h2g2 The Hitchhiker's Guide to the Galaxy: Earth Edition

Entry: Semiconductors for beginners - A687143
Author: S'pe;lug:x // Researcher, Ace!, Scout, Guru, ... // (1+8-3)×(1-1+7)=42 1.7 // see A672554, now a Guru! - U183117

Somehow this was kicked out of PR, to here it is again, not to tattered though, so input would be welcome.

Hi s'pe;lug:x,

Nice entry.

I worked at Texas Instruments for 20 years. Looks good to me. If you have the time and inclination, you could add something of the history. The first transistor in 1946. Kilby and the first "chip".

A little more about how the transistor works as a switch or as an amplifier (the name came from transfer resistor I think).
How about something on the process? The growing of the cystal, the photo-resistance process, the sputtering of aluminium to make the connections. Dicing, alloying, bonding.

It's fine as it stands too.

Awu.

Thanks for the comments Awu. I don't want to start delving into any of the history, as I think this would be best covered in a separate article as it could make this one too big (IMHO 1000 word articles are the best; less is often more). As this is a beginner's guide I also don't want to delve too far into the nitty-gritties of manufacture or into the technicalities of complex (and mystical) things like amplification.

spelugx -- writer *and* a scout

x, did you get this out of the flea market? There are still quite a few fleas.

Here's two to be going on with.

How far do I have to go to find an electronic device without transistors? Not far. A fluorescent tube is an electronic device according to the scientific definition of 'electronic' and yet it does not contain transistors. In fact, that whole first waffly bit makes the reader believe that transistors and semi-conductors are synonymous - which they aren't. What do you mean by electronic?

Conductance. The unit of conductance is the siemens (S) or ohm to the power of -1 (sorry, can't write that properly in a post) or mho. That whole second section fails to support the point it is trying to make and is a load of carp.

There's more yet but I'll save it for later.

?



Amy the (physicist) Ant

I'll add 'facts check' to the list of things to check before pulling something out of PR.

Spelugx the (GCSE physicist) something (suggestions for what I am on a post card please to the regular forum)

There are some even bigger fleas further down the article. Do yourself a favour, pull this out of PR and put it out of its misery.

I'm a bit reluctant to do that immediately, as Awu seemed to like it, but I will say:

TO ALL SCOUTS: DON'T PICK THIS YET, I NEED TO CHECK IT OUT MORE CAREFULLY.

I'll have a through check of it sometime tomorrow evening, when I'll either abandon it or give it a flea bath. (Hey I was only trying help add articles to PR)

Wouldn't it be better to pull it out of PR using the 'remove' link? Then you work on it in you own space - I'll come and help - without any danger of it being whisked out of your editorial control.

Awu can argue his corner if he likes but I suspect he was looking aspects of the work other than its scientific accuracy.

Amy, your concern is noted, but don't worry about it being whisked out of PR yet, with your comments and my replies I wouldn't get past the towers, so stop fretting.

Hi Again,

Amy's right about conductance. I missed that. (I may have skimmed over it too quickly.) It's the reciprocal of resistance. (Handy - you can add them when they're in parallel). I think you meant Greek letter Rho, which, if I remember, is the resistance of a unit cube of material (resistivity?).

The parts about valency and doping are OK. ( I don't know about the lattice though.)

There are more than two types of transistor. Going back in time there was silicon alloy, epitaxial, etc. But maybe you didn't mean how they were made.

I think you have the Zener diode backwards. Every diode has a voltage threshold when forward biased. It's called Vf and is about 0.7V. The Zener is just arranged to have a much higher Vf for, as you say, regulating a voltage.

I have to admit that I only became an authority on testing transistors (H-parameters, switching times, break down voltages, latching, etc.) and not on making them. I later drifted into software so I've been out of touch for years.

How about a mention of the collector, emitter and base and, for the FET, the drain and gate, etc. Just to give the "nubins" names?


I still don't think it would hurt, or add too many lines, to explain that you can either turn them completely on or off (a switch), or use a small current to control a large one (amplifier).

Awu
P.S. I think Bossel is an electronics engineer. He'll put us straight

I definitely meant mho. It's ohm backwards .

Well, that's most of the fleas identified.

Amy,

I wasn't disputing your mho or Siemens.

I thought, because the author was listing ohms/cm, he really meant resistivity (ohms/cm3), and I think the symbol for that is Rho. This was re-inforced by the fact that he was listing it by material. A certain component might have a resistance of 1000 ohms, but you can't say the material wood has a resistance of 1000 ohms, or a conductance of 0.001 mho. You have to say how many ohms per cubic meter or whatever.
(I think we're probably saying the same thing )

Awu

This is good stuff here. I have seen approved guide entries a whole lot worse, not like I'm an old salt or anything. Your description of a FET is so clear, can you ocme up with somehtign as good for bipolar transistor?

MD

Resistivity has units of ?m (that is, ohm-metre. I hope that comes out right). The resistivity of silicon is 2300 ?m. I can't remember any others offhand.An insulator would have a resistivity of around 10 to the power of 14. The fives in the example given in the article are fabricated.

Oh dear, the letter omega doesn't come out right.

May I?

Good entry... BUT

Some minor nits:

mica has a RESISTANCE of... (I think that was mentioned before, the footnote is correct, though)

'hotter normal materials': Conductivity drops with temperature for metals, metals are not normal materials. The passage should read:
'instead ofdecreasing as happens in most 'normal' conducting materials'. But that's just me being pedantic, once.

I have some more problems: Silicon has a valency of 4. The layperson might not understand what you mean with that, maybe replacing the term 'valency' by 'electrons in its outer shell'. Pure silicon forms a perfect lattice: Same here. Maybe you could use the term 'structure' instead of lattice. But then again this (i.e. the valency and the lattice) does not explain semi-conductors. For that you would need to introduce the band theory. It could go like this: - Silicon has 4 electrons in it's outer shell, but to form it's structure the electrons are entirely used up to form bonds. Thus, no electrons remain free to conduct charge, like in metals. -


The next passage of yours goes like: "This also explains the increase in conductivity with heat" No. It is not THIS (the doping) that explains the better conductivity at higher temperature. Again it's the bands. For a layperson you could formulate it like this: At higher temperatures eventually one electron can escape the bond and it is then free to conduct electrical charge.

And: "... in semiconductors the benefits of having more charge carriers knocked out of the electron shells of atoms far outweigh the problems caused by the nuclei of the atoms themselves vibrating more and getting in the way." This is not understandable by a lay person. Maybe you should just skip this part altogether and leave space for an entry that explains conductivity in more detail.

There would be more comments from my part, but I'll leave it for later.

Again. The entry is well-written and the nits mentioned above should be corrected accordingly.

Thanks,

HELL

Spelugx...

Any chance of updating the entry? I would have recommended it except for the crowd of experts clamoring for changes...

Well I'm a complete non-expert but would also clamour for changes!

Whatever the merits of the entry, it is not one 'for beginners'. It started to lose me early on when it mentioned about an MP3 player having millions of transistors, without even saying what a transistor is and where all these millions of them are located in my MP3 player. And then it starts talking about something called 'conductance' - what the hell is that? - as a beginner might well ask. Didn't even mention the old joke about a resistance of two ohms being equivalent to, er, half a mho... (though I did enjoy the doping joke).

I wouldn't know about such things, but this strikes me as one of those entries suitable only for people who already know the subject so don't need to read about it here, as if someone wrote an essay merely to prove to the teacher that they know what the teacher already knows.

I would love to have an entry in the Guide on Semiconductors for Beginners, and I just think you need to focus more on who you are writing this for. I would also be much more interested to find out what they actually do than to get involved in the abstruse technicalities of how they do it.

Does anyone fancy chasing up the author? It's almost there - just needs him to do a bit of tinkering in response to all the good comments above.

A bit of tinkering? It'll need more than that, I'm afraid.