How to size the constant-gm transistor if the current(NM2/NM3) is not 1.25mA in order to reduce the total power consumption? How to design the start-up circuit using gm/id methodology? Thanks!
@@ericyeh3787 If I(NM2/NM3)=0.625mA, do you keep gm/id(PM5) constant or keep gm(PM5) constant? 1) if keep gm/id(PM5) constant, say gm/id(PM5)=12 in your video, since id is scaled to 0.625mA instead of 1.25mA, then gm(PM5) is also scaled half to be 0.5*15mA/V=0.75mA/V. as R0≈1/gm(PM5), since gm(PM5) is scaled half, then R0 should be doubled of the initial parameter. 2) if keep gm(PM5) constant, say gm(PM5)=15mA/V, since id is scaled half, then gm/id(PM5)=12*2=24, is 24 too large? 3) How to consider NM2/NM3? Keep their gm/id constant？ Thanks！
Hello, you mentioned the calculation of bias current circuit in the video, can you recommend some information? I couldn't find any explanation in Allen's or lazarvi's textbooks. I'm looking forward to the following updates. Thank you！
Thanks for your support, due to my new career, it’s no easy to make the lecture in the very rarely rest time. However, if you have any interested issues you want to discussion about gm/id , it’s welcome to leave your email here. I’d love to share what I learned.
Eric,thank you very much,and i want to know you will teach analog design again? please put it on youtube!
老師好，在影片19:20左右，Constant GM電路Nmos 為了讓 Rout 變大，所以把nmos 的L變大，但這樣不就會改變W/L的比例?!
這樣難道不會影響電路的運作嗎?! 流過的電流固定，如果L變大， (W/L)變小，所以Vov變大，有可能沒有操作在飽和區嗎
嗯嗯，所以實務上必須確認一下all corner nmos的vds電壓是否有足夠的margin保證落在SAT
這部影片救了我!!
希望有幫助到你
12:13 的output输出部分该如何设置
Please do op amp design in English, it’s really appreciated and thank you.
讲的太好了老师，非常适合现在的我，感谢！同时老师能不能再讲一下其他的模块电路设计，比如带隙参考、二级运放、三级运放等。
你好，请问仿真器的输出部分要如何设置，有代码吗
谢谢老师~！ 老师能不能再录制一些常用模块的影片，比如带隙基准，比如采样电路之类的。（老师讲话好温柔~~~
老師你好，12:38 的地方可以看到amp的增益只有大概10db左右，想請問要怎麼才可以把增益提高呢?
因為我現在參考您的gm/id的方法設計也有遇到增益不足的情況，謝謝!
你可以把L增加，為了維持相同的ID/W，W也等比例增加
How to size the constant-gm transistor if the current(NM2/NM3) is not 1.25mA in order to reduce the total power consumption?
How to design the start-up circuit using gm/id methodology?
Thanks!
Just simply scale your width. Because of the size is calculated from the id/W, it is easy to scale your bias current by adjust transistor width
@@ericyeh3787
If I(NM2/NM3)=0.625mA, do you keep gm/id(PM5) constant or keep gm(PM5) constant?
1) if keep gm/id(PM5) constant, say gm/id(PM5)=12 in your video, since id is scaled to 0.625mA instead of 1.25mA, then gm(PM5) is also scaled half to be 0.5*15mA/V=0.75mA/V.
as R0≈1/gm(PM5), since gm(PM5) is scaled half, then R0 should be doubled of the initial parameter.
2) if keep gm(PM5) constant, say gm(PM5)=15mA/V, since id is scaled half, then gm/id(PM5)=12*2=24, is 24 too large?
3) How to consider NM2/NM3? Keep their gm/id constant？
Thanks！
Hello, you mentioned the calculation of bias current circuit in the video, can you recommend some information? I couldn't find any explanation in Allen's or lazarvi's textbooks. I'm looking forward to the following updates. Thank you！
volthauslab.com/wp-content/uploads/2016/05/225303report55.pdf
"Analog integrated circuit design"
you can refer to page. 307. thanks.
This link is no longer active
How do you determine the input voltage Vcm to be in VDD / 2?
請問在掃id/w-gm/id圖的時候，影片中VDS好像都固定使用0.6V，需要根據每顆電晶體預計的VDS下去改嗎，不知道影響會不會很大?
你可以用不同的vds 去試試看，在較舊的製程裡，不同的vds 並不會影響gm/Id 太多
@@ericyeh3787 瞭解，非常感謝您的回覆
我等待了一年以来，每个月我来你的频道，看看你是否上传了任何视频-
由谷歌翻译（我不懂中文）:)
Thanks for your support, due to my new career, it’s no easy to make the lecture in the very rarely rest time.
However, if you have any interested issues you want to discussion about gm/id , it’s welcome to leave your email here. I’d love to share what I learned.
老师您好，请问哪本书有constant gm bias这个电路的分析啊
www.wiley.com/en-us/Analog+Integrated+Circuit+Design,+2nd+Edition-p-9780470770108
請教老師一下,通常offset小input pair大部份都不會太小,那左側的偏壓電流的size,不就要跟著一起放大,以達到constant gm的function?
我自己是會這樣設計沒錯，不過反過來說，也要先確認系統是否一定需要constant gm?
如果不需要，其實bandgap 搭配V to I的設計其實比較常見
請問老師，為什麽放大管和電流源管的gm/id取一樣大？前面章節有介紹說，考慮噪聲的影響，放大管的gm/id要大于電流源管的gm/id。請問這裏是沒有考慮噪聲的影響嗎？或是別的原因？請老師賜教
因為Bias的部分是使用Constant gm的設計技巧
@@ericyeh3787 感謝解答，期待更新！
您好，請問一下這樣如果我要考量這個OPAMP的SLEW RATE的規格的話我該怎麼著手呢？
因為這樣會有一個要求的gm, 也會有一個要求的slew rate，這樣的話gm/Id就不是我們決定的嗎？
你可以把gm/id展開為 gm/id = (gm/Cx)/(id/Cx) = W0/(0.5*SR)........(1)
W0是GBW, SR是slwe rate ,Cx=Cc(Miller OP), Cx=CL(Folded OP)
根據你要的SR規格和GBW規格，就能夠反推input differential pair的MOS gm/id值了，其實式(1)也暗示著如果要GBW與gm/id成正比而SR與gm/id成反比。
@@ericyeh3787 我沒有想到可以這樣轉換欸，非常感謝您！
你好，我想知道如果要提高直流增益的话，应该要是要怎样的设计思路呢？
你好，可以考慮增加L並且重新掃一遍電晶體的電流密度(id/w)滿足所需的gm/id即可
老师好，我用tsmc的0.18um库进行仿真，当gm/id=12时，nmos的id/w仅有几百mA，请问这是正常的吗？而您的视频中此时的额id/w为几十A。我试图增大nmos的mul值至10时，id/w值也仅能达到几A
老师您好 我想问一下您还有别的ic相关的课程资源吗 想学习学习
老師你好，請問可以拍opamp常見的一些模擬嗎？(像是CMRR、PSRR、ICMR、Slew rate、noise等等之類的)謝謝
相關的模擬我們有設計在C610的lab課程中，你可以參考看看
@@ericyeh3787 老師你好，請問該課程在哪裡可以看到？
@@ericyeh3787 想請問老師在哪裡呢謝謝
能否出一个buffer的设计教程呢