How transistors REALLY work

[cwalex]

Hi,

I found this article linked to on www.talkingelectronics.com website. http://amasci.com/amateur/transis.html It looks really interesting, does anyone have any comments on the accuracy of the information? It is a very different description than I have ever read from anywhere else about how bjt's work.

cheers,

Alex

[IanB]

From a quick glance, it seems reasonable to me.

Did you ever wonder why the gain (beta) of a BJT varies so much? It's because it is not actually a fundamental device parameter, but rather a derived quantity that depends on other more fundamental parameters.

All transistors are essentially voltage amplifiers. A BJT just happens to be a very poor approximation to an FET.

[w2aew]

I agree with Ian - at a quick glance, a very reasonable explanation.  BJTs are really transconductance devices (base-emitter voltage is what determines collector current).  This is evident in the basic Ebers-Moll model of the transistor which states that:

I_e = I_s * (e^(Vbe/Vt) - 1)

...emitter current is equal to saturation current times e raised to the base-emitter voltage divided by the thermal voltage. 

Current gain (Bete, hFE, etc.) is a by-product of the transistor's design, and is NOT well controlled, which is why you should always try to bias / operate transistors to be largely independent of the variation in current gain.

This video, which I happen to put together tonight for a friend, discusses common emitter amplifier gain and frequency response.  You'll note from the video that current gain isn't mentioned at all - but transconductance certainly is.

[ejeffrey]

It seems mostly right, athough the author fixates on some things a bit too much and has a bit of a crazy person tone.

Anyway, both the current amplifier and the voltage-controlled current (transconductance) model of transistors are "right", but the transconductance model is "more right".  It is both quantitatively more accurate and better reflects the underlying physics.  As a bonus, FETs and valves can also be described as transconductance devices, so you can treat them all on a more similar footing.

The right way to think about a transistors operation is to first understand diodes.  You forward bias a diode by ~0.6 V, and current flows across the junction.  There turn-on is quite sharp, current rises exponentially with voltage according to the ideal diode equation.  In a transistor, the base-emitter diode basically functions as a normal diode.  In fact, the emitter current is accurately described by the ideal diode equation.  The difference is that in a transistor the collector 'steals' almost all of the emitter current and only a tiny trickle makes it out the base lead.

All transistors are essentially voltage amplifiers. A BJT just happens to be a very poor approximation to an FET.

BJTs may be less ideal in terms of drive current and saturation voltage, but they have much higher transconductance.  BJTs transconductance is exponential in Vbe, while FETs are only quadratic in Vgs.

[free_electron]

He's partially right up till about part 2 because there is a part lacking ... The doping of the emitter is different than the collector.

then there's all the drivel about ' transistors do not amplify current.. blabla. of course they don't,  you can't create energy, only convert it... basic law of nature. a transistor is a steering valve.the emitter-base current steers the emitter-collector current.  And the drivel about 'of current is not what flows its charge flowing'. This is semantics.

Electrical current is the movement of electrical charges. Electrical charges are carried by electrons. So when we say current flows we mean : electrical charges move . You say tomeeto i say tomaato.
same with the drivel about the water analogy. yes the pipes already contain water. this is just ballast to make his theory sound interesting. the water representation is just an analogy. you could make the same representation using a system of compressed air and vacuum. and then it would completely fail as an analogy but it would still work right.
It;s like trying to explain why an orange is an orange and an apple is an applle. they are both round and they can both be green ( when not ripe ) but a ripe apple is red and a ripe orange is orange. they are still both round and both have seeds. so there is something else that sets them apart...

anyway. analogies are just that analogies. you could argumentate that the water analogy is false as it cant 'shock you if you touch it', yet eletricity will... of course we are not after that property , we only use the fact that it is a medium that flows through a pipe. if i would have a system that works on the principle of a rond object moving though a pipe io could throw in an apple or an oragne. i only use the aspect that both are round. That's why it is an analogy.

Current is indeed the movement of electric charge ( carried by electrons)

Silicon is a semi-conductor : it is short an electron on its outer shell and want to 'share' electrons with its neighbour. in its natural state , having  one electron short , silicon does not conduct. Feed it an electron ( by seeding it with a dopant) and all hell breaks loose the silicon atoms will hsare this electron and charge can all of a suddne move. it has become conductive !. So, we need to create 2 materials from this pure silicon : one with excess electrons and one that is really short in electrons :

N material is pure silicon doped with an electron-seeding material like Phosphorous , Arsenic or Antimony ( which one you pick depends on production process. Phosphor diffuses quickly and is used for bulk seeding of wafers. Arsenic diffuses very slowly so it is very suitable for selective ion implantation. this is more chemistry and material physics than anything else. ) It has an excess of electrons. This group of donor material is called the Group V of the chemical table or the Nitrogen. All the elements in this group have 5 electrons on their outer shell. Silicon is happy with 4 electrons on its outer shell , but it only has 3 ... By seeding it with material containing 5 electrons , the silicon atom will 'share' the electron.

P material is pure silicon doped with an electron- accepting material . Typically Boron, but could be Aluminum or Gallium ( Shottky diodes actually use the aluminum as the P dopant hence their 0.2 .. 0.3 volt forward voltage. ) Any material in Group III ( having only 3 electrons on their outer shell ) can be used for this.
Silicon is already 'short' one to make it happy , and now we even derive it 'more'

If you slap a piece of N material agains a piece of P material you create a depletion zone. In a very thin layer the sharing of electrons goes 'haywire' . the excess electrons from the N material are absorbed by the 'two-short' P material and you get .. inert silicon... or non-conductive silicon ( there is no carrier available to move the charge. so it isolates )

This process forms a thin layer all by itself. once no more electrons can travel across the barrier the barrier stops widening. Apply pressure and you may get a few extra electrons across only widening the barrier . the thicker the barrier the better it isolates. This is done by applying an outside electron source to the P material will stealing electrons from the N material. The diode does not conduct.

On the other hand if you pull away electrons from the P material there is room for some to come over from the N material. By sucking away electrons from the P material you pull electrons out of the N material ( that's why that is called the emitter : it emits electrons !. You are actively making this depletion layer thinner and electrons can pass.

now , if you make an N-P-N stack and you polarise this thing so that you 'pull ' hard on the top N material , and inject electrons in the bottom using one battery ( + on the top N material , - on bottom N material ) no charge can move as you have TWO depletion layers..... one between bottom N and P and one between top N and P...

using a second battery you pull electrons out of the P while feeding them in the bottom N material. What happens now is that you REMOVE the depletion layers so electrons can flow from bottom N material into the P material and out of the base elctrode. But since the depletion layers have been removed electrons can also flow between bottm N material and Top N material ( the collector where electrons are collected ) . The name of 'base' comes from how a transistor was initially constructed. the P material was the base onto which the rest was built

In an ideal transistor you can reverse the contacts of Emitter and collector. However , such a transitor is very wasteful as the electrons have to travel a long distance across the P material. So in a real transsitor they make the P material very thin , have a large emitter ( as it delivers electrons to both the collector and base ) and they play with the ratio of 3 to 5 doping in such a way that more electrons take the route to flow to collector than flow into the base. this is transistor physics)

Anyway i have posted on the forum here a very long article where i explained this including drawings of how the depletion layers move and how the electrons run.

That guys explanation is semi-correct

[LoyalServant]

Thank you for filling in the gaps, free_electron.
The article and your write-up make for some very interesting reading.

[free_electron]

go read this :

https://www.eevblog.com/forum/beginners/how-transistors-work/msg109166/#msg109166

This is where i give the 'long ' explanation. read all subsequent posts in that topic. As it explains even more details.

Transistors are actually pretty easy to understand , if it is explained correctly. just slapping the equation down doesn't do you any good because it doesn't tell you what is happening in the material. A lot of the operation of the transistor is depeendent on its physical construction. the chemistry gets you to point a, the equiations get you to point b , but to get the thing working you need to carefully craft the shape of the electrodes in respect to each other to make sure the electrons go where you want them to go.
and then there is all the dissipations in and around the depletion area. that's why we have all these weird physical contructions like pinch-gates, mesa's , planar, gunning , tunneling and 400 other ways to stack this tower of NPN. none of whihc is covered by emers moll equations but by ton's of other material physics and field related theory.

in practice : you draw a bunch of shapes your gut tells you should work , spin the wafer and run them through the curve tracer and then pick the best ones. if you got a lot of time and money you will learn a lot and be able to draw a shape that works better than another. that is what semiconductor companies do. they spend a shitload of money designing better transistors. can;t change the laws of physics , you can just build better channels to conudct more electrons and create a better doorway mechanism... one that lifts smoother , has no backlash. ( look up 'snapback' for a transsitor.. you'll go fuzzy eyed trying to understand that one ... especially if you approach it from the mathematics side. the mathematics say it doesn't exist yet every transistor has it... the electron will damn well flow where it wants to go, even if a formula say's it can;t go there... electrons are deaf and blind. but they are damn good at sniffing out a place that has a different charge than they have)

oh and before anyone starts hammering on the 'one electron short' i explained here and the 4-short i explain in the above article ... Here i explained what happens between an intrinsic silicon system and a dopant. the dopant delivers a free electron.

in the intrinsic system the sharing happens between neighbours. you need to think 3 dimensionally ( i sound like Doc Brown... you gotta think four dimensionally Marty, in reality that billboard with the indians isn't there) this stuff sits in a 3d-lattice so electrons come from left, right, top, bottom, front and back. the exchange is not between 1 silicon atom and 1 donor ( there is only one donor per x silicon atoms. how many donors we shoot in controls conductivity capability . ) so this exchange happens between a silicon atom in the center and all its neighbours in the lattice. it is sharing electrons between all its neighbours. semiconductors are very wonky materials. they are essentially natures misfits... ( if you think about what i just said you will find something wrong. silicon wants to go from 4 to 8 electrons.  but it has 6 neighbours... left, right , top,bottom , front and back... so why 4 and not 6 ? well because other crap going on with the weak nuclear forces in an atom )

here is a nice representation of how this sharing works ( in 2 dimensions):

http://hyperphysics.phy-astr.gsu.edu/hbase/solids/sili.html

in reality the sharing is 3 dimensional so if you really want to whack your brain out  ( this shows why 4 electrons and not 6) :
http://hyperphysics.phy-astr.gsu.edu/hbase/solids/sili2.html#c1

And that is just intrinsic silicon ( non-conducting , no excess electron , no short an electron) shoot a dopant in there that steals one electron or delivers a free one and your brain will try to eat it's own ass and desintegrate in a big 'pfffwwrrrrt i'm outta here , this is too complex... '

kind like Data feeding the Borg that funny hyperdimensional figure that cannot exist..

[cwalex]

WOW! Thanks for all the great replies. free_electron, you always explain things in excellent detail and I really appreciate the time you take to make sure nothing is missed. I will read through your reply a few more times to make sure it all sinks in

[NewBeginner]

Electrical current is the movement of electrical charges. Electrical charges are carried by electrons.
[...]
Current is indeed the movement of electric charge ( carried by electrons)

I may be wrong but electrons are not the only charge carriers (ex: in gases and liquids).

The author seems to be a Research Engineer ( Chemistry Dept. , University of Washington, Seattle  http://amasci.com/me.html ), so I tend to think that he knows what he is talking about. But again, I may be wrong .
Anyway the purpose of those articles (if I understand correctly) is just to bring some clarifications on common misconceptions about different electricity related topics (including the common misconception that electrons are the only charge carriers or that a current is always a flow of electrons).

Some of his explanations are harder to digest. Even if the topic about transistors is not fully detailed I still find it helpful. I mean it helped me to better understand transistors.

Thank you

[McMonster]

So it really isn't Art of Electronics' transistor man adjusting the dial in each device?

[Mechatrommer]

Electrical current is the movement of electrical charges. Electrical charges are carried by electrons.
Current is indeed the movement of electric charge ( carried by electrons)

I may be wrong but electrons are not the only charge carriers (ex: in gases and liquids).

the notion of "current" that we know is the movement of +ve charges, the electrons movement (-ve charges) are backward. thats what i understood since i learnt it, pls prove me wrong.

[madires]

P material is pure silicon doped with an electron- accepting material . Typically Boron, but could be Aluminum or Gallium ( Shottky diodes actually use the aluminum as the P dopant hence their 0.2 .. 0.3 volt forward voltage. ) Any material in Group III ( having only 3 electrons on their outer shell ) can be used for this.semiconductor
Silicon is already 'short' one to make it happy , and now we even derive it 'more'

Schottky diodes aren't based on a p-n-junction. They got no p zone, they got a metal film on top of a n semiconductor, i.e. a metal-semiconductor-junction.

[NewBeginner]

Electrical current is the movement of electrical charges. Electrical charges are carried by electrons.
Current is indeed the movement of electric charge ( carried by electrons)

I may be wrong but electrons are not the only charge carriers (ex: in gases and liquids).

the notion of "current" that we know is the movement of +ve charges, the electrons movement (-ve charges) are backward. thats what i understood since i learnt it, pls prove me wrong.

I am not sure what do you expect me to prove . I am not trying to prove something (I mean no disrespect). I am still in the process of learning physics/electronics (and I hope I always be) .
Here are some references about charge carriers:

http://physics.bu.edu/~duffy/py106/Charge.html
http://en.wikipedia.org/wiki/Charge_carrier

Physics Vol (1 and 2) - Page 318 - Q.26
http://books.google.ro/books?id=DUkVFmKxEqQC&pg=PA318&dq=current+charge+carriers&hl=en&sa=X&ei=GSGqUID_KcW0tAbU2oDABg&redir_esc=y#v=onepage&q=current%20charge%20carriers&f=false

Thank you

[ptricks]

Try DIY transistors, makes it really clear how they work.

[free_electron]

P material is pure silicon doped with an electron- accepting material . Typically Boron, but could be Aluminum or Gallium ( Shottky diodes actually use the aluminum as the P dopant hence their 0.2 .. 0.3 volt forward voltage. ) Any material in Group III ( having only 3 electrons on their outer shell ) can be used for this.semiconductor
Silicon is already 'short' one to make it happy , and now we even derive it 'more'

Schottky diodes aren't based on a p-n-junction. They got no p zone, they got a metal film on top of a n semiconductor, i.e. a metal-semiconductor-junction.

I worded that wrong. Aluminum itself has 3 electrons on its outer shell. What happens is that at the junction of a piece of n material and a strip of aluminum you create a depletion zone. So in essence you do not start by creating p material , but you apply the aluminum directly.

[free_electron]

I may be wrong but electrons are not the only charge carriers

Electrons and protons. The temperature has no impact. ( temperature of an item controls if it is solid , fluidic or gaseous )
And the subject of study is electronics, not protonics... Ring a bell now ? If you want to do protonics : talk to the boffins at CERN, they'll sell you a few cyclotrons so you can start ... Cyclotrons that operate on... Electronics..

The author seems to be a Research Engineer ( Chemistry Dept.

i've been making transistors for the last 20 years of my life. (And i've done actual layout of single transistors including making the p-cell generators for them). Granted there's some bits where i too am abit fuzzy and need to rethink it when i am explaining it because it is not something you do routinely. Most often you pull one , characterized, out of a lib and plonk it down. More frequently you pull an entire circuit like an opamp or a logic gate out of the lib and plonk it down. You don't care about individual transistors. Unless you are mucking about making a custom i/o cell finetuned for what you electrically need and the size of it. Then you need to know how to put the n and p wells down, how to shape them and how to run the metallics. And you also need to know where to shoot the substrate barriers and make the via rings to shield one transistor from another.
If you'd ask me to explain that i'd have to go back to my notes as that was 15 years ago i did that last. It's not everyday material... Unless you are a full time chip layout person. Those guys know. Not my job. I did that for 6 months as a learning experience to get a better understanding of what was at play , where the gotchas are and the tricks.

[NewBeginner]

@free_electron
I think you misunderstood me. I was not challenging your knowledge on the subject . It is clear to me that you have more experience than I could ever dream of in the field of electronics.

And the subject of study is electronics, not protonics... Ring a bell now ?

Please don't take this personally but I find this to be funny . I don't think that "electronics" means that we should only deal with electrons or that we should only learn about electrons.
In fact most of the electronics books I read so far explained that the electrons are not the only charge carriers and that a current is not always a flow of electrons.
If I understand this correctly a current is a flow of charge.
I believe that in order to better understand electronics one must also understand the basics of electricity and electrical circuits.

You said that:

Electrical charges are carried by electrons.

So I was just pointing out that this information is not complete (since this is a very common misconception).  For example, if I understand correctly, charge can also be carried by ions (ex: the case of  salt water - http://van.physics.illinois.edu/qa/listing.php?id=2295). In this case the current would be a flow of ions.


Another reference:

Physics Vol (1 and 2) - Page 318 - Q.26
http://books.google.ro/books?id=DUkVFmKxEqQC&pg=PA318&dq=current+charge+carriers&hl=en&sa=X&ei=GSGqUID_KcW0tAbU2oDABg&redir_esc=y#v=onepage&q=current%20charge%20carriers&f=false


Please note that I am not trying to offend you in any way. My mother tongue is not English so I may not always express myself correctly .



Thank you

[IanB]

I don't think that "electronics" means that we should only deal with electrons or that we should only learn about electrons.

It turns out in fact that electricity really is all about electrons, even when you consider things like positive ions. Electrons are the charge carriers that cause electricity to happen. When you have positive charge carriers like holes in P type semiconductors, or negative ions in solution, the positive charge is caused by a lack of electrons. Electrons are always involved one way or another. Just think of a positive ion as a particle where someone took away an electron or two.

From a physics point of view electrons are on the outside of atoms and can be added, removed, and moved around with relatively little energy. Manipulating electrons involves amounts of energy that are accessible and familiar to us in our everyday world.

On the other hand the positive charge carriers in atoms, the protons, are tightly bound in the nucleus at the center of the atom. Adding, removing or manipulating protons involves extremely huge amounts of energy that are far out of reach of any ordinary process or machine. We have to build big expensive "atom smashers" to get at that physics.

[NewBeginner]

Electrons are always involved one way or another

I agree.

I was simply trying to point out that a current is not always a flow of electrons (as some may think; i know I did at first). It can also be a flow of ions for example.
It may seem just a subtle difference but I think it helps to better understand things. It helped me.

If I physically/electrically misunderstood something please correct me. I always try to learn and better understand how electricity and electric/electronic circuits work.

Thank you

[IanB]

Electrons are always involved one way or another

I agree.

I was simply trying to point out that a current is not always a flow of electrons (as some may think; i know I did at first). It can also be a flow of ions for example.
It may seem just a subtle difference but I think it helps to better understand things. It helped me.

If I physically/electrically misunderstood something please correct me. I always try to learn and better understand how electricity and electric/electronic circuits work.

Thank you

It is true that ions may be involved in the flow of electrical current, but if you look at the overall system there are more layers to what is going on. Consider, for example, the following (very simplified) view of a conducting system:



Is it positive ions that are flowing between the electrodes or is it electrons?

[free_electron]

HOLD IT ! what is an Ion ? look it up.
It is an atom that has been forced to either accept extra ELECTRON(s) or had one or more stolen from it.. in essence the number of protons in the core is different from the number of electrons.
so it is NOT the Ion that carries charge. it is still the electron. and you are either short or have too many.

Whie it is possible to do this to an individual atom (i ttakes a lot of 'force') it is most commonly done in molecules.
electrons are already being shared between individual atoms in the 'bond'

iron can lose electrons : Fe²⁺ is an iron atom that is short two electrons
The electrons are always stolen from the outer shell. ( the so called valence electrons )

Natrium (Sodium if you are english/american) ( Na) has only one electron on its outer shell and a bunch on inner shells. it tends to 'loose' this electron very easily.
Chlorine has 7 valence electrons and is actually 1 short  ... So that is how you create salt. Na⁺ + 1e (the natrium ion short 1 electron + 1 free electron ) and Cl^- (Chlorine ion short one electron) readily combine to make NaCl or kitchen salt.

The charge is still transported by the electron. The NET charge of an atom or molucule is the remainder after subtracting the total number of electrons from the total number of protons in the atom or molecule.
since protons belong to the nucleus (or nuclii ina molecule) they can't be moved. electrons ont he other hand can be moved. so the charge carriers are electrons as they are the only movable and exchangeable elements.

of course now some boffin is going to step in and say that, if properly accelerated and smashed the charge is not actually carried by the electron but is a perceived effect of the spin of some quark that makes up the electrons. or some coupling between a charmed quark and another in the core... beyond me. i'm not a chemist o physicist. i deal with electronics. electrons carry a 'charge' which results to a 'force' we perceive as energy.  by taming the electron and letting it flow where we want we can do things. This 'taming' process is called electronics. i dont deal with bosonics protonics quarkonics ( unless it involves raspberry ... mmm quark pie with raspberry sauce ... ) or any other muckery with elementary particles. not my field of work .

and by the way the nacl example i ripped of wikipedia. i'm not a chemist either... i vaguely remeber something. if you want to know more : please look up the term 'ion' in wikipedia( or a more reputable and complete source )

[NewBeginner]

I think we are talking about slightly different things.
I was simply pointing out that a current is not always a flow of electrons.

A quote from amasci.com http://amasci.com/miscon/eleca.html:

1. All electric currents are flows of electrons? Wrong.
Electric currents are not just flows of electrons, they are flows of any type of electric charge. Both protons and electrons possess exactly the same amount of 'electricity.' If either the protons or the electrons flow, that flow is an electric current. In salt water, in fluorescent bulbs, and in battery acid, atoms with extra protons can flow along, and this flow is a genuine electric current. And in fuel cell membranes and in solid ice, electric current is actually a flow of single protons (go and look up "proton conductor.")

Some additional resources:

http://www.protonconductor.com/
http://en.wikipedia.org/wiki/Proton_conductor


Thank you

[jahonen]

Also, as an example, the beam intensity in a particle accelerator such as the LHC is also measured by current transformer, since there is a direct relationship between protons in the beam and measured current. In LHC, each beam can hold about 2800 bunches of protons and each bunch can contain about 1E11 protons, then we have a charge of 2800*1E11*1.6E-19 = 44.8 µC circulating in each beam, and since the beam makes about 11000 turns per second, then we can calculate the beam current as 44.8 µC*11 kHz = 493 mA.

Regards,
Janne

[free_electron]

The name proton conductor is used to indicate an ION that has less electrons than protons in its nucleus.

It is not the proton that moves but the whole ion. There is no way you can move a single proton as they reside in the nucleus of an atom. You would need to split the atom. This requires... A nuclear reaction (nuclear from the word nucleus meaning : core ).

Electrons do not reside in the nucleaus but spin around it. Electrons are movable. Neutrons and protons require carving up the nucleus.

Now, protons do indeed carry a charge. Exactly the same as an electron but with a positive sign. So if you move the whole ion ( which is what is happening in fuel cells and lithium ion batteries ) you can move a charge. But keep in mind you are moving more than just the proton. It is the whole core neutrons and protons ) and the associated electrons that moves.

To move i dividual protons you need to split the atoms and extract the protons. You can indeed mesure those with a current transformer. As they carry charge and are moving they will create a magnetic field. A moving magnetic field induces current in a conductor as it passes by. In essence the flying proton goes through the hoop of the current transformer. As it goes through , an electron in the winding says: hey, i want to come along... You magnetically attract me.... So it starts moving too .. But finds out it is trapped in the copper wire so it has to let go of the passing proton. Its like a kid running along a train waving goodbye. The proton sits on the train as it drives through the station. The kid meets the train as comes in, runs along it to wave and stays behind as the train pulls out of the station. The train is the stream of protons, the kid the electron in the current transformer windin and the station the winding of the current transformer.

Since the kid moved we see this as a current. So yep. The physics work.

This is also how ion implanters work to make integrated circuits and transistors , which brings us back to how transistors work and are made.

We heat a carrier gas spiked with phosphorous, arsenic , or we vaporise antimony. This cloud is ionized, we strip some electrons. So now all these ions can be manipulated. Either electrostatically ( since the ions in this case are positive we can attract them with large negative voltages ) or magnetically. Since they are charged and travel they induce a magnetic field. Once they travel and pass by a static conductor wwith a current flowing through it so this conductor also emits a magnetic field, we can play with current polarity as to reject or attract the ion magnetically.

We pull a beam of ions, control how many we need per surface area,  accelerate them to control how deep we want to shoot them. And aim this beam at the surface of a pure silicon wafer. Just prior to the ions hitting the wafer we position an electron gun to nullify the charge of the ions. They have their final speed, they are where they need to be position wise and they can now hit the 'wall' safely. The reason we need to strip the charge by adding electrons is as not to destroy the small structures we are creating. The stream of ions would induce a current in the wafer. That current could be large enough to burn out a freshly created diode for example. Or blow through the gate isolation of a mosfet. So we strip the charge first.

What a wonderful world we live in.

[IanB]

I was simply pointing out that a current is not always a flow of electrons.

It always is, really. It's just that sometimes the electrons are not front and center where you can see them.

Did you look closely at my illustration above? Consider it to be a gas discharge tube. Electrons enter on the left, flow through the gas discharge, and exit on the right. There is clearly a flow of negative charge, carried by electrons, flowing through the gas. The positive ions are just mediators in the process.

If you tried to measure current by some external means (by looking at the magnetic field, for example), it would be impossible to tell if negative charges were moving from left to right, or positive charges from right to left. Any experiment you performed would just tell you there is an electric current flowing with the same direction and magnitude each time.

I don't think that amasci article you linked to is correct. It is just exaggerating in an attempt to be contentious. When ever "positive" charges seem to be carrying electricity there are always negative charges flowing the other way to keep things balanced. It is no more correct to say the positive charges are carrying current than to say the negative charges are carrying current.

Try a thought experiment. In a copper wire carrying an electric current, the electrons may be flowing along the wire with a certain speed (the "drift velocity"). For a given current and wire diameter let's say this speed is 1 meter per hour. So, let's coil up the wire stretched between two spools and feed the wire from one spool to another at exactly 1 meter per hour in the opposite direction. Now the electrons in the wire are standing still, and the positive charges in the wire are moving from one side to the other. But would we ever say the electric current is not about electrons in this system?

The truth is that sometimes the flow of electric current causes electrons to be removed from their host atoms or molecules leaving a positive charge behind. And sometimes these positively charged atoms or molecules move around under the influence of electric fields. But always, if you trace cause and effect back to the source, you will find electrons also moving around. The moving positive charges are just the shadow, the counterbalance, the mirror image of the electron flow that is always there.