One of the best videos on this and much more understandable than what I have seen in textbooks. Thank you! I know that using a cap in parallel with the feedback resistor is common with inverting amplifier circuits, but I have also seen it in non-inverting amplifier circuits. Of course at higher frequencies that cap will only reduce the closed loop gain to unity, until some other pole reduces the gain further at higher frequencies. So does the cap in parallel with the feedback resistor in non-inverting configurations help with closed loop stability?
Yes indeed. For stability, you only look the feedback network. While the overall circuit is a low pass filter, within the feedback loop is a series capacitor which results in a stabilizing zero. Great question. I should have included that in the video.
Wow, this is the video I wish I had when I was building up one of my first opamp circuits, a high gain narrowband filter for receiving ultrasonic signals. The filter worked well but I never fixed the terrible amplifier oscillations, and looking back I think I probably made a lot of design mistakes. Between this, the gyrator video, and all the other videos I haven't watched yet, I think I'll have retained more information about opamps from your videos than I did in college.
Thanks so much for the comment. Before I did the “From Abstraction to Reality” series, I reviewed some open courseware from several colleges. I found they were all about abstraction and didn’t have anything about the pitfalls of op amps. “The perils of resonant decoupling networks” video is an example.
You can reduce the effect of the isolation resistor on the dc precision by using an isolation resistor _inside_ the feedback loop, but you need a larger value to regain stability, of course. The other downside is then a reduction in the ability of the opamp to drive the output close to the rails and a degradation of its ability to supply current to the next stage.
Exactly. That in-loop compensation uses a capacitor from the opamp output to the inverting input, that provides a zero, to cancel the load capacitance pole. Downside is, the load capacitance must be known and constant. I might be the ticket if you are always driving the same type/length coax in a system. Thanks for the comment.
Another great video, sir. As I'm sure you're aware, many guitar effects pedals and guitar amps use high value feedback resistors. (I can understand it in a battery powered pedal, but in a mains powered amplifier?!) These are often used in parallel with a "treble cut" capacitor; does the capacitor help change the phase margin? Also, what software are you using for the calculations and graphs? It looks like it would be incredibly useful for anyone working with operational amplifiers.
Thanks again. Yes, a capacitor in parallel with the feedback resistor, creating a low-pass filter helps the phase margin. We think of it as single-pole low-pass filter, but that's closed loop. Stability analysis is open loop. Open loop, that capacitor makes a zero, which helps to counter the parasitic pole. I should have thought to include that in the video, but then I would need to lump it in with the parasitics, and that would get complicated real quick. Just remember, that's mostly applicable to an inverting amplifier. A non-inverting amplifier with a cap around the feedback resistor helps, but the gain at high frequencies will still be 1 (or at least until the opamp gain quits). I used Excel for the analysis. The applicable Excel workbook is linked in the video description, available from github.
Great video as always! But I have problems to follow the mathematics. What is the "s" in the equations, and how do I use it in practice? Also I saw a lot of polar diagrams with poles and zeroes but I'm unable to extract any information from them. Could you make videos of these topics? This would be great! Greetings from Germany!
Greetings from the USA and thanks for the comment. s is the Laplace complex frequency variable. It’s used when working in the s domain. en.wikipedia.org/wiki/Laplace_transform s = σ + jω Where: σ is a real number than can assumed to be zero for linear time-invariant systems. ω = 2πf j = imaginary operator. I have a video called “Using Excel in The Laplace s-Domain” where I have an overview of the Laplace transform, and use the imaginary number functions in Excel to create magnitude and phase plots, linked below. I hope this helps. ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-demRliYzc9c.html
