Nubie transistor question ...

[moonbird]

Up to now I have stayed in the tube world (still a bit of a noob there too I think). In trying to learn about transistors, comparisons certainly suggest themselves. For one thing .. can someone help my understand the difference between Hfe and 'amplification factor' in tubes. Is there a formula or something or am I really off base (most likely). thx!

[eyoung]

Mu is the rough tube equivalent to Gm in transistors I'm still learning myself tons of reading on the web and in the forum . Now get to it laddie...

Regards, Elwood

[apls]

tube is input voltage control ouput current
transistor is input current control ouput current
hfe means ouput current / input current
amplification factor means ouput current / input voltage

sorry for my english :-p

[Miles Prower]

Originally Posted by apls

tube is input voltage control ouput current
transistor is input current control ouput current
hfe means ouput current / input current
amplification factor means ouput current / input voltage

sorry for my english :-p

That's not right. BJTs are voltage controlled, and base current is a bug, not a feature. If it were possible to get all those minority carriers across the base without recombination, they'd gladly do it and eliminate base current completely.

As for amplification factor, this is:

deltaVo / deltaVi @ Ip= const (dimensionless)

Output current / input voltage is transconductance.

[Andrew Eckhardt]

Yes miles, but he asked what Hfe is, which is Ic/Ib. Any "noob" should definitely check out the wikipedia BJT entry. Concise.

[Mooly]

Originally Posted by Miles Prower

That's not right. BJTs are voltage controlled, and base current is a bug, not a feature. If it were possible to get all those minority carriers across the base without recombination, they'd gladly do it and eliminate base current completely.

As for amplification factor, this is:

deltaVo / deltaVi @ Ip= const (dimensionless)

Output current / input voltage is transconductance.

That's not a very helpful comment to anyone trying to understand transistor theory. Transistors are current driven devices.

There are many "h" parameters. As a start,
Bipolar junction transistor - Wikipedia, the free encyclopedia

[moonbird]

Well -- I guess what I am after is finding a way to compare small sub-mini tubes to transistors - I have an idea to replace transistors with sub-minis in some circuits I am looking at, however the "gain" that transistors can achieve seems apparently greater (which is why they won out afterall).

Guess I am trying to get some kind of sense of the gain difference to assess feasibility before going off to study in detail. Looks like I have some reading to do. Am I looking for a shortcut? Yeah I suppose I am <grin> ...

The current drive (si) vs voltage drive (tube) comment is very helpful - I obviously need to change my point of view it appears from the start. I have a basic understanding of the P/N junction but getting to the point of understanding specs will require more work. One of the trasistors I am looking at is an oldie -- the BC109. Is there a sub-mini that I could use in its place that uses 130v or less on the plate (Yes thats right -- I am going backwards in time <grin> we tube guys always are)

I asked the question at a more general level not to be coy but rather to offer respondents more "freedom" of though if you will. A more focused question might be ... are there "known" analogs between transistors and sub-mini tubes. I have read that the sub-minis arose at the end of the tube era as a last ditch effort to compete with semiconductors. Does anyone know of any "head-to-head" devices. These rivals would be great fun to play with (great fun for someone kinda strange like me anyway <grin>).

I appreciate the input from all.

[Claude Abraham]

Originally Posted by Miles Prower

That's not right. BJTs are voltage controlled, and base current is a bug, not a feature. If it were possible to get all those minority carriers across the base without recombination, they'd gladly do it and eliminate base current completely.
As for amplification factor, this is:

deltaVo / deltaVi @ Ip= const (dimensionless)

Output current / input voltage is transconductance.

Wrong, dude. Have you studied semiconductor physics at the graduate physics or EE level? I'm not being confrontational, but you don't know this subject as well as you think you do.

Even if all carriers "emitted" from the emitter should survive all the way into the collector, there is still a base current that is non-zero. In order to forward bias the b-e junction and create the b-e E field, some carriers, holes for npn device, are injected into the emitter. Recombination takes place in the emitter and this component of base current is called the injection component. Typically, for silicon, injection is much greater than the transport component. So even if the transport component of Ib was zero, the injection component is still needed. The transport component includes the carriers recombined in the base.

There is yet another all-important component of Ib, the charging component. Any change in E field requires a non-zero displacement current as well a a non-zero voltage. As frequency increases, more base current is needed since charging the capacitive junction is being done faster, hence more charges per unit time are needed. But the charges per time is the base current. That is why all devices have a finite value of "ft" the transition frequency. Above ft, beta is less than 1, and amplification is diminished.

Beta is all-important as is base current. So is Vbe and gm. A bjt needs both to operate, not just 1 or the other. The reason for classifying the bjt as current controlled has to do with external drive.

Again, which quantity, Ib, Vbe, or Ie, is responsible for establishing Ic? The answer is all 3 in unison. But only 1 can be directly controlled, with the other 2 being incidental.

Placing a low impedance voltage source directly across the b-e junction, i.e. controlling Vbe directly, is never done. The current Ic would be established per Ic = alpha*Ies*exp((Vbe/Vt)-1), the temperature increases, and the saturation current Ies, would increase, resulting in a runaway condition.

We can place a current source in the base lead, or the emitter lead. If we set the base current, Ib, to a specific value, we have a thermally stable circuit. But Ic = hfe*Ib. We have a condition known as "beta dependency" as hfe (beta) is not controllable to tight tolerances. Even manually selected devices for hfe exhibit substantial changes in hfe with temp.

Controlloing Ie, emitter current, results in Ic = alpha*Ie. This is both thermally stable and alpha varies very little over temp, current value, and device to device. Thus we control Ic with Ie, hence "current controlled".

Again, at the semiconductor physics level, is Ic established as a result of Ib, Vbe, or Ie? That is an endless vicious circle since all 3 participate. Neither can exist w/o the other 2. An equivalent question is as follows.

If I dim the light in my room using a dimmer, what was responsible for decreasing the light output, the decrease in current or voltage? The decrease in light output is due to which one? What is your answer?

I say 6 of one is half a dozen of the other. I and V cannot decrease independently. They are inclusive and simultaneous.

The contrarian point of view regarding bjt devices is a joke propagated by semi-literate hackers, and wannabes. Some may believe their position, others are likely just stirring up trouble. Every semi physics ref text uses charge control at the micro level. Beyond this level is quantum mechanics. The terms "voltage controlled and current controlled" do not even appear in advanced physics texts. That is because these terms are only used when externally viewing the device as a black box, only examining terminal characteristics wrt temp, freq, I, & V.

All semiconductor makers emphasize that the bjt is current controlled when using a mAcro viewpoint, and charge controlled at the mIcro viewpoint. Again "current control" is the best mAcro model. Current control was never meant to explain device physics at the micro, or atomic, let alone quantum level. As long as we understand that current control only describes the crude 1st order external mAcro level behavior, it is fine.

Ib, Vbe, and Ie are all equally important. A bjt needs all 3 to function. Neither is the "prime mover". Peace.

[SQLGuy]

Originally Posted by Miles Prower

That's not right. BJTs are voltage controlled, and base current is a bug, not a feature. If it were possible to get all those minority carriers across the base without recombination, they'd gladly do it and eliminate base current completely.

As for amplification factor, this is:

deltaVo / deltaVi @ Ip= const (dimensionless)

Output current / input voltage is transconductance.

It's a forward-biased PN junction, how could current flow be considered a bug?

I must say I'm pretty happy for all the buggy rectifiers in my amps.

[Bigun]

Originally Posted by SQLGuy

It's a forward-biased PN junction, how could current flow be considered a bug?

It makes the FET people feel happier to think they've picked the best technology

Actually, FETs are great too.