Noise figure & Phase noise of transceiver

[chihhao]

In mobile phone design with a fully integrated transceiver,
can I say that the transceiver noise figure is associated with sensitivity &
phase noise is associated with phase error ?
Thanks

 

[willyboy19]

I agree with you that noise figure is largely determined by the sensitivity of the receiver, by the famous formula which relates the sensitivity with -174dBm, noise bandwidth and signal-to-noise ratio.

The phase noise is again, largely related with the frequency synthesizer's phase noise performance. It may also depends on other blocks in the whole transceiver chain where phase error come into play. This is, in fact, a universal problem in all communication channels, including the integrated transceiver.

 

[biff44]

Thermal noise, at -174 dBm/hz, is half phase noise, half amplitude noise. In the time domain it is a random noise vector of changing amplitude and phase. If you think of this small noise vector at the end of a bigger signal vector, you can see that the noise vector makes the composite signal bigger and smaller in length (AM noise), and appear to jitter back and forth in phase (PM).

Phase noise is a little unique, though, in that the electronic components can easily impart phase jitter to a signal that will be far in excess of that predicted by -174 dBm/Hz + noise figure. For example, say you have a microwave combline rod filter in the receiver, and it is physically placed close to a cooling fan. As the fan vibrates the chassis, the rods in the filter move around causing minute phase shifts. These minute phase shifts look like phase noise. Phase noise is also present in any sort of regenerative circuit, like a feedback oscillator, where low level noise is fedback on itself many times--resulting in significant phase noise. 1/f noise in an oscillator is an example of this. Finally, a very common form of phase noise is a microwave oscillator that is phase locked to a lower frequency reference--the microwave oscillators phase noise will be around 20 Log N worse than the reference's, where N is the microwave frequency/reference frequency.

 

[Wilson_yu_chen]

Transceiver is a kind of key factor to determine sensitivity where it should be come
out by another key factor,C/N @baseband. With stringent C/N requirement at baseband, sensitivity would be hard to get perfect number even with very good noise figure number.
Phase error is essentially coming from phase shift and phase noise. It's really important essential spec to pass in GSM system where noise level within TX loop plays a lot role.

bif44,
For thermal noise, I am a little confused about "Thermal noise, at -174 dBm/hz, is half phase noise, half amplitude noise" description. Would you pls clear my mind what's the word "half" playing in your points , Thanks !!

With Regard,

 

[biff44]

Half of the noise POWER is phase noise, the other half of the noise POWER is amplitude noise