rx band noise
Do someone has stuffs related to this topic?
What about in EDGE where PA's are linear region?
certainly you can, in fact, although the PA is linear, but the noise floor should come into the receiver; you can measure the output power of the LNA when PA is off (no rx signal ) using the PSA, and then turn on the PA and test the output power of the LNA again using the PSA, and you shall find the noise floor is increased!
in fact, this is why TD_SCDMA system require the off tranmit power of the PA!
Hi,
Besides TX noise floor leakage to LNA which is the main contribution, also one may need to check Vcc noise floor (VCO supply...) and nonlinear distortion for mixing product of 13MHz or TX IF or TX LO to TX RF frequency. The former can be suppressed by better bypass capacitors topology and layout, while the later can be relaxed by proper frequency plan.
If it were polar EDGE structure, then Amplitude loop should also be taken care also.
Welcome anyone to add your points or experience.
wilson
But in GSM the TX and RX never work at the same time. I dont think it will not be like a problem.
In CDMA TX to RX isolation is very important.
In GSM, you may interfer some other use.
So this is an item need to test.
Testing tx noise in rx band is to ensure the overall noise or spurs should not desense the neighbouring cell phones. Mainly the overall noise and spurs comes from the TRX and PA. As in GSM, the ASM and FEM should work in isolating the noise to affect the TRX.
A nice method to evaluate the PA noise power in RX band is to do a conversion gain test (simulation and measurement) using a two-tone test. The difference in dB?s between signals at the input of the PA, Fc and F1 is ~ 40dB (for example if Fc = 0dBm F1 = -40dBm)
In the picture that I am attaching from Philips, the conversion gain in RX band is a little exaggerated. Usually the level of 2Fc-F1 is smaller with a few dB?s than F1.