IHP-GmbH / IHP-Open-PDK

130nm BiCMOS Open Source PDK, dedicated for Analog, Mixed Signal and RF Design
Apache License 2.0
410 stars 65 forks source link

Simulation of MOSFET noise with ngspice #207

Open 0y8w1x opened 1 month ago

0y8w1x commented 1 month ago

Environment

Actual Behavior

I tried to do a noise simulation of a single common source transistor, but found the noise to be 0. In case the schematic is incorrect or I use the noise analysis wrong, please let me know. A simple DC analysis works just fine, so the general setup should be working.

My circuit looks like this: Screenshot from 2024-09-22 20-55-08

And when I plot the onoise_spectrum/inoise_spectrum aswell as the total integrated noise I get the following: Screenshot from 2024-09-22 20-56-23

bmurmann commented 1 month ago

You are measuring the output noise at a node that is pinned to 1V, so I am not surprised that the result is zero. What are you trying to do?

0y8w1x commented 1 month ago

You are measuring the output noise at a node that is pinned to 1V, so I am not surprised that the result is zero. What are you trying to do?

Oh yeah that was a stupid mistake. I want to measure the intrinsic noise of the MOSFET

KrzysztofHerman commented 1 month ago

@0y8w1x could you please provide a correct testbench with some results. After that we could close the issue with a working example.

dwarning commented 1 month ago

Maybe these qucs-s/ngspice (NGSPICE_OSDI_DIR env variable have to be set) examples can help to close the issue: Bildschirmfoto vom 2024-10-09 17-43-31 Bildschirmfoto vom 2024-10-09 17-44-03

lv_nmos_NOISE_id.sch.txt lv_nmos_NOISE_vd_vg.sch.txt

KrzysztofHerman commented 1 month ago

@dwarning we provide a sample .spiceinit, which sets the location of the OSDI file so you do not have to reference it on the schematic. The only thing you have to do is to create a symbolic link to this file under your $HOME direcory. I will also add some more controls to setup the noise plot type.

dwarning commented 1 month ago

Sorry, it was not the emphasis of my post. Anyway - thanks for hint. What about induced gate noise. There are data and test benches available?

0y8w1x commented 1 month ago

thanks @dwarning for the example, I reproduced it and it works great. Screenshot from 2024-10-12 17-44-16

dwarning commented 1 month ago

Another test case in a 50 Ohm environment NF, NFMIN and Rgn =f(Freq) Bildschirmfoto vom 2024-10-14 13-56-15

sergeiandreyev commented 3 weeks ago

Hi @0y8w1x, do you have additional questions, or we can close the issue?