Noise power linearity

[abhijitrc]

WHY IT IS A MANDATORY THAT LINEARITY IS REQUIRED FOR NOISE MEASUREMENT !!
If the system is not linear then the measurement of noise is not possible??? Plz do advice me.

 

[jiripolivka]

From noise figure definition, the device under test must be linear to correctly respond to noise power at input to noise power at output. As the typical noise power is measured in pico- and nanowatts, linearity is no problem in low-noise ampliiers. But in complete receivers you must find the point to measure the output noise power before any detector or demodulator in order to correctly measure the output noise power.
Find Agilent white paper "noise figure basics" or the ARRL Radio Amateurs Handbook. Many engineers do not know the basics and fail in noise measurements.

 

[abhijitrc]

From noise figure definition, the device under test must be linear to correctly respond to noise power at input to noise power at output. As the typical noise power is measured in pico- and nanowatts, linearity is no problem in low-noise ampliiers. But in complete receivers you must find the point to measure the output noise power before any detector or demodulator in order to correctly measure the output noise power.
Find Agilent white paper "noise figure basics" or the ARRL Radio Amateurs Handbook. Many engineers do not know the basics and fail in noise measurements.

Sir, what i mean to say is that in 1dB compression point why we can't measure the noise!! , why linearity is important?, can we measure noise( thermal noise) in the nonlinear region of an active multiplier ..., output noise power in nonlinear region is not possible ???

 

[jiripolivka]

Sir, what i mean to say is that in 1dB compression point why we can't measure the noise!! , why linearity is important?, can we measure noise( thermal noise) in the nonlinear region of an active multiplier ..., output noise power in nonlinear region is not possible ???

Your questions demonstrate that you have no knowledge of basics.
No, we cannot measure noise figure of a nonlinear system. I have explained clearly why.
You can measure noise power anywhere but in nonlinear devices there is no clear link to input noise.

 

[abhijitrc]

Your questions demonstrate that you have no knowledge of basics.
No, we cannot measure noise figure of a nonlinear system. I have explained clearly why.
You can measure noise power anywhere but in nonlinear devices there is no clear link to input noise.

I dont think u had explained the matter properly, whatever u had explained earlier that is a bookish explanation, that i know......,

 

[jiripolivka]

I dont think u had explained the matter properly, whatever u had explained earlier that is a bookish explanation, that i know......,

I am sorry but you fail to understand this "bookish" explanation.
You can measure the noise power at any point you choose. But the thermal noise being Gaussian can only be referred to system input if the system response is linear. Otherwise the noise becomes clipped and is no more Gaussian, nor linked to the input noise.

Read some "bookish" explanation first, please. Otherwise you cannot understand important things.
You are not alone, MIcrowaves & RF are full of failures by "experts".

 

[biff44]

if you are testing a non-linear system, especially one that had compression or clipping, that compression or clipping might reduce or accentuate the noise--leading to a false reading of noise figure. The noise may be clipped, or if the gain compresses may appear enhanced.

So, yes, keep things linear somewhat.

 

[mtwieg]

Well if your noise power alone is enough to cause any compression, then you have other huge problems...

 

[abhijitrc]

I am sorry but you fail to understand this "bookish" explanation.
You can measure the noise power at any point you choose. But the thermal noise being Gaussian can only be referred to system input if the system response is linear. Otherwise the noise becomes clipped and is no more Gaussian, nor linked to the input noise.

Read some "bookish" explanation first, please. Otherwise you cannot understand important things.
You are not alone, MIcrowaves & RF are full of failures by "experts".

Now the explanation is proper , the question had been raised by me as i am also thinking that noise may be clipped as we moved beyond 1dB compression point ,but not very sure of my own explanation,but now it is ok.., anyway thanks for this fruitful discussion.

 

[abhijitrc]

Why close in phase noise of LO for digital communication system is more serious concern than an analog communication system ??

 

[jiripolivka]

Why close in phase noise of LO for digital communication system is more serious concern than an analog communication system ??

There is no relation of phase noise to analog or digital communication system. Phase noise affects signal demodulation. If phase modulation is used, phase noise confuses phase demodulator.

Again it is clear to me that you have never read any textbook on communication systems and signal modulation. Without study you cannot understand our answers on the Forum.

 

[tony_lth]

Phase noise's effect on digi telecomm is very important.
For different modulation, different phase noise is required.
Such as for 3.5MHz QPSK, PN@1KHz may be important, but for 28MHz 64QAM, PN@10KHz is important.
That can be simulated during system design period.

 

[abhijitrc]

There is no relation of phase noise to analog or digital communication system. Phase noise affects signal demodulation. If phase modulation is used, phase noise confuses phase demodulator.

Again it is clear to me that you have never read any textbook on communication systems and signal modulation. Without study you cannot understand our answers on the Forum.

In digital commiunication system close in phase noise( ~1-10 Khz) is important and this hampers phase demodulation process ( eg in PSK QPSk), whereas in analog communication system, modulation information sometimes lies several Khz away from carrier, hence phase noise is of less concern , Plz do refresh ur all text books before make such comments on forum...

- - - Updated - - -

In digital commiunication system close in phase noise( ~1-10 Khz) is important and this hampers phase demodulation process ( eg in PSK QPSk), whereas in analog communication system, modulation information sometimes lies several Khz away from carrier, hence phase noise is of less concern , Plz do refresh ur all text books before make such comments on forum...

I want to know how much phase noise is of concern for several digital modulation technques--- for PSK ( 500 bits/s ) or for QPSK 1Kb/s will be affected by a noisy LO...

 

[abhijitrc]

Phase noise's effect on digi telecomm is very important.
For different modulation, different phase noise is required.
Such as for 3.5MHz QPSK, PN@1KHz may be important, but for 28MHz 64QAM, PN@10KHz is important.
That can be simulated during system design period.

If the bit rate is 500 bps( using PSK modulation) with carrier frequency is 8KHz, then for effective demodulation how much phase noise is tolerable ??, now if i increased the bit rate to 1Kbps with same carrier frequency of 8KHz then what will the phase noise in this condition? Is there any formula existing ?, Plz guide me.