real nice explanation of transistors imho, may be handy

[pinkjimiphoton]

it seems to make better sense to me, but i'm pretty newb...

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

[R.G.]

Interesting. I'd have to study it in more detail to see if I agree with all of it.

My own personal internal model includes "depletion regions" too, but I think it's simpler than what he's saying.

All P-N junctions do cause a depletion region, generated by the P-ness and N-ness of the materials. It insulates because the electric field from the excessive-P and excessive-N sweeps the mobile charge carriers out of the region, making an insulating region where none of the charge carriers are free to move.

The depletion region can be made thicker or thinner by reverse biasing (thicker) or forward biasing (thinner) the PN-region. The inherent voltages that make the depletion region are not very big, and in silicon, you only need 0.7V or so to overwhelm them and force the depletion region to nearly zero. Well, it's still there, but the forward bias forces in so many charge carriers over the objection of the intrinsic voltages that the region is flooded in carriers and conducts merrily.

If a junction is reverse biased, the reverse bias is in a direction that reinforces the depletion region, keeping anything from going through.

A transistor has one reverse biased region (C-B) and one forward biased region (B-E). What makes this "amplify" is that charge carriers going into the base "poison" the C-B depletion region by injecting charge carriers in and proportionately letting charge carriers through. Base current (or voltage changing the depletion regions, etc, etc) weakens the ability of the collector-base to hold off the reverse bias.

[DavenPaget]

I actually truly understood after actually reading the entire article .
I seemed to know a bit that it is indeed everything is voltage controlled , the fact that current can't flow .
It just reinforced what i had in my mind as being very hazy , messy . 

[PRR]

> see if I agree with all of it.

Seems messy to me.

The new concept of "current" should perhaps be a separate essay.

The conclusions get about half-way to Hooper and Cherry's classic book which treats ALL amplifying devices (tubes, JFETs, BJTs, MOSFETS) as "charge controlled". However Hooper/Cherry is a heavy read too. "Charge" is not a familiar concept for most of us, not as en engineering quantity with units.

A key question for students: Do you need to know how a transistor "really" works? Or would you rather use "black box" rules-of-thumb which give useful answers for most practical problems?

While I have delved the "how- it really works" depths, when analysing or designing I never think about holes and depletion zones. The dichotomy between current-driven versus voltage-driven is often swamped by other circuit elements; usually one fiction or the other is simply easier to apply.

I like the "PPPPPPPPPPPS":
This article apparently has triggered extensive debates if not flamewars on multiple hobbyist forums...

[artifus]

[R.G.]

A key question for students: Do you need to know how a transistor "really" works? Or would you rather use "black box" rules-of-thumb which give useful answers for most practical problems?

I actually prefer the explanation that Maxwell's Demon sits inside the base region and only lets through the charge carriers he likes the looks of. No half-spins need apply.

[PRR]

> Maxwell's Demon sits inside the base region and only lets through the charge carriers he likes the looks of

Not a useful design fiction, since "likes the look of" is un-defined.

We could say a demon which needs 0.5V-0.7V of push, passes collector current, and leaks 1/hFE at the base. This suffices for most linear design.

A few more notes about current- and voltage- cut-off, you have switching/logic.

I actually think abstractly of a model which could be described as "bus terminal". You need 0.6V of inducement to hire a bus driver through the door marked "base". That driver can transport 50 passengers entering at the door marked "collector". The 50+1 persons leave through the "emitter" door. You can control the flow of a large number of people into a stadium by controlling the flow of bus drivers.

[frank_p]

Seems messy to me.

I don't know if it's only me but the author announce a big revelation with his own treatment of the subject, but in the end, it's only the charges/holes view.  I've recopied the article to be sure I wasn't missing something and if I remember well there is nothing more than what is in most textbooks like Malvino.

With better graphics, absence of semiotic crusades about incorrect views of 'flow vs current' and less self confidence in the pursuit of truth, it would perhaps be a relatively upright article.  I am no expert on the subject but it has given me that very messy feeling too.  You finish reading the article and you wonder yourself: '... I've missed something or what ?' '...What have I learned about the real physics of that BJT'

I think that my opinion is: DON'T READ IT IF YOU ARE IN A NEED FOR RELATIVELY CLEAR IDEAS ON THE SUBJECT.

[digi2t]

WOW  , that is a really cool video. I just have one nagging question now...

What the hell is that cheesy keyboard, and where can I get one? 

[pinkjimiphoton]

the guy has rants about current, too.

thanks for the feedback...to me, as a newb, it seemed READABLE..that's why i posted it. it's kinda dummied down like torres's "inside tube amps", which actually gave me enough knowledge to work on amps.

the scary thing, is i understood it, which is kinda rare. the website there is a hoot, in general.

thanks guys.

[PRR]

me> Hooper and Cherry's classic book which treats ALL amplifying devices (tubes, JFETs, BJTs, MOSFETS) as "charge controlled".

Beaty has an essay on Charge: http://amasci.com/miscon/fund.html

He argues that Coulomb is the fundamental unit, Amperes is a derived unit.

How do we measure "current" in a familiar case?

"This ditch is flowing 1,234 gallons per hour"

To macroscopic eyes, water is a continuous (no particles) substance, and we measure it in a bucket.

To macroscopic eyes, electric current is a continuous (no particles) substance, and we measure it in a bucket.

The electric bucket is the Coulomb.

People who can count electrons say there are 1,600,000,000,000,000,000 electrons in a Coulomb.

If I have a Coulomb-bucket of electrons, and they pour-out in one Second, that is one Ampere.

What Hooper/Cherry say is that to instigate flow in VT/BJT/FET we must shift the input pin with some number of Coulombs.

VTs and FETs mostly hold this charge and keep conducting after you stop pouring Coulombs, at least for a little while. We can observe this on the bench with CMOS. Large MOSFETs working fast need big drivers to dump Coulombs onto the Gate quickly, but the holding current is nearly negligible. BJT flowing collector-emitter "leaks" charge from the base, so we must pour more Coulombs in constantly. In all low-speed work this leakage dominates design, hence the Beta design parameter.

[frank_p]

People who can count electrons say there are 1,600,000,000,000,000,000 electrons in a Coulomb.

1.618 is the divine proportion or golden rule 
replace by 6.24e18 charges/Coulombs

The Truth, the Most Fundamental and the Divine,
now we are talking about the real mysteries of the Trinity.

[pinkjimiphoton]

pink is so, soo, sooooo hopelessly lost at this point...lol

they speaketh the magic words, we all know them, but they understand them, praise voltaire...

[mac]

As physicist I agree with the author, the only things that have real existance in nature are the charge, the mass and the spin. Voltages and fields are useful concepts but we need to keep in mind that the charges create them.
An example, you learn to divide two numbers using a pencil and a sheet of paper (the charge). After you know how to do it by yourself you write some code to crunch the numbers (the voltage).
I have also seen some people talking about current flow. This is a mistake since current is i=dq/dt, and a current flow is (if by flow we mean flowing in time) di/dt=d2q/dt2, the second time derivative of the charge. Think space, velocity and aceleration.

So strictly speaking, tubes, fets and bjt are charge controlled devices.

But since we are macroscopic creatures dealing with voltages and fields is easier for us.

(Now that I'm old and science is a remote memory, I'd say that everything is money controlled.  )

mac