The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Simulators >> Circuit Simulators >> how to sim the OTA
https://designers-guide.org/forum/YaBB.pl?num=1271318229

Message started by linkernzhang on Apr 15^th, 2010, 12:57am

Title: how to sim the OTA
Post by linkernzhang on Apr 15^th, 2010, 12:57am

hi,all
I'm designing an OTA. I have referred some simulation method, I still can't figure out how to sim the OTA to get the curve like this .

Title: Re: how to sim the OTA
Post by linkernzhang on Apr 15^th, 2010, 1:01am

So far I simulated ,I only get the curve like diagonal. Can anyone tell me how the set the testbench?

Title: Re: how to sim the OTA
Post by Andrew Beckett on Apr 15^th, 2010, 1:49am

I don't mean to be rude, but do you think anyone is likely to be able to understand what you're asking, based on a low resolution jpeg (I can't even read what the expression names are), with no information at all as to what you've referred to or what you think you're trying to simulate?

Please remember that others cannot see inside your head, and so you need to furnish the reader with as much information as possible in order to elucidate your question and increase the chances of somebody being able to answer.

Maybe somebody can guess what you want - I can't.

And by the way, this would have been better in the Circuit Simulation board rather than Analog Design (Ken will probably move it).

Regards,

Andrew.

Title: Re: how to sim the OTA
Post by aaron_do on Apr 15^th, 2010, 3:46am

Hi,

First of all I think Andrew is right in that question and the figure are unclear and the post will probably be moved anyway.

That said, it looks like a PSS analysis with the input signal Amplitude swept from -1V to +1V. The graph appears to be a plot of the transconductance since you say it is an OTA. I assume the plot is trying to illustrate gain compression...it also seems like you have plotted multiple biasing points.

Plotting from -ve to +ve amplitude looks interesting but I don't think there is any difference to just plotting one side (0 to +1 for example).

cheers,
Aaron

Title: Re: how to sim the OTA
Post by linkernzhang on Apr 15^th, 2010, 7:25pm

Hi,
I'm sorry I haven't describe the question clearly,and sorry for post it in a wrong place .
you know ,the OTA is convert the voltage into current. So I think the V-I relationship is a diagonal .and the slope keeps mainly constant in some range ,eg if Vinput range from -100mv ~ 100mv . So I get the V-I curve as follows .

Title: Re: how to sim the OTA
Post by linkernzhang on Apr 15^th, 2010, 7:30pm

But I have referred many books/papers ,their gm curve are usually like this.as follows . I have calculated .the slope the curve above can be transfer into this . But I don't know how to transfer it in ADE. The above the current output(positive)- output(negative). I think ( output(positive)- output(negative))/vinput ,I can get the curve below .But the result is not .Can anyone help me ?
I don't know weather I have describe the question clear or not.

Title: Re: how to sim the OTA
Post by ayounis on Apr 16^th, 2010, 12:45am

What you are plotting is I vs. V. I think what you want is gm vs. V (which is an important characteristic of an ota).
gm vs. V is the derivative of I vs. V.
I believe, if you plot the derivative of the curve you show, you should get what you want!

Ahmed

Title: Re: how to sim the OTA
Post by sheldon on Apr 16^th, 2010, 6:12am

Greetings,

From the plot it looks like the testbench is okay. Just for
review ground the output into a voltage source at the desired
output voltage, Vcommon_mode. Next, plot the derivative of
the current through the voltage source with respect to input
voltage vs. input voltage, gm= delta Iout/delta Vgs. There
might be an inversion required.

Best Regards,

Sheldon

Title: Re: how to sim the OTA
Post by Ken Kundert on Apr 16^th, 2010, 2:44pm

It looks like the transconductance curves for a doublet (the second order form of Gilbert's multi-tanh circuits). Unless there is some switching behavior (CMFB?) there would be no need for PSS analysis. You can get these curves from a simple swept AC analysis, as I describe in my book (Section 3.4.2 starting on page 106 in http://www.designers-guide.org/Books/dg-spice/index.html).

-Ken