Exercise 10.11 in "Microelectronic Circuits" 2nd Ed.

Fred Bloggs expounded in news: snipped-for-privacy@s9g2000yqm.googlegro ups.com:

Yes, thanks. I did eventually figure it out, but due to the nature of usenet, I couldn't cancel my message. But yes, this is how I computed Veb1.

Yes.

Yes.

This is different from how I computed it. This time I used:

Vbb = Vbe1 + Vbe2

Knowing Vbe3 from Ic = 100mA, I used:

Vbe4 = Vbb - Vbe3

Ic4 = Is * exp(Vbe4)

Thanks for your help on this, Warren

Well, I just dummied in those names. I also assumed that all the transistors are matched, though. So I just ignored the part numbers, too. They were just placeholders. The only thing I guessed about using those part numbers was the beta of 150. The rest I took from the example you provided.

The Art of Electronics is kind of fast paced for hobbyists with zero training. And it skips a LOT of the individual, small steps that a beginner needs to apprehend the broader meaning. But there is a Student Manual for it. And it REALLY HELPED. Any non-pro reading AofE should BUY that manual. It walks you through things that the textbook does not.

I need to work out each and every detail, in order to gradually "get" the big picture view that helps you "go faster" and ignore the less important details. In short, it is nice to first take everything into account so as to see, personally, how some things have little effect and others have larger effect. And also where and where not certain effects become more important or less. If you haven't put your own hand to it and are just reading summaries, you will never really "get it." (Okay, maybe the geniuses do.)

This example you provided, when looking at the 100 ohm loaded case, shows how Q3 is important to know about but that Q4 in that case is almost unimportant. Any professional here sees that in a fraction of a second. Been there, done that, they'd say. But a neophyte needs to see that for themselves. And at first, they will see both Q3 and Q4 and be flummoxed about what Q4 may or may not be doing at the time. And that won't necessarily be obvious. But after struggling through it once or twice and being forced to have to think over and over on your own, it deepens into you and you begin to gain that "eye" that helps reduce the thinking load to where it needs to be.

I liked the example. Even to figuring out Ic for Q4 in the

+10V, 100 ohm loaded case. Turns out that in simulation, which I only attempted AFTER trying to write out the calcs on paper, was pretty darned close with those part numbers. The slight differences came from the _actual_ Vbe of each BJT, which wasn't precisely the same as assumed, the slight differences in simulated kT/q (I used 26mV, but I know it is slightly different than that in simulation), better estimates of beta, and some more complex modeling (Gummel-Poon instead of EM1, Early voltage [VA], etc) for the BJTs that I didn't care worrying over. Nice to see that happen, too.

The basic equation I always keep in mind:

I = Is*(e^[Vbe/(kT/q)]-1)

It's really a diode equation. Since the BJT has two diodes, things are a little more complex, but you can ignore a lot of that in simple circumstances. Just think of that one as talking about the BE diode and ignore the BC one when in usual use. Although it talks about the base current relative to the Vbe voltage, if you assume that beta is constant (it isn't and you do need to keep track of the fact that towards the higher end of Ic the beta goes down), then it also tells you something about Ic, too. If Ic varies by 10, then it is likely Ib also varied by 10 (assuming beta fixed.) That's why.

_Is_ is often down around fA levels, so if _I_ is many orders more, you can ignore that -1 term and drop it for simpler use, which is what I did.

The 'T' term in the above equation isn't enough to help you understand what temperature does. That's because _Is_ is, itself, highly temperature dependent. In fact, enough so that it changes the sign of the derivative. So the main thing is not to look at that equation alone if you care about temperature effects. It's 'mostly useless' there, since _Is_ is a function of T.

Oh, type this into google:

(Boltzmann constant * 300 kelvin) / elementary charge

That's for about 27C. You should see:

0.0258520269 volts

Which is why I assumed 26mV. You can set that up for any temperature. Keep in mind that all this works well for unitless, relative ratios. If you are comparing one Ic with another, for example, then the basic diode equation gives you some idea how the Vbe would change. But you need a starting point and that is device dependent. That's why I used the

0.7V at 2.86mA, since the example you gave provided those. At that point, I took everything relative to that. There, the equations tell you about relative behavior, given some known starting point. But without specific knowledge of some part, you don't get absolute values. And you need a LOT of info on the part for that. Datasheets are there for those reference points which allow you to reasonably interpret from there.

GET THE STUDENT MANUAL!!!! I cannot emphasize that enough. It will help you. A lot.

Jon

Jon Kirwan expounded in news: snipped-for-privacy@4ax.com:

I'm not coming at this with zero training or experience. I have a fairly good feel for how these things work from general pov. I have an advanced ham license and been active in electronics since around 8 to 10 yrs of age. But what I've always lacked is the ability to engineer something like this from scratch (when I was growing up, I had no electronics mentors, no Internet, no money, blah blah). With the Internet and the availability of good texts, I finally decided that there is no reason not to learn this properly now.

What I've liked about the Art of Electronics (from the little I've seen), is the little bits of practical advice. There is some of that in the Microelectronics book as well but less. Those rules of thumb are very helpful.

For me, this was fairly obvious. I knew the vo was going to be near +10V, so Q3 was going to be doing all the work while Q4 was pretty much shut off. For me, the challenge was being able to compute the precise voltages/currents in that scenario. So it was a good exercise to go through. The Microelectronics book is just full of these kinds of exercises at the end of each chapter. So I may do some more of these to burn in that training.

But experience can tell you that as well, without being able to produce the numbers. That was me anyway.

The book did say it was "a long exercise but very instructive". The "very instructive" part sold me on it.

I didn't bother to simulate it at all. I knew I could, but the point for me was the "on paper" part. Now there were hours where I did question the book's answers and was tempted to start up LTspice, but decided to trust the book.

Oh ya, I have that burnt in after going through this book earlier this year. But the book seems to have required the student to figure out that he needs to sometimes derive the Vbe = form, given Ic, which was a big leap for me. After hours of struggling with this, that lesson too, is now burnt in!

If you struggle with something, you'll remember it!

Yes.

The Microelectronics book suggests that you use 26mV if the circuit is in an enclosure (higher temp). But otherwise it seems to be using 25mV. If memory serves, it is more accurately 25.2mv at 25C however.

I bought the equivalent of that for the Microelectronics book. I haven't gone through it yet. One problem was that I couldn't get the book matching the edition I have. But if the problems and the answers are both there, it shouldn't matter much.

Would that Art Student manual be useful if it doesn't match the edition of the book? I plan to spend the big bucks on the

3rd edition someday. For now I went cheap and bought the toilet paper (International) edition.

Warren

The student manual I have in my possession works stand alone, I think. So assuming you get the same one, should be okay.

Jon

Jon Kirwan expounded in news: snipped-for-privacy@4ax.com:

That's good to know, thanks. It'll be next on my used books queue. It's fairly easy to find deals on the Student Manual.

Warren