Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

Gain vs SNR

S

Squirtle

Junior Member level 2
Junior Member level 2
Joined
Aug 24, 2021
Messages
22
Helped
0
Reputation
0
Reaction score
1
Trophy points
3
Activity points
298
Hello guys,

I don't understand something about SNR. In signal chain, we know that SNR never increases, in every step it slightly decreases and in every component signal passing through, noise floor get closer to the signal. We use amplifier to amplify coming weak signal but noise signal also increases in same amount. So, why do we use amplifier if it does not have positive affect on SNR.
 
Sort by date Sort by votes
You forgot to include noise figure in your amplifier considerations

There are two cases:
1) incoming signal is already noisy, e.g. short wave radio where noise background is already higher than the amplifier's input noise level
2) incoming signal has high SNR, e.g. microwave antenna pointing to the sky, so that we are limited by the noise figure of the amplifier

For case 2, especially the first amplifier's noise figure will make a difference.

For cascading multiple stages, have a look at cascaded noise figure

Cascaded Noise Figure Calculator - everything RF

www.everythingrf.com www.everythingrf.com
 
Upvote 0 Downvote
Hi,

Imagine the tiny energy coming from a turntables coil. Without amplifier it would be impossible to hear anything (useful).
Imagine you feed an very tiny signal to an ADC ... maybe the input signal is below 1 LSB ... so the output of the ADC would be nothing useful.
Imagine you have the output voltage of a thermocouple temperature sensor. It is in the microvolts.What useful thing you could do with this microvolts signal?
Imagine you have the LNA of a sattelite TV disc... it´s in the microvolts .. how do you think you (digitally) can process this tiny signal and visualize it as a movie on a TV screen?

Klaus
 
Upvote 0 Downvote
So, we use amplifier because receiver cannot hear below a threshold voltage which can be defined as minimum detectable signal. If incoming signal is below a threshold, without an amplifier we will not get signal eventhough we have good SNR. Only good SNR is not enough. This is what i understand.
 
Upvote 0 Downvote
Yes.

To process a signal, we need to amplify, filter, demodulate then perhaps decode.

There will be offsets, nonlinear effects or adjacent channel interference of these latter stages that add error to low signal levels. We must amplify to minimize these effects below the noise floor or achieve the desired quality or error rate from a desired SNR. There are quality factors in an LNA that include the specs for the NF, gain, BW and distortion products defined by THD or 3rd order intercepts that are considered.

Generally the higher the frequency, the greater the Friis losses , the lower the signal level and raw SNR and thus the need for a low NF LNA. Then SNR is increased greatly and optimized by matching the channel BW to the signal BW. This filtering may be done at the RF, IF and baseband as needed. This is still part of the signal chain. Each stage may also include filters to gradually reduce the BW with minimal group delay or phase distortion.
 
Last edited:
Upvote 0 Downvote
Squirtle said:
Hello guys,

I don't understand something about SNR. In signal chain, we know that SNR never increases, in every step it slightly decreases and in every component signal passing through, noise floor get closer to the signal. We use amplifier to amplify coming weak signal but noise signal also increases in same amount. So, why do we use amplifier if it does not have positive affect on SNR.
Click to expand...
Your question make sense, somehow.
If the gain of the first amplifier (LNA) is high enough (>15dB) the SNR of the receiver it follow the noise figure of the LNA by 1 to 1 in dB (e.g. SNR=20dB for LNA NF=1dB, and SNR = 19dB for LNA NF=2dB).
But in the same time the total receiver noise figure it will decrease dramatically if a low noise amplifier is used in the first stage of the receiver.

The receiver sensitivity is not only function of the SNR, but also by the total noise figure of the receiver:
Receiver_Sensitivity[dBm] = –174 + 10*LOG(BW[Hz]) + Noise_Figure[dB] + SNR[dB]
 
Upvote 0 Downvote
Note that this Sensitivity equation is the required minimum threshold per Hz BW of the receiver (Rx), not the actual. The -174 dB comes from Boltzmann's constant using 50 Ohms at 25'C per 1Hz BW.

For more details on the entire path loss budget and receiver discriminator path, you must understand your noise sources; Flicker 1/f noise, thermal noise figure, shot noise and other sources of interference. From that one can compute the requirements or determine the resulting SNR or error rate.

Start here.

Rx_Sensitivity[dBm] = –174[dB] + 10*LOG(BW[Hz]) + NF[dB] + SNR[dB]

The output SNR comes from amp. gain and matching BW to the signal spectrum while the input could have more noise than the signal.

It would be best to look up Friis Loss calculators for results and elsewhere for explanations. Find as many as possible to gain SNR of your understanding. ')
 
Upvote 0 Downvote

LaTeX Commands Quick-Menu:

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top