Re: Channel length in MOS

From: Jim Logajan (JamesL_at_lugoj.com)
Date: 08/17/04 

Date: 17 Aug 2004 04:50:18 GMT

Danny at Chrastina dot net <danny@chrastina.notreally> wrote:
>
> On Sun, 15 Aug 2004, Jim Logajan wrote:
>
>> jokerknowsthetruth@yahoo.com (Nanite) wrote:
>>
>>> I understand that in order to turn on a MOS, say Nmos, we gotta have
>>> a channel of electron form between the source and drain. I don't get
>>> this: why do we need a channel? I always thought that if you want to
>>> attract electron from 1 place to the other, you use positive
>>> voltage. So a channel of electron seems to me that the electron in
>>> the channel will repel one another. Could anybody pls explain on
>>> this?
>>
>> There are always as many protons in the channel as there are
>> electrons.
>
> Until you apply a gate bias, surely?

Well, I was thinking that the gate bias changes the charge distribution in
the substrate, but doesn't actually change the net amount of charge in the
substrate (on its own, anyway). But this wouldn't be the first time I've
posted incorrect (or incomplete) information!

Perhaps you are diplomatically pointing out that the electrons (or holes)
that gather to form the other half of the capacitance with the gate
constitutes the channel, which would have a net charge? If so, I guess I
wasn't paying attention to the difference between channel and substrate.
(Unless you mean something else and I've still overlooked something?)

I'm not sure if any of this helps Nanite - probably adds further confusion.

> When electrons get from the source to the drain despite there not
> being a channel, it's called "punch through". This can be a problem
> because it makes very short transistors difficult to switch off.
>
>> http://www.tf.uni-kiel.de/matwis/amat/elmat_en/kap_4/basics/b4_1_2.htm
>> l
>
> This gives me even me a headache, and I have a PhD in
> semiconductor physics. I think it's the alternation between black and
> red font.

Your PhD trumps my rusty physics BSc. :-) And the links you provided below
are better than the one I quickly grabbed.

> http://ece-www.colorado.edu/~bart/book/book/toc6.htm
> http://ece-www.colorado.edu/~bart/book/book/toc7.htm

Relevant Pages

  • Re: Channel length in MOS
    ... i was asking about the channel length. ... I was thinking that the gate bias changes the charge distribution ... > times as many Si atoms as there are free electrons in a channel like this. ...
    (sci.nanotech)
  • Re: Channel length in MOS
    ... I was thinking that the gate bias changes the charge distribution ... > in the substrate ... > gate constitutes the channel, which would have a net charge? ... times as many Si atoms as there are free electrons in a channel like this. ...
    (sci.nanotech)
  • Lets display in view of the political prisons, but dont cure the furious developments.
    ... rewarding in charge of a inheritance with the midnight is too ... Ayman's legend argues as yet our channel after we comfort ... contractors. ... proceed mortally than wind with Abu's intact assistance. ...
    (sci.crypt)
  • Re: Picture breaks up in RGB but not PAL
    ... RGB works fine with BBC1,2,3,4, ITV1,2,3 and Channel ... any idea what Sky charge for a replacement?". ...
    (uk.tech.tv.sky)
  • Re: Noise level between two ant types
    ... one cannot hear the individual charged particles. ... charged partcles impacting the antenna, but if it walks like a duck ... would seem reasonable that whatever is delivering charge to something ... One thought would be to make a pair of channel 2 dipoles with one ...
    (rec.radio.amateur.antenna)