Apparently transistor Q15 in the following circuit is in common collector mode (well, actually it says it's an emitter follower but my understanding is that this is the same thing as common collector mode). This isn't obvious to me at all, although my textbook offhandedly says this as if it should be obvious. How can you tell that it's in common collector mode?

- 13,593
- 1
- 17
- 48

- 353
- 1
- 9
-
2This circuit element also seems to provide current limiting by means of the 50 ohm resistor and Q22. Current more than about 120 mA will cause Q22 to switch ON and limit drive to Q15. – PStechPaul Apr 13 '23 at 23:31
-
1Urthona, The fact that you were told Q15 is an emitter follower, combined with the apparent resulting comments here that generally agree, points up a serious problem with such descriptions. In fact, Q15's emitter is being held at a fairly fixed voltage since Q22 is measuring the voltage across the 50 Ohm resistor and ensuring that it stays at about 1 vbe above the negative rail. So the base of Q19 will likewise be about 2 vbes above the negative rail, holding Q15's emitter. Q15/Q22/Q19 make up the VAS, which is effectively emitter-grounded in this topology. So not so obvious it seems. +1 – periblepsis Apr 14 '23 at 02:09
-
1Here you will find a dip analysis of a internal circuit of a 741 https://cdn.evilmadscientist.com/KitInstrux/741/741_principles_Rev21.pdf – G36 Apr 14 '23 at 14:16
6 Answers
It's not immediately obvious, but Q15 and Q19 form a darlington pair. Replacing all the gumph around them with basic equivalent elements, you are left with (on the left):
simulate this circuit – Schematic created using CircuitLab
Everything in the blue box is a darlington pair emitter follower, which is sometimes called "common collector". In this case, the collector doesn't look very common to anything, especially considering its collector wobbles up and down, and is actually the output, just prior to buffering by the push-pull output stage. I understand your confusion.
For me, the big giveaway was the low value of R1, 50Ω, indicating that its function is to modulate current through R1, and consequently throughout the entire vertical path via I1, V1 and Q19. The voltage across R1 (and therefore also the current through it) varies in proportion to input potential at node A, a classic "voltage-controlled current sink" architecture, employing an emitter follower.
Since emitter followers are also called "common collector", that's probably why your professor named it such.
I suppose R2 could be confusing too. Its role here is simply to improve Q19's turn-off speed, by providing a path for Q19's stored base and capacitive charge to evacuate, when Q15 turns off, which would leave Q19's base floating.
The other potentially confusing element is Q22, shown above right. This seems to be a simple current limiter. When 12mA or more flows down through R1, it develops 0.6V, enough \$V_{BE}\$ to switch on Q22, and prevent A from rising any further in potential.

- 27,759
- 2
- 16
- 87
-
3
-
There is something else confusing here - two current "sources" (the current source Q13 and the current sink Q19) are connected in series. Will there not be a conflict between them? Or, maybe, this was the Widlar's goal... a "dynamic load" configuration? – Circuit fantasist Apr 14 '23 at 07:27
-
1@Circuitfantasist That's an interesting point. If these two competing current sources were ideal, there would indeed be an unresolvable conflict, but here they are constrained to have compliance voltages of \$V_{CC}\$ (for the top source) and \$V_{EE}\$ (for the lower sink). So, in practice, one or the other pulls harder (up or down), and wins the fight. For some value of \$V_A\$ the two currents will be very similar, and the "output" swings freely between the extremes of \$V_{CC}\$ and \$V_{EE}\$... – Simon Fitch Apr 14 '23 at 07:44
-
1@Circuitfantasist The gain of this stage near this equilibrium point is huge, because small changes in \$V_A\$ produce very large changes in current in R1. Since this current flows through the very large impedance of the upper current source (remember ideal current sources have infinite impedance, this one though is not ideal), it produces large changes in the output potential of that upper source, as it fights to keep current constant. – Simon Fitch Apr 14 '23 at 07:53
-
That is, he made an amplifier stage (Q15 abd Q19) with a dynamic load (Q13)... – Circuit fantasist Apr 14 '23 at 07:59
-
1@Circuitfantasist I guess what I am trying to say is that the top source is constant current, and the bottom sink is variable. We have a tug-of-war between the two, and if these were ideal sources, the universe might implode. But they are not ideal, and their relative impedances are finite, but *very* different, giving rise to extreme output swings as the battle takes place. – Simon Fitch Apr 14 '23 at 08:02
-
1@Circuitfantasist, yeah, the upper source is most definitely a dynamic load! Sorry, I failed to make that connection. Actually, I suppose both sources see the other as a dynamic load. – Simon Fitch Apr 14 '23 at 08:03
-
Great analogy! I think that everything in this world is "tug of war" or "arm wrestling" but the equilibrium is not always reached :-) – Circuit fantasist Apr 14 '23 at 08:16
-
1@Circuitfantasist I think Simon already explain very well why there isn't a "conflict" per se. The question you asked could indeed be also posed for a regular common-source amplifier with a load current source. Imagine there's feedback around it, how do we keep the sources fighting from each other? Feedback works to keep both in saturation, and, typically, the common-source will be "less saturated" than the current source. – Designalog Apr 14 '23 at 10:16
-
1@Circuitfantasist One of the 1st things one of my mentors taught me is that, when designing an input stage (in an IC, that is, but probably applies in discrete as well), you have optimize it and make sure you actually an amplifier rather than a current source. – Designalog Apr 14 '23 at 10:17
-
There is no paradox here because the current sink is modulated by feedback; and for short time scales, the large compensation capacitor (30pF) keeps the voltage well-defined (it changes at, at most, \$\frac{dV}{dt} = \frac{I}{C}\$). This also gives the output a low-ish impedance at high frequencies (via the output voltage follower). Without feedback, indeed it will eventually (within 10s µs) saturate to Vcc/Vee, and you have a poor-performing comparator. – Tim Williams Apr 14 '23 at 11:03
-
Thank you for taking the time to make that diagram. I'm new to all this, so would never have been able to recognise that Q16 with its resistors forms a voltage source and the current mirror is a current source. To put things simply, would you say that the main reason it is a common collector is that the collector terminal is connected directly to a voltage source? – Urthona26 Apr 14 '23 at 11:26
-
I have seen comments on here saying that the reason it is a common collector is that the output is taken at the emitter. The problem I have is I have seen diagrams of common collector transistors where collector is connected to base and emitter, and base and emitter are not connected to each other. It makes sense why the collector is 'common' in those diagrams, but here everything is connected, albeit with other components between the connections. – Urthona26 Apr 14 '23 at 11:27
-
@Urthona26 The "output" of Q19 is clearly taken at the collector, since that's what's driving the next push-pull buffer stage (if we define "output" to be the signal of interest for the end user). I'm happy to say "emitter follower" though, because that's certainly Q15/Q19's role here, in which case the emitter could also be considered an output (of sorts). No argument there. Only in *that* context does the term "common collector" makes sense to me, but personally I wouldn't use the term "common" *anything* in this context, because there's no "common" point I can identify! – Simon Fitch Apr 14 '23 at 12:06
-
@Urthona26 In the sense that I1 and V1 and all those "upper" elements are providing some kind of potential at Q19's collector, that Q19 can draw current from in its role as an emitter follower, I suppose you *could* say that Q19's collector is "common" to that potential. Usually emitter followers *really do* have their collectors pinned to a supply rail, giving rise to "common collector", and even though that isn't the case here, it's still an emitter follower, and I can see why your prof insists on the name "common collector". Can't say I agree, but I do get it. – Simon Fitch Apr 14 '23 at 12:16
Q15 by itself is more or less an emitter follower (common collector), yes.
More precisely, it's in a Darlington pair with Q19, a composite structure, which is itself common-emitter. (Which is also current-limited by Q22.)
Which way is more useful to look at it, depends on what you're after. These circuits are complex enough that an analytical description isn't really feasible, so we tend to use hand-waving descriptions like "common-emitter", and rougher descriptions like, corresponding the input/output voltage/current ranges (discounting nonlinearity of the device(s)), making assumptions about certain devices being "on", "off", in or out of saturation, etc.
I wouldn't think there would be much academic merit in labeling these as CE or CC, but if your instructor is going to stake the success of a test on it, say, I would suggest paying attention to whatever definition they give.

- 22,874
- 1
- 20
- 71
-
5"...if your instructor is going to stake the success of a test on it..." Yup. It'll bump your grade, and give you experience with smiling and nodding at da boss when they've gone around the bend. – TimWescott Apr 14 '23 at 00:41
-
TimWescott, Exactly! "Common-something" is a formal classification and "common-collector" is the most meaningless of the three. In this case, even the collector is not common... so nothing is "common" here... – Circuit fantasist Apr 14 '23 at 06:04
Q15 is in an emitter follower, or common collector circuit.
The signal comes in via base and is output from emitter, and the collector is supplied with constant current from a current source circuit.

- 127,425
- 3
- 97
- 261
-
But isn't that exactly the same as common emitter? Common emitter signal goes into the base and out the emitter with DC going into the collector, just like in common collector. – Urthona26 Apr 13 '23 at 23:09
-
@Urthona26 You got it wrong. Common emitter configuration has the signal going into the base and appearing at the collector, while emitter is either connected directly to the supply line or its AC signal is grounded via capacitor to the supply line. Common collector circuit has collector connected directly to the supply line while the output voltage appears on the emitter and following the base voltage with a difference of about 0.6-0.7V (that's why it's also called emitter-follower). – Edin Fifić Apr 13 '23 at 23:24
-
Do you mean that the output signal is taken at the emitter in a common collector? I understand that. Is it true that current flow is from base to emitter for both of them? – Urthona26 Apr 13 '23 at 23:57
-
@Urthona26 Yes, if there's current in the opposite direction the transistor isn't going to survive much of that. – Hearth Apr 14 '23 at 00:53
One way to tell which transistor electrode is the common one is by looking at whether it's connected to a voltage rail or current supply line OR if the other two electrodes serve as signal input and output.
The Q15 in this case has the signal coming into its base electrode and going out of its emitter into the base of Q19 (darlington connection), which would, by the process of elimination, make its collector the common electrode.

- 5,188
- 7
- 24
-
1The collector is connected to the supply rail but through the "soft" current source, not directly ("firmly"). In addition, it is the output terminal of the transistor. Then what is the point of "common collector"? – Circuit fantasist Apr 14 '23 at 06:24
It is difficult to label Q15. An emitter follower has a voltage output at its emitter. In a Darlington configuration which Q15,Q19 are in, the emitter supplies current to the base of the second transistor. There is no signal output from the Darlington emitters to the load. The 50K resistor is only there to discharge Q19 bas-emitter capacitor. The 50 ohm resistor is used to sense current for current limiting and so is not used as a signal output.
So the emitters are essentially grounded. The Darlington pair is used as a common emitter gain stage.
Since the emitter of Q15 is not used as a voltage output but the collector is, then Q15 is not configured as a common collector amplifier, but is more like a common-emitter configuration.
Like Tim Williams....Call it the way your instructor says.

- 12,496
- 2
- 9
- 34
I've had another look through my textbook, having considered some of the answers here. It seems that they consistently use 'common collector' to simply mean that the input is at the base and the output is at the emitter. Similarly, they use 'common emitter' to mean the input is at the base and the output is at the collector, and 'common base' to mean the input is at the emitter and the output is at the collector. So, that is how I am going to think of it from now on. I'm no longer going to worry about why, if it's common collector, there is a path between base and emitter even though they aren't connected at all in the very basic diagrams of the common collector, or think that the collector is common to both the 'input circuit' and the 'output circuit' and try to identify where the 'input circuit' and 'output circuit' is (try doing that for the diagram I posted - very difficult indeed!)
Thank you all for your help.

- 353
- 1
- 9