You think that only because you don't understand QM.
They're not particles - they're waves. When they're in the conduction band, they don't have a fixed locality, so you can't measure them individually passing anything.
Only because your meter can't count electrons. I don't care exactly where the electrons are. When they pass through my measuring instrument they count as current. If they don't pass through, they don't count. If your instrument can't count electrons then that is your equipment, not the nature of electron flow.
-- Rick C
And speaking of valves, and barriers:
Even what would conventionally be considered quite a large barrier (the surface of a material), isn't much of one, necessarily.
The case being thermionic emission: the noise of a diode or triode, due to electron motion, is much less than shot noise.
How? The space charge pushes electrons back into the cathode material. With the space charge and cathode (bulk) populations in a rapid equilibrium, there is little room for noise generation. Meanwhile, the space charge itself grows and shrinks (in volume and/or "virtual cathode" potential), acting to "cushion" the noise of emission.
Diode noise is maximized only when the cathode is saturated, i.e., the space charge is stripped away from a substantial part of the surface. (Traditional cathodes have a craggy microstructure, so there is plenty of space charge hiding inside pits, or indeed, carrying current between grains -- part of the reason why a cathode of otherwise-insulating ceramic particles carries a current as well as it does. But once the majority of emitting surface is saturated, the noise level is mostly full shot noise.) This is usually achieved by using a high plate voltage, a weak emitter (like pure tungsten), and controlling cathode temperature to set cathode current.
In contrast, screen-grid tubes (tetrodes and up) suffer from "partition noise", where an electron in transit can only either proceed, or gets sucked into the screen. That binary choice creates full shot noise in the screen current. (Coming from space charge theory, it would seem the transiting electron beam doesn't have much 'cushioning' value, which says something about its density I suppose.)
The plate and screen noise are correlated of course, so some audiophiles like to "cascode" the screen current back into the plate, making a composite device with the gain of a pentode and the noise of a triode. (They don't realize that the transistor is necessarily performing amplification work on their precious signal in the process, but that's okay. They can think whatever they like... as they are wont to do.)
Tim
-- Seven Transistor Labs, LLC Electrical Engineering Consultation and Contract Design Website: http://seventransistorlabs.com
One per color per pixel, plus maybe some address decoding along the edges.
They're digital, in the ECL sense: on or off, where "on" is "not off", but not necessarily saturated.
Actually, those should "saturate" okay, because AFAIK, their action is discharging a capacitor (the LCD pixel) to whatever the voltage-being-written is (which may be recharging instead, but that's fine). An analog switch, in the "on" state, with ~0A and 0V on it, could be said to be saturated.
So, a couple million. If there's decoder and driver logic on there, add maybe a few thousand to a few million (the latter if it's buffered, which isn't common on higher resolution panels AFAIK).
There still has to be some "analog" transistors in there, for the voltage-being-written line drivers, but that's a DAC (more analog switches?) and an op-amp per line (so, dozens of transistors at the most).
Tim
-- Seven Transistor Labs, LLC Electrical Engineering Consultation and Contract Design Website: http://seventransistorlabs.com
Shot noise sounds pretty digital to me, too.
A current meter is a loop of wire in a magnet field, which rotates a pointer. The covalent metallic bond that holds the wire together is a set of allowed electron states that extend over long distances. No point particles, no knowable positions.
You can count the electrons, though, if you make your ammeter with a loop of nonconductor wire. Your count will be zero...
All analogue signals are composed of discete electrons, that doesn't make t hem digital. Digital signals have a small number of levels, each of which i s detected by the following input as separate digits. So an 8 bit DtoA with no filter on the output gives an analogue signal, even though it's stepped . The flow of small numbers of electrons only becomes digital when you coun t them individually, or detect separate current level bands.
NT
If I did that's not at all what I meant. I would like to know, at least to an order of magnitude, how many transistors are "linear" (I guess by Don's "if it's not counting 0's or 1's then it's linear" definition), and how many are "digital" (i.e., "acting like switches").
-- Tim Wescott Control systems, embedded software and circuit design I'm looking for work! See my website if you're interested http://www.wescottdesign.com
Since the charge itself is quantized, there are no continuum mode devices. :-)
Best regards, Piotr
And we can measure velocity.
What is your point?
Dan
1nV | .---||-----[1T]----. | | | | | `------------------' wait 4 minutes.
unfortunately last I checked digikey were all out of nanovolt batteries.
-- This email has not been checked by half-arsed antivirus software
But I do have a couple of 1Tohm resistors. According to the typewritten label sellotaped on the side, they are 2.5% high at 100V and 1% high at 500V
Now all I have to do is find a reason to use them :)
A lot of transistors are used in MLC flash chips. They have more than two states I guess.
Wouldn't they be spaced evenly in a vacuum too?
Not necessarily. It would depend on density, temperature and time.
The point I was trying to make, and which I have trouble to get across, is that even though charge is quantized, current need not be.
Jeroen Belleman
Rick, You know it's sorta impossible to convince you of something you think ain't so. You think current is discrete, we say it isn't. The good news is you can measure it! Shot noise is pretty easy to measure. (Schottky did it decades before Johnson measured thermal noise.) The current from a photodiode has full shot noise. The same current from a battery/ resistor has none.
So go measure it if you are interested. If you are not (interested) then talk about something else.
George H.
The world and conduction of electrons is explained by QM... so there's that. But I don't think you need QM to explain the lack of shot noise in resistors. You can make a classical model of electron scattering, and pretty much understand it.
George H.
Which also connects with my earlier comment about equilibria -- in this case, chemical. There's a clear barrier (the metal-electrolyte surface), but the energy barrier for a particle to exchange -- going one direction or the other, reversibly -- is very small, and driven by diffusion, which averages out massively on a macroscopic scale.
Physics is cool (except when it's hot, in which case it's still cool). :)
Tim
I'm sorry, what does QM stand for??? I believe the Q is for quantum, no? What *is* a quantum anyway? I believe it is a discrete packet. So what part of that says you can't count electrons?
In any current the electrons are the current carriers. You can't have a part of an electron pass a certain point in the path. So the current is discrete.
-- Rick C
None so blind as those who will not see.
Jeroen Belleman
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.