Video
Đăng bởi Ban biên tập OpenER ·
Tác giả giữ toàn bộ bản quyền (All Rights Reserved) — không phải giấy phép mở. Bạn vẫn xem và học miễn phí trên OpenER; muốn tái sử dụng cần xin phép tác giả.
PROFESSOR
So hello, everyone. I guess we'll get started, since it's Berkeley time. My name is Panos.
I'm a graduate student here, and I'm also a GSI for this class discussion and content. And I'll be doing this review session together with [INAUDIBLE] and Jack.
So the style that we're going to be following is we're going to present a slide deck that's pretty similar to what you saw in mid-term 2, with the addition of some [INAUDIBLE] circuits.
Since there's a lot of info and we want to [INAUDIBLE] through a lot of content today, we're going to go fairly quickly. But feel free to ask questions.
And if there's a lot of questions, Jack and Ricky are going to come to you and answer your questions.
I will also be taking questions at the end of each submodule, so if you are a bit patient, I can answer your questions then.
Also, you can see — you can take a look at the slides and the webcast from this review session later on.
So without further ado, let's get started, because as I said, we have a lot to cover.
I'm going to be going over most of the stuff that have been discussed in this class module 2 for about 40 minutes, and then we're going to go through some [INAUDIBLE] and open problems at the end of the review session.
First of all, we're going to see some passive circuit elements the fundamentals of module 2 — resistors, capacitors, power, identifying node.
Then we're going to go into some circuit analysis [INAUDIBLE] equivalence, and look at the fundamentals of [INAUDIBLE] and where we can use them. So as a reminder, let's go through the first module.
The passive devices that we've seen so far are resistors on capacitors. The main equations that are related to resistors and capacitors are shown tabulated here.
Essentially, the physical equations in the first line — R is resistivity times length over area. Capacitance is the electric permittivity times area over the distance of the two plates.
The I-V relations are Ohm's law and — so V equals IR for the resistors — Q equals CV for capacitors.
Then when you have connected multiple resistors in series, you have an equivalent resistance between the two end terminals that is the sum of all the resistors.
If you have multiple resistors in parallel, 1 over the equivalent resistance is the sum of 1 over each individual resistance.
Also, there is the formula for the parallel combination of resistances — R1 parallel 2 is R1R2 over R1 plus R2 — only two resistors in parallel, in this case.
Now, with capacitors, as you have seen, the situation is kind of reversed. So when you have them connected in parallel, you just add up the capacitance.
Where you have them connected in series, you just do that 1 over C equivalent is the sum of all the — 1 over all the capacitances is connected in series.
We also have this relationship, which comes if you integrate the I equals C dV dT relation of the Capacitor And so for the voltage, V equals I/C times t, this is through only when you are integrating a constant current.
If the kind is not constant, you cannot really do m, and you need to perform the actual integral.
Now, moving on, some very fundamental useful circuits that you've seen over and over in discussion on lecture and in homework problems are the voltage divider, where you have a voltage across a stack of resistors.
And basically, the voltage across each resistor can be found as VR1 is Vs times R1 over the sum of the two resistances.
The VR2 is just Vs times R2 over the sum of the two resistances.
Then moving on, the current divider is the current dual of this circuit, where you have inversed relation.
Now, IR1 is Is times R2 over R1 plus R2. So you use R2 to calculate what IR1. And IR2 is Is times R1 over R1 plus R2.
This is intuitively, because the current follows the path of least resistance.
And you can also analyze the circuit by finding the equivalent resistance, figuring out what the voltage is at the top of this — at the top node, and dividing by the resistance.
Now, moving onto power, for every cell circuit element, if you have some current flowing across the element and some voltage across it, the power that it dissipates or generates is IV.
Now, if we're talking about the resistors, we can substitute the Ohm's law — V equals IR — to get two more expressions.
So the power is I squared times R, or V squared over R. Power in circuits is always conserved, so the sum of all the powers is 0.
And even this sine convention that we're using, if you find a power to be negative, it means that this element is generating power to the circuit, is providing power to the circuit.
If you find the power to be positive, then the circuit element is consuming power. Now, the final fundamental thing that I want to talk about is how we identify nodes.
A node is anywhere in the circuit where I have exactly the same potential. Now, what does not change the potential is wires.
So only when you cross components like resistors or capacitors, you can change potentials. And you can lump nodes together by essentially using some sort of coloring technique.
You start from this point, and you start calling to one direction until you hit this element, the resistor over here. You start then coloring to the other direction until you hit the voltage source.
This is one node. This is exactly the same node. Then you do the same for the other nodes, the other wires, and you keep coloring and find three different nodes in this circuit.
So how can we tell if two components are connected in series or in parallel? For serious connectivity, there's two ways to identify this.
First, they should have exactly the same current flowing through them — exactly the same physical current — not the same value of the current, as you see in the second picture here, but exactly the same physical current, now, the second way of looking at it is that, if two and only two elements
are connected to one node, like in this case, then they are in series.
If there's something else connected to the node, like in this case or in this case, then the current is going to split up and you don't have series connection anymore.
All right, now, regarding parallel, parallel elements are elements that say exactly the same node on both sides — not the same voltage value, but the same physical nodes on both sides.
So in the first two examples that we have over here, you have the same nodes on both sides of each resistor. So they're in parallel.
The capacitors, even though they're drawn a bit funk, they also have the same property. So they're in parallel as well.
But here, we don't have these resistors connected in parallel, since we have another element here, like — [INAUDIBLE] one node, not a pair of nodes.
What we have is that these two guys over here are connected in series, because you see there's only two elements connected to this bottom right node.
So this resistor and this resistor are connected in series. Now, a final quick note on the fundamentals is, what's ground?
Ground is defined as a reference point where we set our voltage to be 0. Now, a good analogy that was presented in lecture was elevation.
Usually, we use 0.0 for elevation at the sea level. So you could be 1 feet above the sea level or 10,000 feet above the bottom of the ocean.
You would be at the same spot, and you need to properly and very specifically define where your reference point is in order for you to be accurate.
Because what matters in elevation, and in circuits and voltage differences as well, is just the difference. So pick a point. Stick to it.
Make sure that you're consistent throughout your circuit analysis about where your reference point is.
We can choose that point to be anywhere. We just need to know where it is, and then we measure everything relative to that point.
So any questions regarding those fundamental stuff before we move on to the circuit analysis parts? Yes? Everything clear?
Okay, great. So let's move on to the circuits analysis part.
Now, the techniques that we've learned so far, in order to analyze circuits, are KVL — Kirchoff's voltage law — KCL — Kirchoff's current law — Ohm's law.
We've Used equivalence, series parallel resistors and capacitors. We've used NVA — the node voltage analysis — superposition, and also, we've seen the Thevenin and Norton equivalents.
That's what we're going to be talking about in this module. Now, a good advice is that, when you're in doubt about the circuit analysis that you're doing, start fresh.
Go back to the fundamentals. Go back to clear algorithms, like the node voltage analysis. They will never fail you.
After you have finished analyzing a circuit, then try to develop your intuition about how and why is this result.
Now, node voltage analysis is the first technique that was presented in this class.
It is an algorithmic procedure that, if you follow with a religious — if you religiously follow the steps, you cannot go wrong.
So first of all, you select a reference node. Let's say that we pick this node in our circuit. Next, you label the nodes with known voltages.
Known voltages are the voltage is set by independent voltage sources — so this top node, right? Then you label the remaining nodes. This is — it's just the intermediate node.
And then you label element voltages and currents following the passive sign convention.
So usually, you would like to pick a direction for the current, and then use passive sign convention to label the voltages across the components.
So you pick I2, you pick I3, and then it's fixed that the plus and minus of V2 are going to be in this direction, and the plus and minus of V3 of — R2 are going to be — is going to be in this direction.
And then you write KCL at its unknown node voltage. So for Vmid we only have one KCL equation here. We just want to find Vmid.
And we say that I2 is equal to I3 because of the direction we picked this voltage — these currents.
Then you write the circuit element equations, which are for I2 for basically resistors R1 and R2, that I2 is V2/R1. I3 is V3/R2.
And you expand this to use node voltages. Yes, question?
STUDENT
[INAUDIBLE] I2 and I3 PROFESSOR: Yes.
STUDENT
[INAUDIBLE] PROFESSOR: That is correct, and I have — we have an example showing that exactly next slide.
Then you write, as I said — like step 6, writing the element current equations, and then substituting back and solving just for the unknown — which in this case, is only Vmid.
So if you do that, you find that you have just a voltage divider. So Vmid is Vs times R2 over R1 plus R2.
Now, back to your question here, does it matter which way I label the currents? And can I pick independently whichever way I like for labeling my current?
So we have two examples where you picked your currents. So you've picked IR1 here in this case, going from left to right, and IR2 going from right to left in the other schematic.
The only important part and the only place where you can go wrong is if you do not follow passive sign convention.
So since you pick IR1 to go from left to right, you need to say that the plus is on the left, the minus is on the right.
Conversely, if you pick it to go from right to left, the plus is on the right, the minus is on the left. Does that make sense?
Then if you write the KCL, you're going to have IR1 equals IR2 in one case, IR1 equals minus IR2 in the other case. So you have a minus there.
However, when you write out your element equations for IR1, you see that you have another negation here. So IR1 is Vvs minus V1 over 4K, and IR1 is V1 minus Vvs over 4K.
These two negations cancel each other out, so when we back substitute into the KCL, you'll see that, by some rearrangement, you can get exactly the same equation. So you'll find exactly the same voltage.
What will be different is the sign of IR1. IR1 is going to have the same magnitude in both cases, but opposite signs.
That's okay, because it indicates that the real direction of the current is opposite to what you have picked.
Now, some other ways of simplifying circuit analysis is looking for a parallel series equivalence. So here, whatever is marked in blue, we lump them together.
Both voltage sources connected in series are just the sum of two voltage sources.
Two current sources in parallel — you can say that, if the currents are looking opposite directions, you can lump them in one and subtract the currents. Make sure you pick the proper direction.
So I1 goes upwards. I2 goes downwards. If I pick my lump current source to go upwards, it should be I1 minus I2.
Then you look for series and parallel resistances, and also you can convert between Thevenin and Norton equivalents to make circuit simpler.
So for example, in this case, you can simplify — you can convert this part of the circuit to a Norton equivalent, and just get three resistors in parallel, which potentially might be easier to simplify — easier to analyze — sorry.
Then a caveat is that you may need to rewind to a previous state of your circuit after simplification, if you have hidden or lumped during reduction the element that you care about.
But at a certain point, you will have computed a value of voltage or current that's useful for what you're trying to do. Let me give an example.
So in this case, notice that, if you care about these voltage over here between R1 and R2, after simplifying, you can no longer figure out what's that voltage.
But hopefully, if you have solved this circuit, you know what the current is through these resistors, and you also know what the voltage at this point is.
So using Ohm's law, you can go back to your original circuit and calculate — using Ohm's law and the current that you found, calculate what the voltage that you were interested in the first place.
Now, let's move on to the other technique that we've seen, superposition.
Basically, since the passive components that we've been looking at are linear in their I-V relationships, you can solve for its independent source in the circuit individually, and then add up the results and get the same net effect that you would get with NVA
or some other technique. So you can use this when you have a bunch of sources and you're confused.
You just need to nullify or take out of your circuit one at a time the other sources, and make sure that you add up the result to get the net effect.
An important thing to note is that this analysis is going to give you exactly the same result as NVA, so you can pick your favorite.
You can go doing NVA or doing superposition, depending on what suits you best. For example, I prefer NVA. Now, let's see an example of superposition, and why does it work.
Let's say we have this complicated circuit, where we have both a voltage on a current source at the same time.
Now, if we write using NVA [INAUDIBLE] KCL equations on u1 and u2, we end up with these two equations.
We can lumped them together in a system — in a linear system that is very similar to what we have been solving in module 1.
Notice that the left-hand side, the matrix here, depends on how the network is defined, how the network is set up of resistors.
And the right-hand side, your B, essentially — your B vector — is only dependent on the independent current and voltage sources.
So you could break the circuit in two and say that, in the beginning, I open the current source and I look at what I have here.
So I set up again this system of equations. And u11 and u21 are going to give me the effect of the voltage source on u1 and u2. And then look only at the current.
[INAUDIBLE] the voltage source, and I look at the solution u12, u22, which are the effect of the current source on my nodes u1 and u2.
At the end of the day, u1 is going to be u11 plus u12, and u2 is going to be u21 plus u22.
You can, essentially, add your solutions together, and if you multiply them with the same matrix, you're going to get the original B vector.
Any questions so far about the circuit — the two circuit techniques that we talked about? Yeah?
STUDENT
On the slide [INAUDIBLE] PROFESSOR: You mean this slide? Yes?
STUDENT
Can you explain [INAUDIBLE] PROFESSOR: Sure.
STUDENT
[INAUDIBLE] PROFESSOR: Okay, great. Actually, this is going to be explained in the equivalents, but let's see about it in the this case.
So perhaps, I should go to the docucam and write out what we have.
So this R1, R2, and R3. These are points A and B, and this is V1.
Let's say that I want to construct a Norton equivalent for this box — so looking to the left of these two points, A and B. Then what these are the short circuit current?
This is A. This is B. [INAUDIBLE] circuit, you just — V1 over R1.
And this is your I Norton. Does this make sense?
So this is equivalent to — right?
[INAUDIBLE] Thevenin resistance, which is equal to the Norton resistance, is just equal to R1, right because basically, you just short circuit the current source.
Does that make sense to everyone?
Okay. I assume yes. So let's go to equivalents, actually, and repeat this point once more. Why do we care?
Why do we use this technique?
Let's say that we have a very complicated network of resistors with some current and voltage source, and we want to simplify to just one voltage source or current source and one resistor.
This can help us design circuits and make our analysis easier, when we're splitting up our circuits in different blocks.
So perhaps, you're designing another block, and you want to simplify everything that comes before that block with just the Thevenin/Norton equivalent — just one voltage source or current source, and a resistor.
That's where you would use this technique.
Now, the way that we can find Thevenin/Norton equivalents are, essentially, you care about two points — A and B — you open them, and [INAUDIBLE] open circuit is V Thevenin.
You can find that V Thevenin using either superposition or node voltage analysis.
Similarly, you can find a Norton by shorting A and B, and looking at the short circuit current that flows through A and B. That's your I Norton.
Then there are a bunch of ways that you can find R Thevenin and R Norton. You can either divide V Thevenin with I Norton and find these two resistances.
R Thevenin is equal to R Norton — let me clarify that. Or you can turn off all independent sources and find the equivalent resistor using some of the techniques that we have seen previously.
You should keep in mind that this technique fails, when you have a dependent voltage or current sources in your circuit.
And when you have such a thing, you should, essentially, apply a test source and measure what is V test and what's I test, and in that way, determine what's your I Norton.
Let's see in more detail how these are defined. So essentially, let's say that we have a black box, and nodes A and B coming out of this black box.
If we model it as a Thevenin equivalent, we will have just one vote that source and the resistor, R Thevenin.
The open circuit, in this case — since there is no current, meaning no voltage drop across this resistor — is going to be just Voc.
Now, when we actually short the nodes A and B, there's some current flowing through the resistor R Thevenin, which is defined by Ohm's law.
And it's V Thevenin over I short circuit — or the V open circuit over I short circuit. Similarly, for the case of Norton equivalents, we have a short circuit current.
We have basically a current source in parallel with a resistor. If we short A and B, all of the current is going to flow through that short.
Nothing is going to flow through the register, and we're going to have that I short circuit is equal to I Norton.
When we open those, there's going to be a voltage drop across A and B, since some current is going to — all of the current, actually — I Norton is going to flow through resistor Rn, and that V open circuit is going to be Rn times In.
In other words, Rn is going to V open circuit over I short circuit — which means that R Thevenin and R Norton are the same.
Now, some things to remember, when applying this technique is that you should replace voltage sources with a short circuit — a wire — replace current sources with an open circuit — so just nothing — and that dependent sources cannot really be turned off, but you need to figure it out if the thing they
depend on is 0 or not before you turn them on — you turn them off.
So what I would say is, look at what sets the current or the voltage of our controlled current or voltage source, and if that element is 0, then you zero it out.
If it's non-zero, you find what it is and you find what the effect of the controlled source is on the circuit.
Final note is that the Thevenin/Norton equivalent of a network preserves the behavior outside the box. So the question previously, we saw that the two parallel resistors remained there in the circuit.
We just simplified or changed the form of the circuit to the left of A and B over here.
So this portion of the circuit remained the same, and this portion was converted to a Norton equivalent.
Some notes about dependent sources — you can have voltage-controlled voltage sources, which means that you have a voltage across two terminals.
And then the voltage across two other terminals is A times V. A has no units here. You have voltage-controlled current sources, which are important, because they are used to model transistors.
You will see more about this in classes like 105 and 140. Here, you have a voltage across two terminals Vi.
And a current flowing in another branch is going to be g times Vi, where g is conductance. It has units of amperes over volts, or [INAUDIBLE]..
Then current-controlled voltage sources — and you have a current controlling the voltage between two other nodes. So Ij [INAUDIBLE] times Ij.
R here has units of resistance, ohms.
Then finally, you have current-controlled current sources, where you have a current flowing in one branch, and that controls with some constant A [INAUDIBLE] current flowing in another branch.
So you have a aIj flowing in another branch. Now, these can be treated as regular sources, when finding the Thevenin and Norton equivalents, as we said.
And they cannot be replaced by open [INAUDIBLE] when you're calculating R Thevenin or R Norton.
Any questions before — about this before I go into the comparator?
Yeah?
STUDENT
[INAUDIBLE] PROFESSOR: So usually, what you do is just apply at test source at the edge of the two points that you care about, A and B. You assume that you know what V test is,
and you're trying to find V test over I test. So basically, you do your analysis — you do NVA with a V test that's known, and then you find I test.
So NVA should give you all the currents. You'll find I test. You'll find V test over I test.
STUDENT
[INAUDIBLE] you apply a voltage source to the two ends [INAUDIBLE] PROFESSOR: Yes, exactly. And it's going to be a function of V test and element components.
So let's say that I wanted to expand this example.
[INAUDIBLE] V1, R1.
I have something like this, which is a times — let's call this node u1 — a times u1.
This is your ground, R1 — sorry — R2 and R3.
And you want to find this equivalent now. So you apply a test — V test.
And you're trying to define what is this current I test as a function of element components.
So I test is going to be a function of V test.
And R1, R2, R3, R Thevenin is equal to V test over I test.
I can do the analysis after the review session and upload this as well, together with the slides, so that you can get [INAUDIBLE] example. Does it make sense?
Any questions so far? Yeah?
STUDENT
[INAUDIBLE] PROFESSOR: No.
[INAUDIBLE] let's call this V element — so basically, if we call this I element or something, this is I test.
Are you happy now?
STUDENT
[INAUDIBLE] PROFESSOR: I test is going to be equal to minus I element. It's not opposite passive sign convention.
So do I have any more space here? Okay, so I'll do this little aside.
Remember that you might have something like this — R, Vs. Your current is going to flow like this.
It's going opposite to the passive sign convention of the source, but it doesn't mean anything. You can pick also Is going that way — Ivs.
And then Ivs is negative IR.
Now, if there are no more — no questions on this, I'll move on to the comparator. You've seen comparators also in mid-term 2.
Essentially, they're blocks that are — they have very similar schematic to the open, but are — there are some intricacies and some differences between opens and comparators.
A comparator is a circuit with three ports.
You have a V plus, V minus, and a V out — V comparison — which is equal to VDD V plus is bigger than V minus, and VSS otherwise.
Let's analyze a comparator. You should remember that I plus an I minus are always 0, and that there's no current.
In this case, what do we have here? What is V1 going to be? Can someone point out the answer? Since there's no current flowing through this node — yes?
STUDENT
[INAUDIBLE] PROFESSOR: R2 Vin.
STUDENT
[INAUDIBLE] PROFESSOR: Yeah. So essentially, you have a voltage divider.
And if you assume that R1 is equal to R2, then you have that u1 is just 1/2 V. And so when does this comparator trip?
It trips when V becomes bigger than 4 volts. So V1 is going to be 2 volts, in this case.
And after V increases, V1 is going to keep increasing as well, so V plus is going to be bigger than V minus. Does that make sense to everyone?
Okay. Any questions in general, before I move on into the part of op amps? We're good? Great.
Now, op amps are very similar devices. They have three main terminals [INAUDIBLE] so you have U plus, U minus at the input, and you have an output.
Uout is A — gain — times U plus minus U minus. And you also have, as I said, VDD and VSS because you need some power to operate this element.
In negative feedback, you have that U plus is equal to U minus, when av — the voltage gain — goes to infinity, essentially.
And you also have that the current going into the terminals of the op amp is 0.
So the op amp rules that we have in this class — that I plus is equal to I minus is equal to 0 [INAUDIBLE] And V plus is equal to V minus only when you hook this up in negative feedback right, and if the gain is infinite.
There is a simplified NVA that was covered in class for op amps.
The two main steps that you need to go through is A, verify that the circuit is indeed in negative feedback, and B, that the op amp is ideal.
Then you follow your standard NVA procedure, with the following simplification. You write all your KCL equations, but you don't need to write a KCL equation for the output node.
The reason you don't need to write that equation is that you have an extra equation, which is that U plus is equal to U minus.
So you can determine all of the voltages in the circuit apart from the output voltage.
And if you have all the other voltages and all the other currents, you can use Ohm's law and figure out what V out is, when you have hooked up your [INAUDIBLE] negative feedback.
Now, let's look at a very useful negative feedback topology in an op amp — the unity gain buffer.
Essentially, we will go step by step through this analysis.
We have that V out is equal to av times U plus minus U minus — the fundamental equations that describes the operation of the op amp.
Then we see that U plus, in this case, is just Vin, if I hook up my input here at the plus terminal.
And V out is U minus, because I've hooked this up with — in negative feedback without any component in between.
Solving for V out, I get that V out is equal to V times av over 1 plus av. And if the gain goes to infinity, then V out is going to V.
Now, let's see an application of this. An application of this — we have gone through it in discussion — it's when we want to cascade circuits.
So how can we implement this circuit? We can use two different voltage dividers in order to do the two different ratios.
So you can use a 1 kiloohm 1 kiloohm divider to construct Vx, and you can use a [INAUDIBLE] divider to construct the Vy from Vx.
But what happens if you want the full functionality of the circuit? Turns out that you cannot connect the two together because you have loading.
Loading is, essentially, the fact that the second block adds some resistance and changes the division ratio, so it messes up your Vx.
Vx is no longer equal to Vin over 2 in this case. How can you remedy this? Based on what you've seen so far, the solution is to add a unity gain buffer.
So if you add a unity gain buffer — I'm going to throw this [INAUDIBLE] new paper.
So you have V, R1, R2.
This is 2K, 1K, and this is now Vx and Vy.
You have maintained this relationship because the current going into the op amp — I plus is 0. And this is 1 kiloohm 1 kiloohm.
Since this current is 0, you can say that Vx is 1/2 V.
And since this node — V plus is equal to V minus is equal Vx — you can say that Vy is 1/3 Vx, and Vx is 1/2 Vin.
Also, the other advantage is that, essentially, the output voltage no longer depends on the node, since the op amp is probably capable of providing as much current as you need it to provide, in an ideal case.
In a real case, you need to consider how big of a load you connect there in order to avoid burning out — burn your op amp.
Now, we're going to wrap it up with some more useful op amp topologies.
I really suggest that you add those topologies in your cheat sheet so that you can have them for any design problems that you're asked.
The inverting amplifier topology, where you have connected your input to the minus terminal — you have Rs and Rf. Then v out is minus Rf over Rs times v.
If you have a reference voltage, you have this term over here as well.
But the non-inverting amplifier topology is exactly the same, where you have hooked up Vin in an opposite way than V ref.
So V out is v in times 1 plus R top over R bottom times V. And notice that R top is essentially Rf, and R bottom is essentially Rs.
It's the same exact topology. We just hooked up Vin, and we [INAUDIBLE] the other way around.
Does everyone see this? Any question about it? Yeah?
STUDENT
[INAUDIBLE] PROFESSOR: So if you want to implement a certain ratio between the input and output voltage, you can do both negative ratios and positive ratios of your choice, depending on what resistors you choose to have there.
You can have ratios that are lower than 1 or bigger than 1. It's up to you. You can pick your resistors and create that ratio.
That's the main difference with a voltage divider, where you could only divide down.
And finally, the transresistance amplifier — if your input signal is a current, because no current is flowing through the minus terminal of transresistance amplifier, all of this current is going to flow through the resistor here to the output.
So because you have negative feedback, you have that the positive terminal is V ref. So V minus is also V ref.
And from Ohm's law, you have v out minus V ref is equal to minus i in times R. Perhaps I can draw the circuit analysis for this one.
So this is V ref, this is i in, and this is [INAUDIBLE]
Since you have negative feedback, and ideal op amp, you will have that V minus is equal to V plus is equal to V ref.
So this terminal here is at V ref.
Also, I minus equals 0, so from [INAUDIBLE] let me move here — sorry — KCl on V minus, you get that [INAUDIBLE] IR —
IR is equal to i in, and Ohm's law — VR is equal to IR times R.
So you have that V ref minus v out is equal to IR times R, so v out is equal to V ref minus IR times R. Is this clear?
Okay. So unless there are any questions, it's practice time. We plan to go over two or three different problems, depending on how much time we have.
Does anyone have any questions so far? Yeah?
STUDENT
[INAUDIBLE] PROFESSOR: Why would you choose to use the transresistance amplifier is the question.
So the answer to that is that, in many applications, your input signal is going to be a current.
And if you want to convert that current in a voltage to subsequently process it, you need to use this type of amplifier. Any other questions?
Yes?
STUDENT
[INAUDIBLE] PROFESSOR: What's the load?
STUDENT
[INAUDIBLE] PROFESSOR: Load is usually a resistor or an element that you want to provide some current to.
So in this case, usually, you're going to see just resistive loads. You could have a capacitive loads. It's essentially a termination.
So when you have your v out, that v out is intended to be connected to a certain load — either a resistor or a capacitor — and provide some current to that, in order for it to do something useful.
A speaker on your phone is just, perhaps, some 8-ohm, 4-ohm resistor. That's the load of the resistor over there. Does that make sense?
STUDENT
[INAUDIBLE] that previous problem [INAUDIBLE] PROFESSOR: So what's load in this case?
So in this case, this part loads this other part.
The second divider loads the first divider, because we add it to this Vx node. So because we directly add it there, we say that it loads that node.
It changes the effective resistance [INAUDIBLE] from that load to ground, and hence, it changes the divider ratios.
So yeah, essentially, you draw some more current through this branch. And because you draw some more current, you mess up this division ratio. Does that make sense?
I wouldn't get too much hung up on the definition of loading.
It's not that important. It's more important to understand the analysis techniques. This is just showing you why an op amp is useful.
An op amp does away with this effect, because it's isolating the two blocks.
The load does not — does no longer affect the previous part of the circuit. [INAUDIBLE] so focus on the properties of ideal op amps connected in feedback.
Focus on NVA. Focus on superposition, stuff like that.
Are there any more questions? You guys are clear? Okay.
With that, I'll pass it on to Jack, who's going to talk about — or Ricky — yeah, Ricky first, who's going to do [INAUDIBLE] problem [INAUDIBLE] problem that you guys saw in discussion yesterday.
Thank you.
[APPLAUSE] TA 1: Hello.
Is it working?
Is it working? Okay. Okay, guys, I'm going to go over the spring 2017 final.
I forget which problem it was, but it was on the discussion worksheet yesterday.
This problem is about we're trying to reconstruct the sigmoid function, which is a nonlinear function that's described by this equation.
And so graphically, it sort of looks like this. So when the input voltage is less than 1 volt, it's negative 1.
And then when the input voltage is between negative 1 and 1, it's a linear function. And then when it's greater than 1, the output is just 1 volt.
And so we're going to see how we can implement this function using switch capacitors circuit.
So this problem tests the concepts of charge sharing, and also about op amp — ideal op amp circuit analysis.
So we're given this circuit, and so part 1 is we would like to draw an equivalent circuit in phase 1.
So if we examine the circuit — we pretend that all the phase 1 switches are closed,
and the phase 2 switches are open.
Okay, so that's what the circuit looks like in phase 1. Yeah?
PROFESSOR
[INAUDIBLE] TA 1: Oh, okay. How do I fix that?
PROFESSOR
[INAUDIBLE] TA 1: Use a different color?
PROFESSOR
[INAUDIBLE] TA 1: Green pen?
Oh, okay.
Okay, I'll start over.
Can you turn the screensaver off?
PROFESSOR
[INAUDIBLE] TA 1: Okay, so I redrew through the circuit in phase 1,
and as always, with circuit analysis, we should label the polarity of your element. So it doesn't matter which polarity you assign your elements.
You just have to keep the signs the same throughout your circuit analysis. So I'm going to call this plus minus VC1 and plus minus VC2 there.
And so that's part 1.
So then part 2 is drawing it in phase 2.
And you have to keep in mind that — to keep the polarity the same.
Okay, now we'll start to analyze it.
The charge on a capacitor is defined as Q equals CV.
and so this V, I like to denoted by subscript VC to be more specific that it's the voltage across the capacitor and also this Q — notice that the Q is positive.
So this is the charge on the positive side of the capacitor.
So if we have a capacitor like this, and you label it plus minus here, this charge is the charge on the positive plate. That's why the charge is positive.
And because we're in this class that we're analyzing steady state, and so the charge on the opposite plate will always be negative Q.
So that's an important thing to keep track of.
And so if we look at phase 1, we have the charge on capacitor 1 in phase 1
is equal to C1 times Vc 1 and we notice that VC1 is exactly equal to the V in.
So this is equal to C1 times V in.
And so our charge on capacitor 2 is equal to C2 times VC2 — oh, sorry — this should be phase 1.
And we notice that both terminals of C2 are connected by a wire, so there's no voltage across C2.
So this is equal to 0.
STUDENT
[INAUDIBLE] TA 1: Lower? Oh, you mean like this? Oh, yeah.
Oh, man.
Okay, so that's the charge of phase 1. And then when we go to phase 2, we have to — in order to solve for the charges, we have to use charge conservation.
So charge conservation — you can only use it at nodes that are not connected to ground or any voltage sources.
So this can only be used at nodes that are not
to ground or voltage source.
As shorthand, we call this a floating node.
So charge conservation, intuitively, if we look at phase 2, the charge — we notice that the charge at the minus terminal of the op amp has nowhere to go.
So if it tries to go left, it will get stuck on this plate of the capacitor. And remember, the plates of the capacitors are insulators, so it can't go.
It can't escape to the left, and can't go into the op amp, because there's no current going there — and also can't go — escape through the capacitor C2.
So it's stuck. So whatever charged was at this node in phase 1 has to be the same as the total charge at this node in phase 2.
So we have this equation. We have to charge — the total charge.
So now, we're looking at this node [INAUDIBLE] total charge in phase 1 has to equal the total charge in phase 2.
[INAUDIBLE] So now, we go look at phase 1 and we see that the node in phase 1 is — right now, it's split between here and here.
So in phase 2, these nodes get connected together, so we have to find the total charge at these two nodes.
And so we calculated that Q. We had C1 V in at this plate, which means that we have negative C1 V in at this plate.
And then over here, we had 0 charge, so then adding it up, we get — we have negative C1 V in.
And then our total charge in phase 2 would be equal to C — we have to be careful here.
So we're looking at the negative plate, because I defined VC1 to be plus minus.
So the charge here is negative C1VC1 plus C2VC2.
And now, we need to find our voltages, VC1 and VC2.
So we look at the circuit again, and this op amp — we're treating it as ideal, and also has infinite gain [INAUDIBLE] negative feedback.
There's a path from the output to the negative input, so therefore, we can say that the voltages at the two terminals are equal.
And so the voltage at the positive terminal is 0, so therefore, we have 0 volts at this node.
And so then we see that VC1 on the left is 0 — on the right is 0, so it's 0.
There's 0 charge on the first capacitor, and then the second capacitor, it's — we have V out here, but then we —
and the negative polarity to V out. So then we have negative C2 V out.
PROFESSOR
Are there any questions about the process?
TA 1: Yeah?
STUDENT
[INAUDIBLE] TA 1: Yeah.
So this note is V out. So then VC2 is equal to 0 volts minus V out.
So we finally finished the equations, putting the total charge from phase 1 and phase 2 together.
And so we have negative C1 V in is equal to negative C2 V out. And we get V out is equal to C1/C2 V in.
So now, we're given another circuit that's sort of similar, and it's claimed that this circuit can create negative output voltage.
Well, it can have a negative gain. So now, we'll analyze this circuit. So I'll draw the circuit in phase 1.
Man, I can't erase.
PROFESSOR
Do you need more paper? TA 1: No, there's a lot over here.
Okay, I think I did it right that time.
Yeah?
STUDENT
[INAUDIBLE] having the wire [INAUDIBLE] TA 1: This one?
STUDENT
[INAUDIBLE] TA 1: You mean this?
STUDENT
Yeah [INAUDIBLE] TA 1: So that resets the charge on C2. So it discharges C2 during phase 1.
All right, so now I'll draw the circuit in phase 2.
Okay, so now we can begin analyzing it. So again, I'll label the polarity on the capacitors.
And so now, we find the charges on each capacitor in phase 1.
So charge on the first capacitor is this, which is V in.
And the charge on capacitor 2 is 0, and the charge on capacitor 3 is also 0.
And in phase 2, we have the charge on capacitor 1 again.
So this node is also at 0 volts, so the charge on capacitor 1 is 0.
The charge capacitor 2 is C2 V out, which is negative C2 — oh, sorry — VC2, which is negative C2 V out.
And then the charge on capacitor 3 is C3VC3, which is equal to C3 V in.
So now, we use charge conservation.
We look at the floating node, and we find the total charge in phase 1 is equal to — so we have negative C1 times VC1 here, and no charge on C3
and no charge on C2. So the total charge at the floating node is negative C1 V in.
And the total charge in phase 2 is equal to the positive — is equal to Q2, which is negative C2 V out.
Let's see. And then the negative Q3 — so it's minus C3 V in.
So we equate them, and then we can solve for V out.
PROFESSOR
[INAUDIBLE] TA 1: Oh, yeah. Okay.
Yeah. So then we get this as our final answer. Were there any questions? Keeping track of the signs is the tricky part.
STUDENT
[INAUDIBLE] TA 1: Which part specifically are you referring to?
STUDENT
[INAUDIBLE] can you pull back the diagram?
[INAUDIBLE] voltage on the left side [INAUDIBLE]
TA 1: Sorry, I'm not quite sure what you're referring to.
PROFESSOR
In this case, [INAUDIBLE] because he
[INAUDIBLE] plus on the left and minus on the right, he's — when he's saying Q1 equals C1VC1, he means that, on the left plate, you have a positive charge of [INAUDIBLE] Could
you explain more [INAUDIBLE] STUDENT: [INAUDIBLE] TA 1: So you're saying it's V in here and 0 here?
STUDENT
[INAUDIBLE] TA 1: Yes.
PROFESSOR
[INAUDIBLE] equal to C times V [INAUDIBLE]..
So plus [INAUDIBLE] the positive side [INAUDIBLE] on the negative side. The positive is the side that has bigger potential. [INAUDIBLE] positive, so it's bigger than 0.
[INAUDIBLE] so on the the top, you have plus [INAUDIBLE].. On the bottom, you have minus [INAUDIBLE]..
Does that make sense?
That's how the [INAUDIBLE] STUDENT: [INAUDIBLE] PROFESSOR: Yes.
[INAUDIBLE] is only applied [INAUDIBLE] nodes that are not connected to any sources [INAUDIBLE] STUDENT: [INAUDIBLE] PROFESSOR: So the floating node becomes floating on phase 2,
because [INAUDIBLE] some charge on it, and then you [INAUDIBLE] so that charge has nowhere to go.
You look at phase 3, the right plate of C1 does not have a path to discharge. So that net charge on the right plate of C1 is going to stay there.
STUDENT
Does that mean that, in phase 2, all of the capacitors [INAUDIBLE] PROFESSOR: Yeah, exactly.
STUDENT
[INAUDIBLE] TA 1: Yeah.
STUDENT
[INAUDIBLE] calculated the [INAUDIBLE] during phase 2? TA 1: Yeah, yeah.
STUDENT
[INAUDIBLE] TA 1: Yeah. So we notice, in phase 1, our output is directly connected to the input — the negative terminal.
And so is this topology is like the buffer. But then our positive terminal's grounded, so V out is 0.
So there's one last part, but I won't go over it, because we're running out of time.
The solution is combining the circuit from the first circuit I showed and the second circuit.
Combine them together, and then you can form a more general circuit that has an output of this form.
So now, I'll let Jack go over another problem.
PROFESSOR
We can over time [INAUDIBLE] TA 2: Also, do people have a preference of another charge sharing problem?
There's also another more resistor, capacitor physics problem. People have a preference for another charge sharing one? Charge sharing? And physics of resistors, capacitors?
Okay, that was pretty split.
One second. I think I'll just go over the charge sharing one, because I think that one won out a little more.
So the setup of this question is we have our capacitive touch screen that we did in lab, and so now we want to add — I don't know — the iPhone pressure sensor so it actually can detect how much you're actually pushing the screen.
So how we're going to do this is we're going to have these capacitors. And between plates 2 and 3 here, there's actually a distance, and you can actually compress that distance.
And so you're decreasing the distance between — of a single parallel plate capacitor, and that's affecting its capacitance.
And so for the sake of time, I'll go through and try to quickly go — do the problems.
So parts A and B are going through another charge sharing problem.
And so we have phase 1 here, which is plus minus — and this is 5 volts, and this is Cf.
So this is one, two [INAUDIBLE] and this is grounded.
Two, this one is floating.
And three, it's grounded again.
And so part A is asking to find a solution such that, in phase 2, V touch is equal to 2.5.
And so the first thing maybe to notice is that in phase 2, we actually have tried sharing, because notice V touch here — even though we're measuring that voltage, it's not actually connected to anything, so it's not drawing any current.
And so this node up here is a floating node or a floating wire, and so charge gets stuck on here in phase 2.
Then notice, in phase 3, it's once again connected down, and so it's discharged in phase 3.
And so going through that, we're going to look at the charges, and we're also going to label our capacitors.
For this one, the labels turn out a bit nicer, if you just keep plus minus on top and bottom.
Đăng nhập để đánh giá học liệu này.
Đăng nhập để bình luận — chưa ai bình luận về học liệu này.
Không có khóa học liên quan.