Voltage or current operated devices?

[BrownOut]

Or the lowered voltage allows more charge to flow because there is less voltage to oppose the diffusion voltage.

Diffusion current neutralizes charge in the junction, lowering the voltage which in turn allows more charge to flow.

 

[crutschow]

crutschow,



Compliment is the correct word. But I took your remark in the best possible light.

Ratch

Thanks for correcting my spelling.

I should have realized you are immune to sarcasm. (The previous sentence is also sarcastic, in case you missed).

 

[kavan]

Hello crutschow,

Kavan here. Do you mind if I ask you what it means "functionally"?
I am still trying to get the distinction between voltage and current operated devices. Both types seem to work if a voltage is applied to them. I also understand the difference between linear and nonlinear devices, where the current and voltage are not proportional: a change in voltage causes a nonlinear change in current...
still I don't get the difference between voltage and current operated devices. What do you mean that some devices are "functionally" current operated?

thanks
kavan

 

[crutschow]

Kavan here. Do you mind if I ask you what it means "functionally"?
I am still trying to get the distinction between voltage and current operated devices. Both types seem to work if a voltage is applied to them. I also understand the difference between linear and nonlinear devices, where the current and voltage are not proportional: a change in voltage causes a nonlinear change in current...
still I don't get the difference between voltage and current operated devices. What do you mean that some devices are "functionally" current operated?

A good example of a functional current-operated device is a bipolar junction transistor (BJT). The solid-state physics theory is that the base-emitter voltage is what determines the collector-emitter current. But since the input impedance of a BJT looks like a forward biased diode and thus has a very low impedance, the BJT acts much like a current operated device with the collector-emitter current related to the base-emitter current by the current gain (beta or Hfe). Thus functionally the device acts like a current-operated device with a relatively low input impedance, at least for large-signal operation (switching or bias-point calculations). Thus you must always have a resistor in series with the base to limit the base current with a BJT (for a common-emitter circuit).

But just to confuse things slightly, for small-signal AC voltage gain calculations the BJT is often viewed as a voltage amp with a transconductance gain (Gm) and a moderately low input impedance of typically a few kΩ.

So how the BJT is viewed depends mainly upon what aspect of its operation you are interested in.

This is different than a MOSFET, for example, which has a very high input impedance and the drain-source current is controlled by the gate-source voltage. It is strictly a voltage operated device under all conditions.

 

[MrAl]

Hi,

From what i've seen in my time my only lasting conclusion can be that all electronic devices operate both on voltage and current, so to provide an accurate description of a device we have to know what it's electrical characteristics are, and those characteristics almost always include BOTH voltage and current specs, and the operation is ALWAYS dependent on both voltage AND current simultaneously.

As Carl (aka 'crutschow') notes in his post nicely is that the most attention is paid to one spec over the other (current over voltage or voltage over current) only after we had chose a given application. That's because almost every device that runs on electricity depends on both at the same time, but often one is more important than the other or we have to run it based on one more then the other and we dont care that much about the other within reason.

To illustrate, an LED is run based mostly by providing the correct current, because the basic operation says that the more current you provide the more light output you get up to the limit of the device. But the LED will also have an approximate voltage specification that has to be met as well. So one without the other just doesnt work, even though it's more like a current operated device than voltage.

A BJT is no different. It is often said to be a current operated device because a huge number of applications out there depend mostly on the behavior of the device with respect to the current levels within the device. But it's no different because there are also applications and even theories that depend highly on the voltage characteristics such as the base emitter voltage rather than current.

The MOSFET is not really an exception here. Normally we think of driving the gate at a particular voltage so that we can effectively get the device to turn 'on'. But then we run across an application where we dont only have to turn it on, but we have to turn it on FAST. This leads to the need to understand both the voltage characteristics of the gate combined with the current characteristics of the gate, which then tells us that we need a much higher current to turn it on fast than to just keep it turned on after it's already been turned on.

So in short, we tend to call something 'current operated' when the voltage doesnt change much with current or just isnt as important for correct operation in a given application, and 'voltage operated' when the current doesnt change much with voltage or just doesnt matter as much for the correct operation in a given application. So you can see why some devices will have a dual description: one application is more sensitive to the voltage behavior while another application is more sensitive to the current behavior.
Mathematically this might look like this:
Current Operated when the response with: di/dx is high, dv/dx is low
Voltage Operated when the response with: di/dx is low, dv/dx is high
So the MOSFET is voltage operated when dv/dt causes a large change in behavior, but current operated when di/dt causes a large change in behavior, or both at the same time when both cause a large change in behavior. For fast turn on, we have to provide a large current to get the voltage to rise fast enough, so it's sort of both at the same time during that period because if we lack one or the other it just doesnt work.

 

[Claude Abraham]

Ratch right here is the source of your confusion. Vbe is NOT the external voltage "applied". First of all we never apply Vbe directly to a constant voltage source, CVS! Vbe does not make charges move. We need an example. Sue, a singer, uses a mic, connected to an amp. The 1st bjt in the preamp stage will serve as our discussion example device. Sue imparts acoustic energy from her lips. This energy impinges on the mic diaphragm resulting in generated I & V. The signal propagates through the mic cable at finite speed. When it reaches the b-e junction, what happens? The charges are already flowing, Ib is already established. The Vbe at the junction is there because the bjt device is biased from a dc network to establish a good quiescent point for the bjt.

The dc Ib/Ie consists of charges transiting through the junction, forming a depletion zone, DZ, resulting in a barrier, Vbe. Does Vbe "control" Ib? We must examine the actions at the junction to determine what controls what, knowing that it's possible neither Ib nor Vbe nor Ie is the quantity ultimately in control. Sue's mic adds small signal charge carriers, the ac current will be denoted in lower case, "ib/ie", with voltage signal as "vbe".

At the b-e junction we have a fixed dc bias Ie/Vbe/Ib. When the signal carriers transit through the mic cable & arrive at junction, what happens? Since the charge carriers moving through said junction increase, as we now have Ib+ib, Ie+ie, the no. of accumulated charge carriers in the DZ also increase. With more carriers present & finite lifetime, there is an increase in excess stored charge in the DZ. Hence the barrier potential increases. Thus vbe is a direct consequence of ib/ie, that is the increase in ib/ie forced vbe to increase. Similarly, a reduction in ib/ie produces a reduction in vbe.

Of course, it is true that to a very limited degree, vbe has a 2nd order influence on ib/ie. If the external source providing the drive for the network is constant voltage, such as a generator, Hall sensor terminated in a high impedance, the current is determined by (Vg-vbe)/R. Let's say a generator outputs 1.00 volt, the junction drops 0.650 volt, with a resistance of 1.00 kohm. Current is 0.350V/1.00kohm = 0.350 mA. If the generator output increases to 1.10 volts, the initial current is (1.10V-0.650V)/1.00kohm = 0.45 mA. But when the charges arrive at the junction, the DZ changes, & the forward voltage drop increases to 0.655 V. So when equilibrium is reached, the current is now (1.10V-0,655V)/1.00kohm = 0.445 mA. So Vd(Vbe, whatever), does have a slight 2nd order influence on the current.

But in this example, the generator output voltage was of the same order of magnitude as the junction forward voltage drop. If the generator outputs 10 volts, the junction drop has less influence. If the generator was constant current source, then Vd/Vbe has no influence at all.

Ratch you have to examine this carefully. Vbe is not an active emf providing energy to charges. The E field in the b-e junction due to DZ charge layer accumulation is not an active emf source driving charges into motion. That is the function of the signal source. Vbe is not controlling Ie/Ib except in the 2nd order where a voltage source drives the input at the base of the bjt. The output voltage of said voltage signal source is Vg (dc), & vg (ac). If the emitter is connected straight to common w/o any emitter resistance, then Iib = (vg-vbe)/R, where R is the total resistance in the base network including Rb, rbb, Rg, etc.

When a junction diode or bjt b-e is driven by a voltage source through a resistance, then the current is related to Vg-Vbe. But an effective topology reduces errors due to Vbe as well as temp, etc. An effective biasing scheme uses either current sources/sinks so that Vbe has little influence, or a voltage source plus resistor scheme where the voltage supply rails are way larger than Vbe. If the supply rails are tens of volts, and Vbe incremental changes are tens of millivolts, then Vbe has only a slight influence on current.

So, Ratch, regarding Vbe having an influence on Ib/Ie (& ultimately Ic as well), via (Vg - Vbe)/R, I concede on that point w/o an argument. Take an LED powered from a +5V source & 330 ohm resistor, with a forward drop, Vd, of 1.80V. Id = (5.00V - 1.80V)/330 ohm = 9.70 mA. Say an ambient temp increase occurs, & Vd drops to 1.70V, then Id = (5.00V - 1.70V)/330 ohm = 10.0 mA. A 5.56% drop in Vd resulted in a 3.0% increase in Id. But if the source is +12V instead of +5V, with R=1.0 kohm, we get:

Id = (12.0V - 1.8V)/1.0kohm = 10.2 mA. Then temp increase forces Vd to 1.7V: Id = (12.0V - 1.7V)/1.0kohm = 10.3 mA. So the same 5.56% decrease in Vd resulted in only a 1.0% increase in Id. No doubt, when a p-n junction is driven by a voltage source plus series resistor, the forward drop Vd (or Vbe for a bjt) definitely has a SLIGHT influence over I. But well designed circuits minimize the influence of Vd over Id. A constant current drive is immune to changes in Vd as long as Vd does not reach values too large for the current source to comply with.

On that 1 point Ratch, we agree. But do the numbers & you will see that V exerts a very minimal influence on I. BR to all.

 

[MrAl]

Hi Claude,


I think you may be talking more fundamentally then we want to get into for this discussion, if i understand you right. In which case we seem to need a push before we can get anything to move. Nothing moves on it's own unless it is first pushed from some external force or had been pushed on in the past and is now in motion because of that past (perhaps very distant past ) push.

It's true that when we push something it usually moves at least a tiny bit, so we could say that the push and the move have to be simultaneous. But theoretically it is possible to push something and not move it at all, if it is considered entirely rigid. So it either moves or it doesnt, but if it doesnt, then we can definitely say that the push came before the move because we've observed things being pushed that do not move, and it's entirely feasible that someone else was pushing the other way and when they let go it moves. So the original push came first, then the movement. So do we think of the case where the push and the movement occur at the same time as a separate case? That's possible i guess, but it seems more reasonable to think that in order to move something it requires a push first even if there is no delay. Because there is no delay it's still arguable i guess, but to me it seems more reasonable to look at it as push comes first.

If we want to make charge move, we've got to provide a push. We've at least got to provide something else that had a different charge distribution if we are not to push directly. That means we had to 'push' something else first even if it wasnt the charge itself, but the charge holder.

Ultimately, the field has to be set up first before anything moves, and the only way to set up the field is to push the charge into place or allow the charge holder's field influence to reach the target.

 

[Claude Abraham]

Hi Claude,


I think you may be talking more fundamentally then we want to get into for this discussion, if i understand you right. In which case we seem to need a push before we can get anything to move. Nothing moves on it's own unless it is first pushed from some external force or had been pushed on in the past and is now in motion because of that past (perhaps very distant past ) push.

It's true that when we push something it usually moves at least a tiny bit, so we could say that the push and the move have to be simultaneous. But theoretically it is possible to push something and not move it at all, if it is considered entirely rigid. So it either moves or it doesnt, but if it doesnt, then we can definitely say that the push came before the move because we've observed things being pushed that do not move, and it's entirely feasible that someone else was pushing the other way and when they let go it moves. So the original push came first, then the movement. So do we think of the case where the push and the movement occur at the same time as a separate case? That's possible i guess, but it seems more reasonable to think that in order to move something it requires a push first even if there is no delay. Because there is no delay it's still arguable i guess, but to me it seems more reasonable to look at it as push comes first.

If we want to make charge move, we've got to provide a push. We've at least got to provide something else that had a different charge distribution if we are not to push directly. That means we had to 'push' something else first even if it wasnt the charge itself, but the charge holder.

Ultimately, the field has to be set up first before anything moves, and the only way to set up the field is to push the charge into place or allow the charge holder's field influence to reach the target.

Mr. Al, your contributions to this topic have been very helpful & accurate. I know I tend to get very deep into fundamentals, but I felt that the contrarian point of view always claim that p-n junction devices like diodes, LEDs, bjt, SCR, etc., "mimic" current controlled behavior, but the "underlying physics" says they are voltage controlled. But I've examined the underlying physics for 35 years & every reference published does not support the contrarian's claims. I will plead guilty to the charge that I take things down to an excessively deep level. But when critics say they have semicon phy on their side, I must call them on it. They don't have physics on their side, they are blowing smoke, whereas my intent is to clear the air.

Your points about fields & motion are well noted & I do agree with you on those points. I had to bring singing Sue into the picture to illustrate that all disturbances in fields, all initiation of charge motion, must start with an independent source of power. Sue's singing is not controlled by I or V or charge distributions, or fields. Sue is an independent generator. Her energy in acoustic form undergoes transducer action & gets translated into electrical form. That is where it all starts. The change in I is "caused" by Sue ultimately. Is that fair?

 

[Ratchit]

Claude Abramson,

Ratch right here is the source of your confusion. Vbe is NOT the external voltage "applied". First of all we never apply Vbe directly to a constant voltage source, CVS! Vbe does not make charges move. We need an example. Sue, a singer, uses a mic, connected to an amp. The 1st bjt in the preamp stage will serve as our discussion example device. Sue imparts acoustic energy from her lips. This energy impinges on the mic diaphragm resulting in generated I & V. The signal propagates through the mic cable at finite speed. When it reaches the b-e junction, what happens? The charges are already flowing, Ib is already established. The Vbe at the junction is there because the bjt device is biased from a dc network to establish a good quiescent point for the bjt.

No, I am not confused. I am looking at just the device, not the circuit. Vbe is the external voltage applied. I am not establishing a quiesent operating point, or worrying about singing swinging Sue. I am just saying that Vbe will cause Ic and Ib to exist in a exponential relationship.

The dc Ib/Ie consists of charges transiting through the junction, forming a depletion zone, DZ, resulting in a barrier, Vbe. Does Vbe "control" Ib? We must examine the actions at the junction to determine what controls what, knowing that it's possible neither Ib nor Vbe nor Ie is the quantity ultimately in control. Sue's mic adds small signal charge carriers, the ac current will be denoted in lower case, "ib/ie", with voltage signal as "vbe".

That ground has been plowed many times. Sedra and Smith shows that the Ib depends on Vbe, as does Ic. Again, it is the forward voltage vs the barrier voltage.

At the b-e junction we have a fixed dc bias Ie/Vbe/Ib. When the signal carriers transit through the mic cable & arrive at junction, what happens? Since the charge carriers moving through said junction increase, as we now have Ib+ib, Ie+ie, the no. of accumulated charge carriers in the DZ also increase. With more carriers present & finite lifetime, there is an increase in excess stored charge in the DZ. Hence the barrier potential increases. Thus vbe is a direct consequence of ib/ie, that is the increase in ib/ie forced vbe to increase. Similarly, a reduction in ib/ie produces a reduction in vbe.

Accumulated charge slows down the response. It does not change Vbe controlling Ic or Ib. Vbe does not increase or decrease if an external voltage source is applied.

Of course, it is true that to a very limited degree, vbe has a 2nd order influence on ib/ie. If the external source providing the drive for the network is constant voltage, such as a generator, Hall sensor terminated in a high impedance, the current is determined by (Vg-vbe)/R. Let's say a generator outputs 1.00 volt, the junction drops 0.650 volt, with a resistance of 1.00 kohm. Current is 0.350V/1.00kohm = 0.350 mA. If the generator output increases to 1.10 volts, the initial current is (1.10V-0.650V)/1.00kohm = 0.45 mA. But when the charges arrive at the junction, the DZ changes, & the forward voltage drop increases to 0.655 V. So when equilibrium is reached, the current is now (1.10V-0,655V)/1.00kohm = 0.445 mA. So Vd(Vbe, whatever), does have a slight 2nd order influence on the current.

According to Sedra and Smith, Vbe is directly controlling Ib and Ic. I am not talking about hooking other things to the BJT like Hall sensors or anything else.

But in this example, the generator output voltage was of the same order of magnitude as the junction forward voltage drop. If the generator outputs 10 volts, the junction drop has less influence. If the generator was constant current source, then Vd/Vbe has no influence at all.

A constant current generator has an equivalent high external resistance. That is is violation of a single device setup. In any case, the Vbe, regardless of its size in relation another source in series with it, controls Ib and Ic.

When a junction diode or bjt b-e is driven by a voltage source through a resistance, then the current is related to Vg-Vbe. But an effective topology reduces errors due to Vbe as well as temp, etc. An effective biasing scheme uses either current sources/sinks so that Vbe has little influence, or a voltage source plus resistor scheme where the voltage supply rails are way larger than Vbe. If the supply rails are tens of volts, and Vbe incremental changes are tens of millivolts, then Vbe has only a slight influence on current.

No, Vbe controls Ic and Ib in the active region of a BJT.

So, Ratch, regarding Vbe having an influence on Ib/Ie (& ultimately Ic as well), via (Vg - Vbe)/R, I concede on that point w/o an argument. Take an LED powered from a +5V source & 330 ohm resistor, with a forward drop, Vd, of 1.80V. Id = (5.00V - 1.80V)/330 ohm = 9.70 mA. Say an ambient temp increase occurs, & Vd drops to 1.70V, then Id = (5.00V - 1.70V)/330 ohm = 10.0 mA. A 5.56% drop in Vd resulted in a 3.0% increase in Id. But if the source is +12V instead of +5V, with R=1.0 kohm, we get:

Id = (12.0V - 1.8V)/1.0kohm = 10.2 mA. Then temp increase forces Vd to 1.7V: Id = (12.0V - 1.7V)/1.0kohm = 10.3 mA. So the same 5.56% decrease in Vd resulted in only a 1.0% increase in Id. No doubt, when a p-n junction is driven by a voltage source plus series resistor, the forward drop Vd (or Vbe for a bjt) definitely has a SLIGHT influence over I. But well designed circuits minimize the influence of Vd over Id. A constant current drive is immune to changes in Vd as long as Vd does not reach values too large for the current source to comply with.

You are talking about design again. Fine, but still, the voltage across the diode determines the current, and everything will adjust to Schockley's equaton.

On that 1 point Ratch, we agree. But do the numbers & you will see that V exerts a very minimal influence on I. BR to all.

Vd is what I am talking about.

Ratch

 

[kavan]

Hi Ratch,

what do you think about this explanation by J. Beaty? does it make sense?

http://amasci.com/amateur/transis.html

thanks
kavan

 

[MrAl]

Hi,



Shouldnt we be talking in only one of the two threads? I just gave a clear example of two applications that each would have to declare the BJT to be EITHER current controlled OR voltage controlled, not one or the other. This was in the other thread.

The BJT is considered current controlled when we are designing a current amplifier, but is considered voltage controlled when we are designing a voltage reference diode. So it is both.

 

[Ratchit]

kavan,

Hi Ratch,

what do you think about this explanation by J. Beaty? does it make sense?

http://amasci.com/amateur/transis.html

thanks
kavan

Yes, I have known about that site for a long time now. I might not agree with everything he says, but he certainly states a lot of things correctly. One more thing, you come across a lot of talk about models, and how this model is used for this and another model for that. Just remember, models tell you what the device will do, but it does not tell you how or why it works the way it does. For that, you need to dig into the physics of the device.

Ratch

 

[misterT]

Just remember, models tell you what the device will do, but it does not tell you how or why it works the way it does. For that, you need to dig into the physics of the device.

You keep repeating that in every debate you are involved with.. and mostly everybody agrees with that. Physics will explain in detail how the device works. But why can't you agree that the detailed physics are not important in practical use. You keep pushing the physics over the models. And you keep arguing against the models also. The models tell an engineer how to use the device, not the physics. If I'm using a microcontroller, I need to know the architecture; the memories and instruction set etc. I don't need to know how the silicon chip was made and how a single transistor or memory cell works in physical level.

 

[Ratchit]

misterT,

You keep repeating that in every debate you are involved with.. and mostly everybody agrees with that. Physics will explain in detail how the device works.

I keep repeating it because new OPs are asking questions.

But why can't you agree that the detailed physics are not important in practical use. You keep pushing the physics over the models.

Because folks keep trying to use and extrapolate design methods into explaining how a device works. I do agree with design methods for utilizing devices in applications, but when folks start saying things like "base current controls collector current", then I feel obligated to correct that statement by saying that base current is only an indicator of collector current.

And you keep arguing against the models also. The models tell an engineer how to use the device, not the physics. If I'm using a microcontroller, I need to know the architecture; the memories and instruction set etc. I don't need to know how the silicon chip was made and how a single transistor or memory cell works in physical level.

No, I am against the misuse of models, such using a design model to explain how a device works. Otherwise, models are very useful.

Ratch

 

[MrAl]

Hello,


If you try to increase the collector current by applying a 'voltage' to the base, you find that the diode equation says that I and V are so inextricably related that you can't just 'apply' a voltage because the diode, in order to do this, requires current as well. In point of fact, you can do it the other way around, you can apply a current and not yet see the voltage rise until the input capacitance has been also satisfied. Once the voltage does rise, then we see a change in collector current, but the voltage didnt get there by magic...it took significant current to get there, and the current has to be maintained in order to keep the collector current going.

So the basic junction requires both current and voltage.

Another way to think about it...how can we accomplish 'anything' without at least some energy (power)? Voltage or current alone do not constitute energy, it takes both. We can set up a magnetic field with current, but if that magnetic field accomplishes anything physical it's going to take voltage too.

 

[Ratchit]

MrAl,

If you try to increase the collector current by applying a 'voltage' to the base, you find that the diode equation says that I and V are so inextricably related that you can't just 'apply' a voltage because the diode, in order to do this, requires current as well. In point of fact, you can do it the other way around, you can apply a current and not yet see the voltage rise until the input capacitance has been also satisfied. Once the voltage does rise, then we see a change in collector current, but the voltage didnt get there by magic...it took significant current to get there, and the current has to be maintained in order to keep the collector current going.

I assume you are directing your remarks to me. Yes, there is a one to one correspondence of voltage to current. I said so in my remarks to Claude. But, the physics of the junction diode shows that it is the external applied voltage that lowers the barrier voltage, thereby enabling a continuous diffusion of charge carriers to take place. If you apply a external current source, the diode will react according to the voltage that the current source applies to the diode. Therefore, that makes a junction diode a voltage controlled device. You cannot discern this through models, equations, or graphs. You have to delve into the the internal workings or physics of the diode itself.

So the basic junction requires both current and voltage.

Certainly it does, but that does not negate what I avered.

Another way to think about it...how can we accomplish 'anything' without at least some energy (power)? Voltage or current alone do not constitute energy, it takes both. We can set up a magnetic field with current, but if that magnetic field accomplishes anything physical it's going to take voltage too.

Certainly a little energy is required to control a lot of energy, but that does not negate what I avered.

Ratch

 

[MrAl]

Hi there Ratchet,


Yes you can say that, but then when you say "physics" you're talking about the PN junction itself, not about the diode as a whole. So you're still applying some physics yet ignoring other physics.

A better illustration maybe...
You have a diode with a PN junction aboard a spacecraft 1 light year away from Earth. Assuming a radio signal travels at the speed of light, how long does it take to forward bias a PN junction from the ground base on Earth (by pressing a button which energizes a circuit which sends out the signal that eventually reaches the spacecraft). The answer is about 1 year. Now why didnt the voltage of the button affect the PN junction immediately? Because we had other physics to consider. And this 'other' physics we had to consider was not too trivial to the application, it was paramount. It took a whole year vs microseconds on Earth.

So it appears that you want to isolate the discussion to focus on the PN junction alone, the theory behind it. That's ok i guess, but it is a little removed from reality because we always have other things to consider when we talk about an *actual* physical device. As im sure you know, we have inductance and capacitance across and though all distances, no matter how short. And we also have resistance too. So we've always got time constants to think about.

Also, is there any proof that voltage alone can perform some function in a solid state device like a PN junction? Once the voltage gets there it's too late, because then we've already seen the power dissipated and that event is over. Once the event is over we cant say that the voltage did it because that would be leaving out what happened just before that, which is also part of the picture. The 'theory' will ignore this physical aspect because theories are there to do just that, ignore some things and concentrate on other things so that our understanding can come more immediately. If we had to concentrate on voltage and current at the same time it would be a more difficult learning process, and would also impede some design procedures.

I tried to give another example removed from the present discussion, based on magnetics. If we generate a magnetic field we need current but not voltage, but as soon as that magnetic field starts to do anything we can call physical (like pick something up or push something) then some energy has to be expended. We can say that the current caused the magnetic field and maintains the magnetic field, but if the magnetic field operated on something it would require at least some voltage otherwise we'd have an efficiency of over 100 percent.
For example, say we have a magnetic field generated by a current and that field squeezes a conducting strip made of semi conductive foam tighter and tighter. As the foam compresses the conductance goes up more and more. Thus, a current flows through the strip from end to end that increases with the magnetic field.
Did the current causing the magnetic field control the current through the strip? Yes, but without the voltage needed to overcome the resistance in the magnetic coil we could not have generated the field. So some energy was expended in the coil. How about if we use a superconductor? If we use a superconductor then we are just introducing another block in the control scheme, where whatever controls the current in the superconductor has to dissipate energy, or we have to move the superconductor closer to the strip (as an alternate to changing the current) but that takes energy too.

So it appears that you are talking about PN junction theory, removed from an actual device, but the actual device doesnt behave that way. And even with the PN junction alone some energy has to be applied to get anything to happen therefore it can not be solely controlled by voltage. The only thing controlled solely by voltage is the theory

BTW, if anything was truly controlled by voltage alone, then that same voltage would be able to control an infinite number of said devices.

 

[Ratchit]

MrAl,

Yes you can say that, but then when you say "physics" you're talking about the PN junction itself, not about the diode as a whole. So you're still applying some physics yet ignoring other physics.

Yes, I am concentrating on the junction itself, not the lead connections or the bulk resistance.

A better illustration maybe...
You have a diode with a PN junction aboard a spacecraft 1 light year away from Earth. Assuming a radio signal travels at the speed of light, how long does it take to forward bias a PN junction from the ground base on Earth (by pressing a button which energizes a circuit which sends out the signal that eventually reaches the spacecraft). The answer is about 1 year. Now why didnt the voltage of the button affect the PN junction immediately? Because we had other physics to consider. And this 'other' physics we had to consider was not too trivial to the application, it was paramount. It took a whole year vs microseconds on Earth.

That sounds like sophistry to me.

So it appears that you want to isolate the discussion to focus on the PN junction alone, the theory behind it. That's ok i guess, but it is a little removed from reality because we always have other things to consider when we talk about an *actual* physical device. As im sure you know, we have inductance and capacitance across and though all distances, no matter how short. And we also have resistance too. So we've always got time constants to think about.

These "other" things are irrelevant or insignificant with respect to what I am I am saying with respect to voltage control.

Also, is there any proof that voltage alone can perform some function in a solid state device like a PN junction?

Yes, obviously voltage causes current to exist in a junction diode.

Once the voltage gets there it's too late, because then we've already seen the power dissipated and that event is over.

Who cares if energy is used or dissipated? The voltage caused the event to happen.

Once the event is over we cant say that the voltage did it because that would be leaving out what happened just before that, which is also part of the picture. The 'theory' will ignore this physical aspect because theories are there to do just that, ignore some things and concentrate on other things so that our understanding can come more immediately. If we had to concentrate on voltage and current at the same time it would be a more difficult learning process, and would also impede some design procedures.

That sounds more like a philosophical argument than a scientific one.

I tried to give another example removed from the present discussion, based on magnetics. If we generate a magnetic field we need current but not voltage, but as soon as that magnetic field starts to do anything we can call physical (like pick something up or push something) then some energy has to be expended. We can say that the current caused the magnetic field and maintains the magnetic field, but if the magnetic field operated on something it would require at least some voltage otherwise we'd have an efficiency of over 100 percent. For example, say we have a magnetic field generated by a current and that field squeezes a conducting strip made of semi conductive foam tighter and tighter. As the foam compresses the conductance goes up more and more. Thus, a current flows through the strip from end to end that increases with the magnetic field....

I am not going to expand the discussion into the physics of semiconductor foam and superconductors, where I would have to study the phyics of each of those materials.

So it appears that you are talking about PN junction theory, removed from an actual device, but the actual device doesnt behave that way. And even with the PN junction alone some energy has to be applied to get anything to happen therefore it can not be solely controlled by voltage. The only thing controlled solely by voltage is the theory

I beg to differ. Any breadboard will show that a junction device does behave in accordance with Schlockey's equation and semiconductor theory. The fact that energy is used does not abrogate the fact of voltage control.

BTW, if anything was truly controlled by voltage alone, then that same voltage would be able to control an infinite number of said devices.

Yes, and the point is?

Ratch

 

[misterT]

So, Ratchit, how would you light up a general red LED from 5V supply? What is the design procedure you go through?

 

[MrAl]

MrAl,



Yes, I am concentrating on the junction itself, not the lead connections or the bulk resistance.



That sounds like sophistry to me.



These "other" things are irrelevant or insignificant with respect to what I am I am saying with respect to voltage control.



Yes, obviously voltage causes current to exist in a junction diode.



Who cares if energy is used or dissipated? The voltage caused the event to happen.



That sounds more like a philosophical argument than a scientific one.



I am not going to expand the discussion into the physics of semiconductor foam and superconductors, where I would have to study the phyics of each of those materials.



I beg to differ. Any breadboard will show that a junction device does behave in accordance with Schlockey's equation and semiconductor theory. The fact that energy is used does not abrogate the fact of voltage control.



Yes, and the point is?

Ratch

Hello again,


Well if energy is used, then that's the true physics, isnt it? Energy is the prime mover, not voltage. Voltage can not cause anything to happen on it's own, it needs current to accompany it even if it is a small current.

In effect you are taking the Shockley Equation and stating that voltage is the key controller. But i said over and over that voltage cant do anything without current, because it's the energy that does anything, and that's the only thing that can possibly do anything real. Why then does the energy have to exceed a certain minimum value in order to move an electron through the band gap? It's not the voltage, it's the energy. If we scrape off the sides of the diode and hold the voltage just below some level we could accomplish the same thing by shining a light onto the die (cause more conduction). That's because we added more energy, not voltage. Also, with no voltage applied we could take that same die and actually produce a voltage at the terminals just by shining a light on the die. So we've actually 'created' a voltage with the energy from the light source.

If you take the S Equation and look at the voltage, it looks like voltage is controlling something. But it's got to be the energy when we look at the total "underlying physics". You can say it is voltage controlled because voltage is in the equation, but if we solve for current then we can say it is current controlled. Why the dual role? Because it's the energy that does it all. And the energy takes the voltage and at least some current. It might be a small enough current to ignore for many roles, but if we want the full picture we dont want to ignore it totally.

The point to setting up an infinite number of devices controlled with the SAME voltage source is quite simple:
If we can control an infinite number of devices with the same voltage source then we must be ignoring the current.
Yes we can control an infinite number of *theoretical* devices, but we can not control that many *real* devices because we could never find the energy to do so. Even if we consider only those electrons within this one universe.