which influences the thickness of the depletion layer and, thus, allows a variable current.
We can manipulate each equation - but not change physical properties and dependencies.
Therefore: Shockley has introduced his formula only in this form (Vf=forward voltage): Forward current If=Io*[exp(Vf/Vt)-1]
-
To simplify the debate, let's ask a simpler question:
We have an ideal resistor of a given value. The resistor is connected in a DC electric circuit. This will give us a voltage across the resistor, and also a current through the resistor. Again, all components are ideal, and all measured values are stationary values, and not zero.
For our given resistor, is the current controlled by the voltage, or is the voltage controlled by the current? -
To simplify the debate, let's ask a simpler question:
We have an ideal resistor of a given value. The resistor is connected in a DC electric circuit. This will give us a voltage across the resistor, and also a current through the resistor. Again, all components are ideal, and all measured values are stationary values, and not zero.
For our given resistor, is the current controlled by the voltage, or is the voltage controlled by the current?
Only one correct answer is possible: The current is controlled by the voltage.
More than that - it is DETERMINED by the voltage.
I am aware that we all speak about something like a "voltage drop" caused by a current.
But this is a misconception - on the other hand, it helps to calculate and analyze electric circuits.
With other words: In electronics, there is nothing like a "current source" - we only can realize voltage sources having a large internal resistance.
Now - when we connect a load resistor to this "current source" we have, in fact, a driving voltage and a resistive VOLTAGE DIVIDER.
In each conductive body (like a resistor) we nead an E-field allowing moving of charges (which we call "current").
And such an E-field is the result of an applied voltage (in BOTH resistors).
A current never can "produce" a voltage. This would be a false understanding of Ohms law.
And this applies, of course, also to the B-E junction of a BJT. It is the voltage which determines the width of the depletion layer and, hence, the amount of emitter current IE which then is split into IB (small part) and IC (major portion). -
OK, but an electric current can happen as well without a voltage. Even more, we can also show how a current can "determine" the "appearance" of a voltage.
But how can we make the charges to flow without a voltage? It doesn't matter how, but if we manage to pump the charges in a circuit, then a voltage will be "produced".
It's not mandatory to have a voltage in order to move charges. We can pump charges in many ways, for example mechanically, as in a Van de Graaff generator. Charges are attached to the belt of the electrostatic machine. The belt is moving, so the charges are moving too. This will produce, by definition, a current. This current is as good as any other current, and if we push this current through our resistor, the voltmeter will indicate a voltage. In this situation, since all has started by the mechanical belt, I guess we can say the current is the cause of the voltage.
Another example of a current produced by something else than a voltage will be the "electrical wind" produced by a heated filament.
Another one will be the Seebeck effect.
Another one will be the photoelectric effect, and so on.
Almost any form of energy can be determined to "pump" some charges for us, and this flow of charges is exactly a current, which current will create a voltage. Now it seems like the current is the "cause" of everything, so there are no voltage controlled devices, just current controlled devices.
Can this be true? I guess not.
The logic fallacy here is, in my opinion, the assumption that one of them, either the current or the voltage, must determine or "produce" the other one.
Since both the voltage and the current are tied together inside the same equation, why not considering them as different faces of the same law?
Why there must be a causality between them?
And after all, how we define causality? -
The abbreviations are getting out of control. What do the abbreviations in those equations stand for? Biggest hurdle to learning is when you have to look up each abbreviation, then find you looked up the wrong one. MrCarlson is never wrong!!!
Maybe I should just stick to organic chemistry, even though its technically a harder subject, for me its alot easier to understand then electronics. -
So what does that mean bias? I see where people have to "rebias" old radios when they put in silicon diodes from germanium. I always thought the NPN is more useful because it goes before the load; simplification:battery-> transistor ->load ->then ground, where as the PNP is the last step before ground so its less useful.
Biasing sets the operating point of the transistor. I'm not going to paraphrase the much better explanations on Google but there are many including tutorials and videos. Some bias conditions, like for switching, are trivial. Class A amplifiers take a little more thought (as do the other amplifier configurations).
-
But how can we make the charges to flow without a voltage? It doesn't matter how, but if we manage to pump the charges in a circuit, then a voltage will be "produced".
OK - perhaps I should restrict my answer to the circuits and the parts we are speaking of in this thread: Simple elctronic circuits with BJT`s.
Let`s concentrate on our subject instead of discussing Van der Graaff generators and the Seebeck effect.
Our point was and still is: Which quantity controls/determines the collector current of a BJT, right?
It is not my intention to convince anybody - I just have tried to answer a question.
Of course, I do understand that nobody simply will believe me and my arguments because: Who am I already?
Therefore, I like to quote another - more reliable - person:
„BJT is a voltage-controlled current-source; the base current is purely incidental (it is best viewed as a „defect“).“.
This quote comes from Barrie Gilbert - one of the worlds most known designer and inventor of novel BJT applications .
If somebody in the forum is interested to read more about the circuits and effects I have mentioned to supporti my view, I can communicate some details.
-
Sigh. Not this soporific argument again!
Instead of debating about which is current or voltage controlled, how about helping the original poster how to understand transistors and the models used to describe them, rather than having a pissing contest?
And bear in mind this is the beginners section, so don't be too quick to post lots of maths.Current controlled device vs voltage controlled? What is the difference? I saw in MrCarlson's lab that he said a transistor was a current controlled device and a FET? was a voltage controlled device? I thought potential was potential. I know the difference in current and volts. But wouldn't this mean that a PNP was current controlled (Say it was used where you had the device then transistor then ground, so the pnp acts as a path to ground) and NPN was voltage controlled (you had the NPN then your load then ground so it feeding the load). You send voltage to the base of the NPN to turn on while current draw on the PNP turned it on.
BJTs whether NPN or PNP behave in the same way. The difference is the polarities of the voltages and currents are reversed.
The same is true for P vs N channel FETs.
If you build the same circuit with the NPN transistors swapped with PNP transistors and vice versa, it will still work, as long as the polarity of all the power supplies and other polarised components such as diodes and electrolytic capacitors, is reversed.
FETs have a high impedance (open circuit) gate, consisting of a reverse biased diode (J-FET) or an insulating layer of polysilicon (MOSFET).
A BJT has a low impedance base, consisting of a diode junction.
Both BJTs and FETs need to be treated differently, depending on whether they're being used in the active region (analogue) or as a switch (digital).
In the case of both BJTs and FETs, when biased in the active region, changes in base-emitter (BJT) voltage or gate-source voltage, causes changes in the collector or drain current. In this mode, both of the transistors can be treated as voltage to current converters (transconductance amplifiers).
When used as a switch, the transconductance is of less importance. To turn a BJT fully on, the base needs to be driven with a high enough current to support the required collector current. The base-emitter voltage is a parasite. It's generally assumed to be the highest value specified on the datasheet. In the case of a FET, the gate voltage needs to be high enough to achieve minimal drain resistance and once the transistor is on, no current is drawn from the gate. -
All I can state is that some post threads around here are just real timesinks. No real learning or teaching going on here, just ego exercises.
-
When used as a switch, the transconductance is of less importance. To turn a BJT fully on, the base needs to be driven with a high enough current to support the required collector current. The base-emitter voltage is a parasite. It's generally assumed to be the highest value specified on the datasheet. In the case of a FET, the gate voltage needs to be high enough to achieve minimal drain resistance and once the transistor is on, no current is drawn from the gate.
Good point that a lot of introductory articles and videos are about how to use transistor as a switch. This is one source of "BJT is current controlled, FET is voltage controlled."
BTW, for MOSFETs the insulating layer is typically SiO2. The gate is polysilicon.
Just to, there are a lot of ways of looking at it, not just one true way. Another way is "all transistors are charge controlled devices." Yet another is "BJTs are current controlled" because the collector current is controlled by the emitter current.
-
Current controlled device vs voltage controlled? What is the difference?
Any voltage source can be seen as a current source, and any current source can be seen as a voltage source, so there is no real difference. They are equivalent. As long as you don't break the math, you can use whatever you like most, or whatever will ease your calculations.
See https://en.wikipedia.org/wiki/Source_transformation. This is sometimes called the Thevenin-Norton equivalence.I saw in MrCarlson's lab that he said a transistor was a current controlled device and a FET? was a voltage controlled device?
I don't know if you saw that, or not. I guess you saw. Anyway, FET stands for Field Effect Transistor, and BJT (which I guess will be the other kind of transistor from your question) stands for http://bfy.tw/A4bU
Yes, people like to think about BJTs as current controlled devices (again, BJT means "normal" transistors, like NPN or PNP transistors) , and about FETs as voltage controlled devices. Because of Thevenin-Norton, it doesn't really matter how you want to see those transistors. Be careful here. Your teacher might want to see each device exactly as he, the teacher, teach you to see it. So, after all, no matter what Thevenin-Norton said, there is only one right answer: the one that your teacher is expecting from you.I thought potential was potential. I know the difference in current and volts.
I have no idea what are you talking about.But wouldn't this mean that a PNP was current controlled (Say it was used where you had the device then transistor then ground, so the pnp acts as a path to ground) and NPN was voltage controlled (you had the NPN then your load then ground so it feeding the load). You send voltage to the base of the NPN to turn on while current draw on the PNP turned it on.
No. For a device to be current controlled, it doesn't matter if the control current is going to the ground, or is coming from the load. It doesn't matter if you "push" a current into the Base of a NPN, or if you "pull" a current from the Base of a PNP. As long as you vary a current in order to control the transistor, then we are saying the transistor is current controlled.
In conclusion:
- What parameter are you tweaking when you want to control a PNP transistor?
- I am tweaking its Base current. Aha, a current! So this device is controlled by a current, so we call it a current controlled transistor.
- Now, what parameter are you tweaking when you want to control a NPN transistor?
- I am tweaking its Base current. Aha, a current! So this device is controlled by a current tweak, too, so a NPN transistor is also a current controlled transistor, like the PNP one.
- And finally, what are we tweaking when we want to control a FET transistor?
- Well, a Field Effect Transistor (AKA FET, or MOSFET) does not have a Base, in the first place. Instead of Base, the FET has a Gate. The Gate does not allow any current. Since the Gate has zero current, all we can tweak in order to control a FET is the voltage applied to its Gate. So, we say the FET is a voltage controlled transistor.
