Register to reply

Why building a BJT transistor out of two diodes is impossible?

by iScience
Tags: building, diodes, impossible, transistor
Share this thread:
iScience
#1
Jul17-13, 12:05 AM
P: 310
i had a thought to stick two diodes together such that the electronic setup was similar to that of a BJT transistor. however i have been told by multiple teachers/professors that the two were not equivalent. I have drawn a picture to show you what i mean. the top two pictures are just the circuit setups. the bottom two pictures are the physical electronic setup that is being represented.
Please explain why these are not electrically equivalent.

http://i.imgur.com/6ssocN3.png

so basically my rationale was, that since a conductor is supposed to carry charge, ie, where ever there is an excess of charge, there becomes a potential difference and thereby allows the charge to travel through the conductor. so whenever one of the P material has an excess or a defficiency of charge, it will get transferred over to the other P type material in the other diode. and.. it gets transferred over through a conductor which of course can be tapped, thereby able to be turned into a BJT transistor setup.
Phys.Org News Partner Physics news on Phys.org
Technique simplifies the creation of high-tech crystals
Working group explores the 'frustration' of spin glasses
New analysis of oxide glass structures could guide the forecasting of melt formation in planetary interiors
DrDu
#2
Jul17-13, 01:47 AM
Sci Advisor
P: 3,555
The basis has to be very thin in a transistor so that the charge carriers which enter from the collector into it have a chance to leave it through the emitter before recombining in the basis.
With two separate diodes, this isn't possible.
marcusl
#3
Jul17-13, 09:31 PM
Sci Advisor
PF Gold
P: 2,059
Correct answer, except that charge carriers are actually injected from the emitter and diffuse through the thin base to the collector. (John Bardeen, coinventor of the transistor, purposefully chose names for these regions that are descriptive). The process is formally called minority carrier injection because the carriers injected from the emitter (electrons for an NPN transistor) form a minority in the base (which is P type for an NPN).

The minority carriers start to recombine in the base but many diffuse through to the collector if the base is thin enough. This is why two separate pn junctions can't display transistor action.

erst
#4
Jul17-13, 09:43 PM
P: 21
Why building a BJT transistor out of two diodes is impossible?

What a BJT does:

There's a bias between the emitter and collector. But current can't flow because the base presents an energy barrier to carriers (e.g. for NPN, the P region is higher than N in an energy diagram). But if you were to inject current into the base, this charge would build up and applies a potential that lowers the barrier. Carriers rush in from the emitter to restore local charge neutrality, but most of them shoot through the base without finding an opposite carrier to recombine with in time. This is why a small I_b causes a large I_c and I_e.

In contrast, the wire is a metal. It takes a long time to move through a metal because of scattering and the sheer number of carriers means it's very easy to find a recombination partner. If you shoot current into the wire in between the two bases, it just goes out the two diodes. No amplification is taking place because that's a manifestation of base transit times being shorter than recombination time. Also the currents I_b and I_c are in opposite directions, instead of the same as in a BJT (diodes only conduct one way and these two are facing opposite directions). There's no way to make a BJT like this.
marcusl
#5
Jul17-13, 11:11 PM
Sci Advisor
PF Gold
P: 2,059
Quote Quote by erst View Post
What a BJT does:

There's a bias between the emitter and collector. But current can't flow because the base presents an energy barrier to carriers (e.g. for NPN, the P region is higher than N in an energy diagram). But if you were to inject current into the base, this charge would build up and applies a potential that lowers the barrier. Carriers rush in from the emitter to restore local charge neutrality, but most of them shoot through the base without finding an opposite carrier to recombine with in time. This is why a small I_b causes a large I_c and I_e.

In contrast, the wire is a metal. It takes a long time to move through a metal because of scattering and the sheer number of carriers means it's very easy to find a recombination partner. If you shoot current into the wire in between the two bases, it just goes out the two diodes. No amplification is taking place because that's a manifestation of base transit times being shorter than recombination time. Also the currents I_b and I_c are in opposite directions, instead of the same as in a BJT (diodes only conduct one way and these two are facing opposite directions). There's no way to make a BJT like this.
There are a few misconceptions here. First, a bias is an externally applied potential, not an internal one. Second, the EB barrier is lowered by the applied bias, not by a charge build up. Third, a large Ic is not exactly caused by a small Ib. Instead, a large current is sourced from the emitter. You are correct that most of this current flows to the collector, so that only a small base current is needed to make up for charge carriers lost to recombination in the base. It is said that the BJT is a current amplifier, but internally it is in fact a voltage-controlled amplifier.

Despite scattering, net current flow through a metal is nearly instantaneous (it propagates at close to the vacuum speed of light.) Furthermore, there is no recombination to speak of in a metal conductor, so the description in the second paragraph misses the mark. The reason that two discrete diodes don't make a transistor is because the base region must be abutted to the collector region and must be thin compared to the minority carrier diffusion length.
erst
#6
Jul18-13, 12:01 AM
P: 21
marcusl -- The current picture is fully equivalent to the voltage picture. See e.g. Streetman Ch. 7. It's meaningless to say I_c is not caused by I_b; local (steady-state) neutrality demands it, so it's just as causal as a bias-based description. It's helpful to think of what would happen if a current was injected into the base and flood of carriers *didn't* come in from the emitter. The carrier build up would change the chemical potential, i.e. a bias has been applied. Which way cause-and-effect goes is not meaningful since chemical potential is a bookkeeping tool.

Metals do have recombination and it is the transit time of individual carriers that matters, not the signal velocity. There's a thermal distribution about the Fermi energy with constant scattering in and out of states above and below it. That's simply generation and recombination in semiconductor terminology; the term is not used because the incredibly high scattering rate (relaxation on the order of ~10 fs) makes it nearly meaningless. That scattering rate is also precisely why it wouldn't work, not the absence of a gap. A short enough piece of high relaxation time metal will preserve and transmit a hole or electron current (if the Fermi level lies within the contacting materials' gaps). It's not the material category (metal v. semiconductor), but lifetime v. length scale that prevents BJT action.
iScience
#7
Jul19-13, 02:32 AM
P: 310
guys guys!! would if i cut into the cathode part of both diodes far enough such that the total amount of P-type material is small enough in comparison to the N s.t the charge carriers can pass right through the base to the collector, and only a few electrons would get carried out through the base?

like so!

http://i.imgur.com/eKCvl2n.png


...well...except... just imagine that the P-type was cut into alot more
marcusl
#8
Jul19-13, 10:42 PM
Sci Advisor
PF Gold
P: 2,059
Well that's the idea. However sticking two pieces against each other creates huge impediments (scattering centers and surface traps) to current flow. There's a reason that classical transistors are made of extremely pure and nearly perfect crystals of semiconductor material.
iScience
#9
Jul22-13, 08:22 PM
P: 310
Quote Quote by marcusl View Post
Well that's the idea. However sticking two pieces against each other creates huge impediments (scattering centers and surface traps) to current flow. There's a reason that classical transistors are made of extremely pure and nearly perfect crystals of semiconductor material.
i couldn't get a simple definition/description of what S parameter is; what is the S-parameter?
marcusl
#10
Jul22-13, 09:51 PM
Sci Advisor
PF Gold
P: 2,059
You mean microwave S-parameters? Please start a new thread--this is a new topic.


Register to reply

Related Discussions
Does PNP transistor share the same formulas as an NPN transistor? Electrical Engineering 5
Typical building materials for an office building Materials & Chemical Engineering 0
Just How Impossible is Impossible?(Picking a subfield) Academic Guidance 5
Throwing a ball upward on top of a building, find height of building Introductory Physics Homework 1
Velocity to jump building to building Introductory Physics Homework 4