Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Receiver's Noise Input =K*T*BWeq ?

  1. Jun 2, 2013 #1

    My transmitter send a signal with BWsignal=17KHz..

    The receiver has an antenna with BWantenna=30MHz...

    I work in a laboratory so T~300k.What is the Input Noice to receiver?

    P=KTBWeq=>-174(dBm/Hz) + 10logBWeq

    BWeq? what is this?

    What i will see in my S.A?it depends from the SPAN,right?
  2. jcsd
  3. Jun 2, 2013 #2
    it should be the the input noise power from the transmitter side multiply by the transmitter's gain + the noise generated from the transmitter itself (due to temperature) multiply by its gain. it should be one or the other.
  4. Jun 3, 2013 #3
    On your spectrum analyzer you will see -174 + 10log(equivalent_noise_BW) + noise_figure_of_analyzer_or_preamp

    Use the noise marker on the spectrum analyzer to obtain equivalent noise bandwidth.
    Noise figure is often specified as DANL (displayed average noise level).
    Last edited: Jun 3, 2013
  5. Jun 3, 2013 #4


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    The operating bandwidth of the antenna doesn't directly contribute to the noise - it will merely affect the received signal level.

    The noise at the receiver is determined by the temperature and the noise figure of the receiver (giving an equivalent temperature) - modified by the receiver bandwidth.
  6. Jun 5, 2013 #5
    Hi and many thanks for the reply! Yes i read about DANL..the equivalent noice bandwith that you are refering is called (maybe) Resolution Bandwidth. The noice figure of the S.A is depend on the device maybe take 20-30 dB values right?
  7. Jun 5, 2013 #6
    I don't understant why the antenna is not contribute to the noice floor of the S.A .

    I have the S.A alone and the noice floor is :-174dBm - 10log(RBW=10^6)+NF(=30dB)===-80dB.

    When i connect only one antenna with some BW in the S.A the noice floor is again the same:O
  8. Jun 5, 2013 #7
    Resolution bandwidth is not the same as equivalent noise bandwidth. Does your spectrum analyzer have a noise marker? (Not all of them do). What model is it?
  9. Jun 5, 2013 #8
    I don't know if it has a noice marker. I will check tomorrow and i will tell you the model exactly!
  10. Jun 7, 2013 #9


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    Thermal noise power in just a matter of the temperature - whether the source is a resistor or an antenna. Having two antennae will increase the signal level because you will (/can) get coherent addition of signals. You may think that noise would add up too but it's an equilibrium issue and temperature is the only thing that counts. Some of the noise power 'generated by' one antenna will go into the other one and no more will 'go into' the receiver. In the same way, the actual noise power is not dependent upon the value of the source resistance. The RMS noise voltage may be different but, when the receiver is designed properly, this doesn't affect the CNR.
  11. Jun 12, 2013 #10

    I think my analyzer don't have noice marker.

    I have HP 8590 Series made in 1991
  12. Jun 12, 2013 #11
    question_zps45a75a76.png "] question_zps45a75a76.png [/URL]
    Hi !

    On the analyzer i think i see the - 174dBm + 10log(RBW) + NF of S.A

    for RBW=1KHz and NF of S.A = 30dB ..Noice Input (analyzer with input at


    I think this is what i see in the display screen. Maye there is some also -( 2-3 dB) of noise that the Sample dETector adds..

    Also i think the real display BW is 1.2*RBW ,because of gauss IF filter's shape
    Is this right?

    So today i did a simple experiment with an LNA . I so that the output noice power was not input +GAIN + NF .... i don't know why this...

    Check the attached please
    Last edited: Jun 12, 2013
  13. Jun 12, 2013 #12
    Hit MKR button, then select MKNOISE ON.
  14. Jun 12, 2013 #13
    Not sure..i will check tomorrow morning. Can you please help with answer?:O
  15. Jun 12, 2013 #14
    This looks correct. My old 8591E has noise figure of 29dB.

    If you don't have a noise marker, then use display averaging.

    They do differ, this is why it is best to use noise marker.

    Not sure how you got -102dBm at output of amplifier, I calculate -109dbm (-111dBm thermal noise from amplifier + 114dBm from SA front end). This would give SNR of 49dB using amp and cascade noise figure of 4.8dB.

    You don't calculate cascade noise figure by subtracting noise figure with and without preamp.
  16. Jun 12, 2013 #15
    Dear friend can you please explain with details your calculations? i don't have a preamp.

    i had signal generator----->S.A and then
    Signal Generator-----LNA---->S.A just this.!
  17. Jun 12, 2013 #16
    First let's figure out exactly what you have.

    LNA is same as preamp.

    Can you provide model number of the LNA? Are you sure it is 3dB NF?

    Where did you get -102dBm, measured?

    The noise you will see on the SA will consist of:

    The SA thermal noise (-114dBm). Same as you would see if you terminated SA input with 50 ohms.


    The noise arriving through the LNA (which will be a "signal" from the SA perspective).
    Its value is -174 +10log(1KHz) + 30 + 3 = -111dBm.

    Convert these to mW, add, then convert back to dBm and you get -109dBm.
  18. Jun 12, 2013 #17
    Dear Emi guy i go for sleep right now (in Greece it is 01.35 the morning). i will come back tomorrow with photos from the S.A and more details. Thanks for all friend!
  19. Jun 13, 2013 #18
    Hi again.

    as you said we must think the system as a cascade.

    ---LNA with NF1,G1 ---- S.A NF2,G2---display screen


    So Noiceout=-174dBm+10logRBW+NFtotal+Gtotal=-174dBm+30+4.8+(gaintotal??)= ?

    This will show as the noice floor when we add an lna ..

    But what is the Gain of S.A?

    Also i was wrong. I Have a HP 8595A and there is NOICE MARKER :biggrin:
  20. Jun 26, 2013 #19
    Sorry, I had not noticed that you responded back.

    Glad you found your noise marker!

    The spectrum analyzer gain is 1 (0dB). What you read on the screen is what it sees on its N-connector. For example, if you enable 20dB of SA internal attenuation, the screen does not drop 20dB. Software in the SA adjusts the display to compensate for any internal attenuation/preamp settings.

    Using 0dB for SA gain:

    So Noiceout=-174dBm+10logRBW+NFtotal+Gtotal=-174dBm+30+4.8+30= -109dBm
  21. Jun 27, 2013 #20
    Νο problem about the delay many thanks for the help!

    I did again the experiment so i notice that: the noise of S.A at RBW=1KHZ and 0 atten is: -113 to -123 not stable. With LNA added goes to -90dBm.

    With noise market without LNA is:-149 to -150 dBm and with LNA goes to -128dBm.

    Now is more clear what is going on ? i don't understand how this -128dBm goes out from noise marker but i think i must use noise marker because the simple market is not stable !!!
  22. Jun 27, 2013 #21
    If you are not using the noise marker you should use RMS detector and turn on display averaging, otherwise the noise floor is quite jumpy.

    Noise marker result is normalized to 1Hz bandwidth (regardless of your resolution bandwidth setting). In other words, do not add the [10log(1KHz) = 30dB] factor when using noise marker.
    Last edited: Jun 27, 2013
  23. Jun 27, 2013 #22
    with noise market -150dBm. how this came from? -174dBm +NFs.a=-150dBm so the NF~24dBm? this is not right i believe. but how can i find the NF of my s.a to do the maths to estimate the NFtotal
  24. Jun 28, 2013 #23
    Why do you think this is not right?
    With your video averaging enabled does the marker go from -150dBm to about -120dBm when you turn noise marker off? (@ 1KHz RBW).

    SA NF does depend on your center freq.

    -150dBm noise marker tells me that your attenuation may be set to 0dB? This does give you better NF for spectrum analyzer alone, but normally you want to use -10dB (this is default) and good low noise preamp in front of it. SA amplitude calibration is done with -10dB attenuation and VSWR of SA input is unspecified with 0dB attenuation and may cause mismatch errors.
  25. Jul 1, 2013 #24
    Hi !!! i am confused because the NF of s.a is different with noise marker on and off?both attenuation 0dBm and 50 ohm terminal at s.a input.

    1)with noise marker on; -150dBm/hz=-174+NFs.a => NFs.a=24dBm ?

    2)with noise marker off: -113 dbm = -174 +10log(RBW=1KHz)=30 +NFs.a=>NFs.a=30dBm...?
    OR -123dbM=-174DBm +30+NFS.A=> NFs.a=20dBm.... ?

    which one is real ?

    i want to make a measure Signal to Noise to the antenna of receiver. The signal is ok.
    About the noise i don't know what value it is...

    Theory says that . The noise of the signal measured in the antenna is
    :-174dBm +10log(BWof signal=20KHZ)=45dBm=-130dBm

    How can i compare the measurements with theory?

  26. Jul 1, 2013 #25
    Use the noise marker, that is what it is there for.

    The reason the SA provides the noise marker is to compensate for various errors which arise when simply viewing the noise floor on the screen (DANL):

    1 - Noise is an RMS measurement. 859X series use envelope detector, not an RMS detector. This causes under response of about 1dB.

    2 - 859X series performs averaging of the output of the log amp. For proper noise measurement averaging should occur before log amp (log of the average is not equal to average of the log).
    This causes under response of about 1.5dB.

    3 - 859X series uses resolution BW filter that is not a brickwall filter. Noise measurements are normalized to a 1Hz brickwall filter. This causes over response of about 0.5dB (brickwall filter captures 0.5dB less noise power than RBW filter of same 3dB bandwidth.)

    Thus we expect a net ~2dB lower DANL vs. noise marker which is what you are seeing.

    Your -113dBm reading (without noise marker) is probably because you have 10dB attenuation turned on. This is the SA default. Make sure attenuation is on manual (not auto). BTW you should be using 10dB attenuation as I mentioned in my previous post, particularly since you have a preamp.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook