Noise figure in quadrature path receivers

[Kamran1992]

Hi everyone
I'm deriving the NF of a quadrature RF front-end. We all know how to calculate the input-referred noise in a single path receiver chain. Because it has one input and one output. But what about a quadrature front-end?
I just found this thread but there wasn't a good answer to it:
https://www.edaboard.com/showthread.php?287497-noise-figure-definition-for-IQ
Thank you all

 

[vfone]

I think that in the mentioned link you have a lot of "good answers" so don't need to bother asking for more relating to this subject.

 

[frankrose]

The pessimistic NF is SNR_in/SNR_I=SNR_in/SNR_Q. More accurate when modulation scheme is known and the combined output of the I and Q channels are used, however becasue of non-linear detection it can be very difficult in the design phase to calculate with combined signals of certain modulation schemes. Normally the receiver noise ends at the output of 1st IF filter, and calculated/simulated NF too, so quadrature mixing shouldn't affect the final result. Nobody will be angry for you if you use the pessimistic calculation.

 

[Kamran1992]

Thanks for your beautiful replies

I think that in the mentioned link you have a lot of "good answers" so don't need to bother asking for more relating to this subject.

there are 2 clear answers to the subject. one is to do a BER test and the other suggests using splitter/combiner. both of them are for simulation.

The pessimistic NF is SNR_in/SNR_I=SNR_in/SNR_Q. More accurate when modulation scheme is known and the combined output of the I and Q channels are used, however becasue of non-linear detection it can be very difficult in the design phase to calculate with combined signals of certain modulation schemes. Normally the receiver noise ends at the output of 1st IF filter, and calculated/simulated NF too, so quadrature mixing shouldn't affect the final result. Nobody will be angry for you if you use the pessimistic calculation.

I don't need an accurate calculation of noise in my circuit. I just want to formulate noise sources and NF to get a sense of how everything affects the output noise.



I found a related subject in a book named: "CMOS Cellular Receiver Front-Ends" by "Johan Janssens" Appendix A
He formulated the relation between the SNR of single output(I or Q) and quadrature output for both Zero-IF and Low-IF receivers(My front-end is a low-IF). no filters used so far and no correlation between I/Q mixers:
For a Zero-IF:

But for a Low-IF:

So it means that the noise factor of the whole I/Q front-end is half of what we measure at a single I or Q output!

 

[frankrose]

So it means that the noise factor of the whole I/Q front-end is half of what we measure at a single I or Q output!

No. If the SNR of the I/Q frontend is half of what you measure at a single I or Q output the noise factor will be double because of higher noise.
And missing from above that if you have a low-IF system with image rejection (typically low-IF is used to reject image) the SNR is not half, because image noise is also cancelled by the image filter.

 

[Kamran1992]

No. If the SNR of the I/Q frontend is half of what you measure at a single I or Q output the noise factor will be double because of higher noise.

No, as the formula shows, The SNR of the I/Q frontend is twice what we measure at a single I or Q output. So the nf halves. Please take a look at the reference I mentioned and tell me what do you think.

And missing from above that if you have a low-IF system with image rejection (typically low-IF is used to reject image) the SNR is not half, because image noise is also cancelled by the image filter.

No image filter yet. Let me tell you what I want to do. I have a circuit like I attached(it is simplified) and I wanna find the nf of the circuit. So first I need to find out about how nf in I or Q outputs relate to the nf of quadrature output in a Low-IF front-end without any further IF process and minimum knowledge on modulation scheme (I'm talking about the pure noise itself, not how immune is a modulation to the noise).