MIT Electronic Feedback Systems (1985) View the complete course: ocw.mit.edu/RES6-010S13 Instructor: James K. Roberge License: Creative Commons BY-NC-SA More information at ocw.mit.edu/terms More courses at ocw.mit.edu
Not sure if anyone noticed the Neil Young song in the end as an example of distorted music that we cannot hear the difference. You can see his face glowing there for a second!
How can one be certain that audio distortion is the result of a lack of a suitable feedback path and not the result of a spooky ghost haunting? The world may never know.
An extremely crappy presentation... It is not clear why the nonlinear transfer function is straightens up... The guy knows his stuff but unable to explain.. At one point he explains a whole electronic circuit with words rather than showing us a schematic..
what's not clear? he even showed you feedback signal which corrects the nonlinearity. also demo follows good example where slopes are 1 and 0.1 initially (very large difference) and after adding feedback they are 9.9 and 9.1 which is very close. you probably need to watch lecture from beginning.
@@PavelKrupets The problem with his presentation is: the nonlinearity he presents is without dynamics (ie, it is not a differential equation), and it wont work as he presents it as the signals will travel through it with infinite speed. Try it in simulink and you will see. But when you put a very fast pole just before the nonlinearity [ ie, something like 0.0001/(0.0001s+1) ] it works marvelously. AFAIU the guy simpli assumes this delay and never tells us about his assumption. So it took me a week to wrap my mind around that.. I still maintain that his presentation is crappy. But make no mistake, this is one of the best control theory lecture series around, if one takes his time to decipher it.
@@PavelKrupets I dont think you get what I say.. The the feedback loop around the nonlinearity, as he gives it @14:54, wont work. A delay is needed. If you dont believe me just try it on simulink... But if you just put a fast pole, like 0.0001/(0.0001s+1), between the gain block and nonlinearity, it will work.
@@sahhaf1234 you probably should try to understand the math everything works in simulink as well - simulink shows exactly the error he shows on oscilloscope (vertical lines around zero, spikier sine tops) - output is pretty much sine wave feedback / gains all help reduce or attenuate non linearity
