noise floor and LNA added gain
What is correct:
1. On spectrum analyzer marker read out is -85dBm, due to the cable loss ref level offset (lets assume loss is 2dB) will lead that value to -85+2=-83dBm, but if LNA is added in the path loss should this be added as follows or substracted?
-85+(Gain) or -85+(-Gain)-can anyone comment?
Any application note on this?
Thanks
The LNA will reduce the noise floor by its gain bud it will add its amplified noise. You should consider LNA bandwidth to estimate the noise-floor increase above that of your spectrum analyzer minus LNA gain.
What an application note is for? Use the Friis' formula and noise power calculation, e.g. Pn =-174 + NF + 10 log BW (Hz) + G, to determine the LNA output noise. The spectrum analyzer usually has a poor NF, so the presented noise floor will be higher than the calculated Pn.
In my opinion LNA decreases the noise figure not noise floor see Friis noise figure formula. But the LNA reduces the overall noise figure of the system by providing gain.
Use the freeware program AppCAD, where you can set the number of stages, and play with characteristics as Noise Figure, BW, Gain, Noise Floor, etc.
AppCAD
(go to Signals-Systems / NoiseCalc)
After that you can do the math using Friis formula to see if matches with program results.