Driving a led with mains. / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Driving a led with mains.

first previous 9 10 11 12 next last

Move Thread

LAN_403

Ash Small

Sun Oct 06 2013, 07:03PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Mattski wrote ...

(I'm also wondering if sometimes when you talk about electrons you're referring to electrons in the valence band instead of holes. I don't recommend this since it's nonstandard and you'll confuse people, much like you'd confuse people if you used electron potential and electron current instead of voltage and standard current in circuit analysis.)

This is my point, Mattski. The electrons that undergo re-combination become valence electrons, don't they?.....and these are the electrons that form the base current. In the model I described above, we're ignoring the conduction band electrons that form the collector current, and just modelling the 'base current' electrons.

What I'm saying is that removing these valence electrons as 'base current' allows the conduction band electrons to 'diffuse' from emitter to base, where only some of them recombine and become valence electrons, the rest form the collector current..

Steve Conner

Sun Oct 06 2013, 07:38PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

This all goes to show how you can design lots of circuits that work fine, without ever having a clue how the devices really work! :) When I use a transistor, I just have a vague idea of electrons and holes jostling around inside it.

I do remember how the junction works from semiconductor physics class, though. When you join N-type material to P-type material, the surplus electrons in the N side don't just fall into the holes on the P side, because the holes are in a higher energy band than the electrons.

Applying a forward voltage across the junction gives the electrons the extra energy they need to hop over the potential barrier (or quantum tunnel through it) and drop into the holes.

Ash Small

Sun Oct 06 2013, 08:15PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Steve Conner wrote ...

Applying a forward voltage across the junction gives the electrons the extra energy they need to hop over the potential barrier (or quantum tunnel through it) and drop into the holes.

Yes, but applying the forward voltage causes the base current to start flowing, all I'm saying is that the electrons don't 'hop over the potential barrier' until you remove some valence electrons when the base current starts to flow. It's this that raises the energy level of the electrons at the junction in the emitter sufficiently to start hopping.

Steve Conner

Sun Oct 06 2013, 09:08PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Nope, that's not quite right. A transistor works just the same as a diode. Electrons come out of the emitter into the base when the applied voltage is enough to help them over the potential barrier. The difference is that most of them get sucked up by the collector and only a few stragglers get trapped on their way through the base.

I believe they mainly "get trapped" by hitting defects in the crystal lattice and a transistor made of perfect materials would have no base current whatsoever.

Mattski

Sun Oct 06 2013, 09:38PM

Registered Member #1792 Joined: Fri Oct 31 2008, 08:12PM
Location: University of California
Posts: 527

There are two contributors to base current, recombination of electrons in the base and diffusion of holes from base to emitter.

Impurities and defects do scatter electrons or create traps which make recombination easier, but even with perfect quality material (not possible since doping is an impurity) some electrons will recombine, though lifetimes can reach milliseconds in Si.

Holes also diffuse from the base into the emitter because it is a PN junction so you get both electron current and hole current. The hole current doesn't contribute at all to collector current, it's just wasted energy. One of the tricks that heterojunction bipolar transistors brings to the table is that you can create an additional valence band barrier to this hole current with a wide bandgap emitter. This helps improve the current gain by cutting off hole current.

Steve's right, once the barrier is lowered the electrons flow from the n-type emitter into the p-type base. The conduction band is nearly empty so there's plenty of states for the electrons to flow into without needing to remove any valence band electrons. Electrons are actually above the conduction band minimum in a distribution that tails off exponentially with energy, so that's how they can get over a barrier. Essentially some of the electrons are "hotter" than others.

wrote ...
When you join N-type material to P-type material, the surplus electrons in the N side don't just fall into the holes on the P side, because the holes are in a higher energy band than the electrons.

Actually the holes are in a lower energy band than the electrons. And if you could just join a p-type and n-type material then probably what would happen is the electrons would diffuse across to the P-side and fall into the holes. Similarly the holes would diffuse to the N-side and recombine with electrons. Then the process stops, or rather reaches a dynamic equilibrium, because the uncompensated dopants have a charge which creates an electric field, which creates a drift current to balance out the diffusion. In the process it forms the built-in barrier.

Ash Small

Sun Oct 06 2013, 09:54PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Mattski wrote ...

The conduction band is nearly empty so there's plenty of states for the electrons to flow into without needing to remove any valence band electrons.

I agree that the conduction band is nearly empty, but you need to upset the equilibrium that exists after the migration that occurs when the junction is formed before any electrons will migrate from emitter to base. I'd suggest that the reason that FET's work in the way they do is because some mechanism either eliminates re-combination, or prevents electrons from the 'base' region reaching the gate terminal (I've not looked into it yet), and I'd suggest that in order for anions (holes) to physically move, you either require a liquid electrolyte or a plasma.

Mattski

Sun Oct 06 2013, 10:32PM

Registered Member #1792 Joined: Fri Oct 31 2008, 08:12PM
Location: University of California
Posts: 527

Ash Small wrote ...

Mattski wrote ...

The conduction band is nearly empty so there's plenty of states for the electrons to flow into without needing to remove any valence band electrons.

I agree that the conduction band is nearly empty, but you need to upsdet the equilibrium that exists after the migration that occurs when the junction is formed before any electrons will migrate from emitter to base. I'd suggest that FET's only work in the way they do is because some mechanism eliminates re-combination (I've not looked into it yet), and I'd suggest that in order for anions (holes) to physically move, you either require a liquid electrolyte or a plasma.

Equillibrium is upset by the applied bias voltage, this allows the current to flow because the emitter is at a higher electron potential energy than the drain is. Fets don't need to worry about recombination because they're majority carrier devices, i.e. when they are conducting from source to drain all of the regions are either n-type or p-type. Therefore for an n-channel there aren't any holes around for the electrons to recombine with, so the lifetimes in the channel under the gate are very long.

You can think what you like about holes, I'd just say that the concept of classical motion applies equally poorly to holes and electrons. Does the bubble float, or the water sink? Take your pick. But we know that we can treat holes as moving particles and get a correct analysis, and that's really the important thing.

Ash Small

Sun Oct 06 2013, 10:55PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Mattski wrote ...

You can think what you like about holes, I'd just say that the concept of classical motion applies equally poorly to holes and electrons. Does the bubble float, or the water sink? Take your pick. But we know that we can treat holes as moving particles and get a correct analysis, and that's really the important thing.

I agree, it's a very useful model.

I also agree that as much energy is expended (and heat created) by trying to pull holes out of the crystal lattice as energy is expended (and heat it created) trying to energise the electrons to move.

The only 'effective' mechanism you have to 'create' the 'potential gradient', or whatever you want to call it is by 'moving electrons' (same as in a capacitor, for example). I'm not trying to challenge traditional models that have proved 'easy to understand' and 'produce credible results', I'm just trying to understand it in terms that I find easy to relate to.

How much energy is required to pull a nuclei out of a crystal lattice?

Mattski

Sun Oct 06 2013, 11:52PM

Registered Member #1792 Joined: Fri Oct 31 2008, 08:12PM
Location: University of California
Posts: 527

Good points Ash

wrote ...

I also agree that as much energy is expended (and heat created) by trying to pull holes out of the crystal lattice as energy is expended (and heat it created) trying to energise the electrons to move.

The phrasing is a little weird here so I'm going to clarify although this may already be your understanding. When holes move, they are not pulled out of the lattice, nor do you need to pull an electron out to create a hole. Holes are formed by other methods (dopants, thermal energy, light) and then they move under an electric field. Also I feel like "energizing" an electron to move is implying that it needs some minimum energy to move, like a forward voltage, but it does not.

wrote ...
The only 'effective' mechanism you have to 'create' the 'potential gradient', or whatever you want to call it is by 'moving electrons' (same as in a capacitor, for example). I'm not trying to challenge traditional models that have proved 'easy to understand' and 'produce credible results', I'm just trying to understand it in terms that I find easy to relate to.

True, only the motion of electrons can create an electric field in a semiconductor (unless you do something like shoot positive ions at a semiconductor). But the positive charge of the nuclei usually does play a role in the electrostatics even though they are immobile. Holes are a strange concept which take some getting used to, but it's a good way to do analysis. You could get the same results without holes but you'd get weird things like conductivity of p-type regions being inversely proportional to the electron concentration.

Probably quite a lot of energy is needed to pull a nuclei out, so much that things are already going very wrong in your semiconductor device :)

BigBad

Mon Oct 07 2013, 12:19AM

Registered Member #2529 Joined: Thu Dec 10 2009, 02:43AM
Location:
Posts: 600

yeah, yeah, yeah, holes, crystals and electrons are cool, but does it blow up when you wire it in the mains or not?

first previous 9 10 11 12 next last

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob