Talk:Common collector

From Wikipedia, the free encyclopedia

Contents

[edit] Collector resistors

Common collector amplifier (earlier version is referred to in discussion)
Common collector amplifier (earlier version is referred to in discussion)

Collector resitors are not normally used in CC ccts are they. THis diag needs altering by someone who can do it.Light current 22:59, 6 August 2005 (UTC)

It's a diagram of the most general case. Set Rc to 0 if you don't want it there. - Omegatron 14:21, August 7, 2005 (UTC)
We could put in all sorts of extra components that are not needed and set their impedance to either 0 or infinity as required but I feel that, whilst you and I may be able to understand it, it would confuse the issue for the general reader.

Everything should be as simple as possible ( but no simpler) A. Einstein Light current 23:04, 7 August 2005 (UTC)

Please refer to Horowitz and Hill's The Art of Electronics Ch.2 (transistors)for a good run down on emitter followers. Generally, they do NOT show resistors in the collector and I think we should follow them. Light current 01:19, 8 August 2005 (UTC)
On this occasion, I tend to agree, most emitter follower circuits simply connect the collector to the supply rail, and though in the general case they might have a resistor, in practice they almost never do. I think simpler diagrams are better. The same wpould apply to the common base diagram - it is more obvious to lay it out with the base clamped to 0v and the transistor turned so that the input is on the left, the output on the right. Graham 01:40, 8 August 2005 (UTC)
Yes. The reason I used this layout was to show the similarities between the three configurations. It's not obvious to a newcomer who sees the sideways version that they are all variations on a theme:
I can draw a sideways common base, too, if everyone wants, and we can put them both in the common base article? And I'll delete the resistor and cap from this image if you insist, although if it is ever used in practice, it should really stay. - Omegatron 02:12, August 8, 2005 (UTC)
The diagrams are very nice actually, but I feel that deliberately making them all look the same might tend to obfuscate rather than enlighten, since to the casual observer the key differences are not obvious. Remember that electronics engineers will already know this stuff, so it needs to be aimed at a lay-audience. Personally I'd leave out the bias network, and just illustrate the circuits with the bare minimum of components, which basically means that only the load resistor should be shown. Waveforms to show phase and gain relationships, and perhaps arrows pointing "into" the inputs and outputs giving the impedance (high/low is all that's needed) would also help. This is all I did in my diagrams over at electronic amplifier, though I notice that I've recently come in for some criticism there. My belief is that simpler is best (but no simpler than necessary of course) and while these circuits can be made to look the same, it is probably more helpful if their differences are exaggerated. Graham 02:35, 8 August 2005 (UTC)
Must say I totally agree with everything Graham is saying. Light current 05:26, 8 August 2005 (UTC)
Yes I would like O'tron to do sideways version of CB as this tends to be the way its normally drawn (unless part of a cascode). Can you draw one of those as well while you're at it?? Light current 05:38, 8 August 2005 (UTC)
I disagree with stripping important components off it unless we're showing a small-signal model. - Omegatron 10:57, August 8, 2005 (UTC)
There's nothing wrong with showing several circuits on the page. They're real easy to make. We can show an actual practical circuit, a small-signal model, show various transistor models on the BJT page, etc.
I'm going to redo all the schematics with proper subscripts later today. (I was in a rush when I made these articles.) Is there anything else I should change in these ones while I'm doing it? (The sideways CB will be a different file.) - Omegatron 12:51, August 8, 2005 (UTC)
Compromise suggestion. would it be possible to draw the bias components in 'greyed out' form so that they are visible but not prominent. THese components could then be referred to if necc in the text. Light current 13:57, 8 August 2005 (UTC)
I think that's an excellent idea - I guess we are so used to the limitations of printed paper that the idea of taking advantage of our medium (web pages) to do something different rarely occurs. However I'd also be happy with multiple diagrams - perhaps one which just shows the very bare essentials, and another that is a practical circuit. Graham 22:58, 8 August 2005 (UTC)
Hmm.. Let's think about what we want exactly before I redraw anything. They're relatively easy to draw, but I don't want to waste my time doing something and then changing it later.
In defense of my images, here's the book I learned from (though it's been a while):
Image:Universal BJT amplifier.jpg
350px


Microelectronic Circuits by Adel S Sedra, K C Smith, Third edition. The word "universal" only appears twice in the most recent, fifth edition, and not next to the word amplifier, so they've apparently abandoned this particular teaching method...  :-)
How about I make a version of the universal amplifier and we can mention it, and then change the three current images to represent the most typical cases? Then we can add some other things like small-signal models or your stripped down unbiased versions. Then everyone will be happy. Give me a roadmap of what images you want, though. - Omegatron 23:54, August 8, 2005 (UTC)
Oh! Yes greyed-out components would be great and save some room. Just the bias resistors or the caps too? - Omegatron 23:56, August 8, 2005 (UTC)
Isay just the bias chain. THe capacitors are part of the signal path. But I'm not fussy (who said?) it depends on the accomp text really. :-)Light current 00:07, 9 August 2005 (UTC)
Well we can always change the grayed part. Find some images [1] [2] [3] you'd like to emulate or describe the circuits you'd like to show, with exactly the components you want, and we can decide which ones to gray out later. - Omegatron 00:42, August 9, 2005 (UTC)
I like this one for emitter follower. hyperphysics.phy-astr.gsu.edu/.../ npncc.html

Sorry I do not know how to make this into a link. Will let u know of the others as I find them Light current 00:53, 9 August 2005 (UTC)

THe one youve already got is OK for CE, IMHO Light current 01:01, 9 August 2005 (UTC)
Just copy and paste the URL: http://hyperphysics.phy-astr.gsu.edu/hbase/electronic/ietron/npncc.gif - Omegatron 01:01, August 9, 2005 (UTC)
No decent CB pics yet.But Found not a bad BJT cascode here

http://people.deas.harvard.edu/~jones/es154/lectures/lecture_5/miller/bjt_cascode.jpg

[edit] Added CC image

Added the CC image. Naming convention is the same as the original but removed the collector resistor/capacitor. The image is the one most commonly used to depict CC stages. I hope it is OK for inclusion in the article. Rohitbd 13:02, September 12, 2005 (UTC)

[edit] Simpler circuit and expressions

The previous circuit and the corresponding expressions were too complicated to be useful (IMHO). They're much simpler now without all the biasing details. I do think its important to show a practical circuit though, so I'd like to add another section maybe showing the previous circuit and explaining the biasing (the simple formulas will still be valid once you account for loading). Roger 03:55, 30 May 2007 (UTC)

[edit] How does this table look?

I think the small-signal characteristics should be grouped together in a table. I don't have any experience with wiki-tables, but I just came up with this:

Current gain {A_\mathrm{i}} = {i_\mathrm{out} \over i_\mathrm{in}} = \beta_0 + 1 \approx \beta_0 \quad (\beta_0 \gg 1)
Voltage gain {A_\mathrm{v}} = {v_\mathrm{out} \over v_\mathrm{in}} = {g_m R_\mathrm{E} \over g_m R_\mathrm{E} + 1} \approx 1 \quad (g_m R_\mathrm{E} \gg 1)
Input resistance: r_\mathrm{in} = r_\pi + (\beta_0 + 1) R_\mathrm{E}\ \approx \beta_0 R_\mathrm{E} \quad (g_m R_\mathrm{E} \gg 1 \wedge \beta_0 \gg 1)
Output resistance: r_\mathrm{out} = R_\mathrm{E} \| {r_\pi + R_\mathrm{source} \over \beta_0 + 1} \approx {1 \over g_m} + {R_\mathrm{source} \over \beta_0} \quad (\beta_0 \gg 1 \wedge r_\mathrm{in} \gg R_\mathrm{source})

How does it look (I know its a bit jumbled)? Can someone help me improve it? Roger 03:44, 30 May 2007 (UTC)

[edit] Alternative

How about this one?

Current gain {A_\mathrm{i}} = {i_\mathrm{out} \over i_\mathrm{in}} = \beta_0 + 1 \beta_0 \quad (\beta_0 \gg 1)
Voltage gain {A_\mathrm{v}} = {v_\mathrm{out} \over v_\mathrm{in}} = {g_m R_\mathrm{E} \over g_m R_\mathrm{E} + 1} 1 \quad (g_m R_\mathrm{E} \gg 1)
Input resistance: r_\mathrm{in} = r_\pi + (\beta_0 + 1) R_\mathrm{E}\ \beta_0 R_\mathrm{E} \quad (g_m R_\mathrm{E} \gg 1 \wedge \beta_0 \gg 1)
Output resistance: r_\mathrm{out} = R_\mathrm{E} \| {r_\pi + R_\mathrm{source} \over \beta_0 + 1} {1 \over g_m} + {R_\mathrm{source} \over \beta_0} \quad (\beta_0 \gg 1 \wedge r_\mathrm{in} \gg R_\mathrm{source})

I just put the approximate expression in a different column. I could also put the definition of each function in a separate column (e.g. a column with only Av = Vo/Vi next to "Voltage gain"). Comments? Roger 03:54, 30 May 2007 (UTC)

here is my suggestion (Alessio Damato 08:40, 15 June 2007 (UTC)):
definition exact expression approximated expression conditions
Current gain {A_\mathrm{i}} = {i_\mathrm{out} \over i_\mathrm{in}} {A_\mathrm{i}} = \beta_0 + 1 \ {A_\mathrm{i}} \approx \beta_0 (\beta_0 \gg 1)
Voltage gain {A_\mathrm{v}} = {v_\mathrm{out} \over v_\mathrm{in}} {A_\mathrm{v}} = {g_m R_\mathrm{E} \over g_m R_\mathrm{E} + 1} {A_\mathrm{v}} \approx 1 (g_m R_\mathrm{E} \gg 1)
Input resistance: r_\mathrm{in} = \frac{v_{in}}{i_{in}} r_\mathrm{in} = r_\pi + (\beta_0 + 1) R_\mathrm{E}\ r_\mathrm{in} \approx \beta_0 R_\mathrm{E} (g_m R_\mathrm{E} \gg 1 \wedge \beta_0 \gg 1)
Output resistance: r_\mathrm{out} = \frac{v_{out}}{i_{out}} r_\mathrm{out} = R_\mathrm{E} || \left( {r_\pi + R_\mathrm{source} \over \beta_0 + 1} \right) r_\mathrm{out} \approx {1 \over g_m} + {R_\mathrm{source} \over \beta_0} (\beta_0 \gg 1 \wedge r_\mathrm{in} \gg R_\mathrm{source})
That looks better, thanks. If no one else has any ideas I'll start incorporating that one. Roger 16:41, 18 June 2007 (UTC)
Very good concept. — Omegatron 20:11, 5 August 2007 (UTC)
I think we should include the simpler approximations in the text itself, though, and segregate the more specific models to their own section, as I've tried to do to this one. (Or maybe give each transistor model its own article?) — Omegatron 22:35, 5 August 2007 (UTC)

[edit] Error in expression for Rout

Rout is deficient by the absence of re (the internal resistance of the emitter). This is equal to 25/Ie ohms where Ie is in mA. This should be added to the expression for Rout for correctness. —Preceding unsigned comment added by 88.109.121.209 (talk) 00:27, 7 October 2007 (UTC)

"re" is equal to 1/gm (where gm is IC/VT), so the current expression is indeed correct (look in the " Approximate expression" column). -Roger 06:35, 7 October 2007 (UTC)