The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Simulators >> RF Simulators >> flicker noise with switched current source
https://designers-guide.org/forum/YaBB.pl?num=1196265720

Message started by aaron_do on Nov 28^th, 2007, 8:02am

Title: flicker noise with switched current source
Post by aaron_do on Nov 28^th, 2007, 8:02am

Hi all,

there is flicker noise reduction technique which involves using a switched current source. Just wondering, is this taken care of when simulating with spectre and the BSIM3 model?

thanks,
Aaron

Title: Re: flicker noise with switched current source
Post by ACWWong on Nov 28^th, 2007, 2:13pm

If the BSIM3 model card is coded with flicker noise, then spectre and spectreRF will calculate it (in noise or pnoise or transient noise).
So the impact of flicker noise reduction techniques should be seen in simulation. Spectre/spectreRF allows you to see how any individual device noise sources like fn contribute to the output noise (i.e. "noise summary").
In pnoise with noise soures set to time domain, you print the noise summary at different time points of the pss (shooting) steady state, to see at what parts of the cycle fn of each device is contributing.

Title: Re: flicker noise with switched current source
Post by vivkr on Nov 28^th, 2007, 11:05pm

Hi Aaron,

When you say "switched current source", do you mean that the current source is switched ON and OFF or just commutated? The latter would be modelled by spectreRF, but for the former, you would need a change in the model. I don't believe that the former is modelled by any model at present.

I am speaking of the technique presented by Eric Klumpernik et al. where they turned current sources ON and OFF and the reduction in 1/f noise exceeds the reduction you would expect based on just the reduction in the duty cycle of the switch waveform turning the current ON and OFF. I don't remember the paper title, but it was in JSSC a few years ago. You can Google Klumpernik.

Regards
Vivek

Title: Re: flicker noise with switched current source
Post by aaron_do on Nov 29^th, 2007, 3:54am

Hi Vivek,

that's exactly what i'm talking about. The current source is switched on and off. So i guess it is not modeled. BTW what do you mean by "just commutated"?

thanks,
Aaron

Title: Re: flicker noise with switched current source
Post by Frank Wiedmann on Nov 29^th, 2007, 4:22am

The article you probably want to read is http://icd.el.utwente.nl/publications/get_file.php?pub_id=299. Currently, the effect is not included in any models.

Title: Re: flicker noise with switched current source
Post by aaron_do on Nov 29^th, 2007, 6:58am

thanks for the help. I was able to get the paper from IEEE xplore

Aaron

Title: Re: flicker noise with switched current source
Post by borisk on Dec 13^th, 2007, 3:37pm

Aaron, if I remeber correctly, the key in that paper was that the source is switched to negative voltages such that the device is taken to accumulation region thereby somehow lowering flicker noise.

If switched on/off (say below/above transistor Vth) then the noise is being averged by the duty cycle of the switching and should be taken into account by Spectre, but going negative wouldn't as it's a physical mechanism that isn't captured by the BSIM3/4.

Cheers,

Boris