input-referred port noise

[Jack]

Hi,
I am doing spectre noise analysis and I am expecting the input-referred
noise power of a power to be just 4kTR, where R is the port impedance.
However, I keep getting kTR, which is 1/4 of what I expect. Can anybody
help explain why?
Thanks.

Best,
Jack

 

[Andrew Beckett]

Jack,

On 19 Feb 2006 10:01:21 -0800, "Jack" <jack.peng@gmail.com> wrote:

Hi,
I am doing spectre noise analysis and I am expecting the input-referred
noise power of a power to be just 4kTR, where R is the port impedance.
However, I keep getting kTR, which is 1/4 of what I expect. Can anybody
help explain why?
Thanks.

Best,
Jack

The input referred noise is referred back to the source within the port - and so
it might be something to do with the voltage gain factor of 2 on that internal
voltage source - that gain factor is there to ensure that if the port drives a
matched load, it give the correct signal level at the output of the port. If
driving a high impedance load, it will generate double the signal level.

If you have access to http://sourcelink.cadence.com you may want to look
at solution 11025268 which gives a worked example of nise analysis.

Regards,

Andrew.