Radio frequency amplifier and error rate

[thavamaran]

Hi, I am designing an RF link. Before my spectrum analyzer, I have an RF low noise amp with a NF of 2.5 dB. Without this amplifier, my error vector magnitude or error rate is 2.1%, after an amplifier it goes down to 0.7%.

My question is, the receiving signal will have its own SNR and with that SNR I achieve 2.1%, after amplification of the signal power which is relative to SNR, cause the signal will have the same amount of signal power and noise power and amplification will linearly amplify both of these factor.

With taking this into consideration, how can the error rate improve cause the amplification is with respect to signal power and noise power? I can't figure this out, please advice.

 

[yungman]

Simple! Without the amp, even though your SNR is better, but the amplitude is too low, the reciever front end is not sensitive enough to capture all the data.

Think of it this way, if you are driving on a remote road farther and farther from the radio station and you listen to the radio. As you drive farther away, the sound start to break up because the signal is getting weaker and weaker. The detector starting to to fail because the signal level is too low. But if you boost the signal from the antenna, even though you add noise, but the signal is stronger. The extra noise in the signal do not hurt because you can at least detect the signal.

 

[fleem]

yungman's right, but I'll throw my wording in, also...

Each amplifier adds some noise. The noise contributed by a given amplifier is, of course, amplified by the following amplification stages. So if you have two amplifiers and one adds less noise than the other, then you should place the lower-noise amplifier earlier in the sequence. If you placed the noisier amplifier first, then its noise gets amplified by the following stage, and you end up with more noise. So swapping the order of the amplifiers doesn't change the total gain of the signal, but it certainly changes the total noise. So you can also add a lower-noise amplifier to the front end of an existing device and increase the final SNR for the same reasons.

 

[yungman]

Agree.