For the first time in my life, I am convinced and satisfied to have learnt about the current mirrors, what a fascinating way of explaining and teaching Prof. Razavi. All my due respect to you from India.
absolutely amazing. the way you teach, I just wish all professor in the world do the same. explaining step by step and going down to the student level of thinking. some professors think that if they explain in a very complex way then that means they are very smart. just learn how to teach from this legend.
01:25 - Intro and Review 05:17 - Need for Current Sources 07:55 - Problem of Biasing of Current Sources 15:45 - Current Mirror Concepts 27:33 - Diode-Connected Transistor 41:48 - Current Mirror of Band-Gap Current Reference
This is one of the best lectures in the whole series. After deriving the output current and finding out it is independent of temperature and supply voltage made me smile and appreciate this wonderful idea. Current mirrors are brilliant concepts and it is beautifully explained by the GOAT Razavi.
the best teacher i have seen since the creation of youtube, your videos are remarkable, you are alone in front of the camera with a soothing, slow voice. I really like! Other teachers film their lessons from afar, it's unpleasant and illegible
Wonderful Lecture. I'm so appreciate that you give this good lecture to students for free. Watching your lecture make me get interested at Electronics. Thanks a lot!
Although the currents between the two transistor could be made the same by their constant width-length ratio, i think it should be also emphasized that in order for temperature variation of the process parameters VTH and uCox to cancel, the two transistors should be properly matched in layout (i.e. close to each other, common centroid layout )so that their temperatures are the same and the the process variation of their W/L ratio varies the same so that the change due to temp and process variation, cancels.
Thanks a ton for the lecture series sir. Ever indebted!! Can someone clarify here @ 40:32 , to effectively cancel out, we are assuming Vth1 = Vth2. But in some instances it is considered that Vth is a function of W/L, W mostly. So to generate multiples of Iref, when (W/L)1 != W/L)2 wont that cause a current variation due to Vth variation?
There is a circuitry called "Tie High Cell" in VLSI. They say it is used to provide "always high" to other parts of the circuit. Main goal of this circuit is in-circuit protection from power surges/spikes from mains, and provide constant 1. There are only one pmos and one nmos in Tie High cell and nmos is diode connected. I know how how diode connected nmos works but i dont completelly understands how this circuit provides "Always high" over the source of the pmos.
Professor, where is the book that you would write concerning analog circuits? I have the "Gray and Myer", and have had it for over 30 years, and it is of little practical value.
I have a doubt when you flipped the current source, previously the current source base is connected to the source of the MOSFET which is at ground, and now to Vdd. I did not get this!!!
Hi prof.Razavi. is it needed that Vth be matched in current mirror , in equation they cancel out if they are the same , observation nr3 M1 Vth and M2 Vth are the same matched.
Won't biasing a transistor with current source as in cascode amplifier also block a/c since current source won't allow fluctuations in current? How can varying signal then be amplified? Can someone please explain?
Actually Vth mainly depend on the source body voltage and in our n type mosfet fabrication, in general all the body is made by same substrate , so only headaches of Vs. now the problem is done because of location of current mirroring…which is in general done at bottom of the rail mean at lowest potential like ground. And if you use at upper rail so all p mos source connected to Vdd, which is ac ground. So that’s now no longer issue in general.
For the first time in my life, I am convinced and satisfied to have learnt about the current mirrors, what a fascinating way of explaining and teaching Prof. Razavi.
All my due respect to you from India.
The best teacher that I've ever seen. You are really a respectable teacher.
Do you know anyone like him for communication course?
do you know anyone like this for communication course?
The way you teach us walking from baby steps is truly unparallel !
absolutely amazing. the way you teach, I just wish all professor in the world do the same. explaining step by step and going down to the student level of thinking.
some professors think that if they explain in a very complex way then that means they are very smart. just learn how to teach from this legend.
01:25 - Intro and Review
05:17 - Need for Current Sources
07:55 - Problem of Biasing of Current Sources
15:45 - Current Mirror Concepts
27:33 - Diode-Connected Transistor
41:48 - Current Mirror of Band-Gap Current Reference
really really appreciate it!!
This is one of the best lectures in the whole series. After deriving the output current and finding out it is independent of temperature and supply voltage made me smile and appreciate this wonderful idea. Current mirrors are brilliant concepts and it is beautifully explained by the GOAT Razavi.
the best teacher i have seen since the creation of youtube, your videos are remarkable, you are alone in front of the camera with a soothing, slow voice. I really like! Other teachers film their lessons from afar, it's unpleasant and illegible
Wonderful Lecture. I'm so appreciate that you give this good lecture to students for free. Watching your lecture make me get interested at Electronics. Thanks a lot!
your book plus your videos are the best teaching devices! thank you!
Sir u remind me of Albert Einstein...🙏and ur method of teaching is soo simple...
what a fascinating way of explaining and teaching Prof. Razavi.
Like here, who are following playlist in 2024
Do not have words to express my feelings.
Best electronic teacher I have ever met(6 in total)
Can you name others so that I can refer those lectures as well.
This lecture was a beautiful adventure 😊 ❤️
You made me madly love this course!!
So Awesome and interesting thought process about the evolution till Current Mirror. Beautiful, Thank you So Much Prof Razavi Sir..
give that man a nobel!!!
pause at 13:00 sharp and you can see a hyena.
when you deep understand some thing you can explain it it to a baby and that what you have done sir ,thank you
greetings
I want to say thanks for ur all of the videos
Thanks for uploading and sharing. Excellent lectures well explained.
love you ... you are great... salute to you...
Best explanation
Although the currents between the two transistor could be made the same by their constant width-length ratio, i think it should be also emphasized that in order for temperature variation of the process parameters VTH and uCox to cancel, the two transistors should be properly matched in layout (i.e. close to each other, common centroid layout )so that their temperatures are the same and the the process variation of their W/L ratio varies the same so that the change due to temp and process variation, cancels.
thanks, was wondering about this.
Amazing Teacher and great Professor.....
Blessed with these lectures:)
Amazing lecture, Thank you sir
Got it lecture -5. Thanks a lot sir.
Thanks for this wonderful lecture.
God level teaching 🙏
Thanks a ton for the lecture series sir. Ever indebted!!
Can someone clarify here @ 40:32 , to effectively cancel out, we are assuming Vth1 = Vth2. But in some instances it is considered that Vth is a function of W/L, W mostly. So to generate multiples of Iref, when (W/L)1 != W/L)2 wont that cause a current variation due to Vth variation?
There is a circuitry called "Tie High Cell" in VLSI. They say it is used to provide "always high" to other parts of the circuit. Main goal of this circuit is in-circuit protection from power surges/spikes from mains, and provide constant 1. There are only one pmos and one nmos in Tie High cell and nmos is diode connected. I know how how diode connected nmos works but i dont completelly understands how this circuit provides "Always high" over the source of the pmos.
Professor, where is the book that you would write concerning analog circuits? I have the "Gray and Myer", and have had it for over 30 years, and it is of little practical value.
Name of the advanced analog course that sir told in lecture ?
I have a doubt when you flipped the current source, previously the current source base is connected to the source of the MOSFET which is at ground, and now to Vdd. I did not get this!!!
put the speed on 1.25. veryy good lecture !
Can a mos produce voltage when excited by a current source?It is a unilateral device!Please correct if i am wrong.
Hi prof.Razavi. is it needed that Vth be matched in current mirror , in equation they cancel out if they are the same , observation nr3 M1 Vth and M2 Vth are the same matched.
Are there exam samples for this material? Please upload your previous exam files or give us a link to previous exams
Won't biasing a transistor with current source as in cascode amplifier also block a/c since current source won't allow fluctuations in current? How can varying signal then be amplified?
Can someone please explain?
we can connect The other terminal of Iref to either Vdd or Ground right?
여윽시 라자비형님 최고십니다
Annaa, devr chanag ittirli ninna🙏
38:05 the circuit doesn't bite, you can play around😂
I want to ask that whether the Vth1 and Vth2 (in current mirror ) is the same all the time.
Actually Vth mainly depend on the source body voltage and in our n type mosfet fabrication, in general all the body is made by same substrate , so only headaches of Vs. now the problem is done because of location of current mirroring…which is in general done at bottom of the rail mean at lowest potential like ground. And if you use at upper rail so all p mos source connected to Vdd, which is ac ground. So that’s now no longer issue in general.
love you
where i can find the slides for these lectures ??
some one answer me plz
did u find them?
Thanks
thanks
take a moment to digest this
The dislikes are from my professors
23:28
1.5x is best