Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.
Hum ,never I did do this simulation but I guess this is not the right way.
I give you an example , how you perform the analysis of gm in case rail to rail input amlifier ?
,because constant gm play a big role in a such amplifier
As zmliu mentioned, that way doesn't work. You can just get one point of gm. Or have you tried it in this way? 'Cause in the Analog Artist, the simulation window won't let you save gm in dc sweeping. Any other thoughts?
As zmliu mentioned, that way doesn't work. You can just get one point of gm. Or have you tried it in this way? 'Cause in the Analog Artist, the simulation window won't let you save gm in dc sweeping. Any other thoughts?
That sure works. You just didn't do it right. "Save all" will cost a lot of disk space though. So the best way is to include the "save" command in a seperate file (if you are running simulation in ADE). In fact this can be done easily from the command line. Use Specrre -h or read cdsdoc for more details. Good luck.
Hello Blue2lucy ,
I will try tomorrow what you said.
If I have well understood ,I need to perform a sweep DC analysis , and saving all.
I use spectre , I will try it tomorrow .
1. specify analysis as OP
2. go to output/setup in ADE and specify ur gm/Id exp using calculator
3. do parametric simulation with one or more variables
what the version of your cadence? i use 5.033 but can not find op analysis.
opamp741 said:
better to do this way..
1. specify analysis as OP
2. go to output/setup in ADE and specify ur gm/Id exp using calculator
3. do parametric simulation with one or more variables
Hi i tried plotting gm/id vs Vov for nmos with 1µ / 20 µ in 0.18 tsmc Vth= 434mV
i found that week inversion according to the plot is somewhere arnd 150mV isnt it too low also cadence shows it to be in cuttoff ( region 0 ) for vgs arnd 200mV ??
I am uploading the plot
thanks
From what I see on the horizontal axis you have Vgs, not Vov. Yes, I also found that transistors go into weak inversion for quite negative Vgs-Vth. While this doesn't sound strange to me for shorter channels, I'm kind of surprised that you got the same result for 20um. But then are you sure that the models in 0.18 are correct for L=20u?
i am using TSMC 0.18 model file and Lmax is restricted to 20u So i guess it should be accurate also i foind that for lower value of L ( 0.22u and 0.8u )it enters week inversion earlier. Is this right?
Good to see that people are still looking at gm/id stuff. A simple way is to plot in hspice. Originally I was trying to do it in Cadence while following Prof. Murmann's gm/id design methodology. You can setup the expression in the artist environment with calculator, and then plot it. The new model is also including a parameter called "gmoverid". But it's more convienent to char your device in Hspice. Check Stanford EE214 (prof. Murmann) Notes.
Try my tutorial, it explain how you can do it in detail
**broken link removed**
edajason said:
Good to see that people are still looking at gm/id stuff. A simple way is to plot in hspice. Originally I was trying to do it in Cadence while following Prof. Murmann's gm/id design methodology. You can setup the expression in the artist environment with calculator, and then plot it. The new model is also including a parameter called "gmoverid". But it's more convienent to char your device in Hspice. Check Stanford EE214 (prof. Murmann) Notes.
Good to see that people are still looking at gm/id stuff. A simple way is to plot in hspice. Originally I was trying to do it in Cadence while following Prof. Murmann's gm/id design methodology. You can setup the expression in the artist environment with calculator, and then plot it. The new model is also including a parameter called "gmoverid". But it's more convienent to char your device in Hspice. Check Stanford EE214 (prof. Murmann) Notes.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
