effect of noise figure on phase noise

[young.microwave.eng]

hi all
I'm going to design some local oscillator.During design I run in to the effect of noise figure and phase noise.If some component have bad noise figure,Does it degrade phase noise?
If yes would you please tell me about the equation between noise figure and phase noise??
THANKS IN ADVANCE

 

[WimRFP]

The noise figure of the active device at the oscillation frequency is just a part of the story.

The active device's noise production (spectral power density of noise voltage and noise current) increases with reducing frequency (so-called 1/f noise below certain frequency, also called pink noise). Though this LF noise is outside the band of interest (when looking to an RF circuit), it has large influence due to non-linear behavior of the active device.

The LF noise, modulates the bias current and voltage of the oscillator. The LF noise at voltage and current changes the active device parameters (like capacitance and transconductance). This changes the phase and amplitude of the gain of the active device, hence the frequency of oscillation is modulated with some of the LF noise. This is because to satisfy the oscillation condition the phase of the feedback network (incorporating the resonator) needs to correct the active device phase shift.

The steep skirts left and right of the frequency of oscillation are produced by the LF noise. The less steep skirts further away from the carrier are produced by the in-band noise production of the active device.

So for making a low phase noise oscillator, you need a device with low 1/f noise.

I know this doesn't answer your question completely, but I hope it will help you to find a complete answer.

 

[tony_lth]

The phase noise floor is equal to -174+NF(dB)-Pout(dB). So if you have bad NF, so you have bad noise floor. If your power level is very low, you have bad noise floor also.

 

[jiripolivka]

The phase noise floor is equal to -174+NF(dB)-Pout(dB). So if you have bad NF, so you have bad noise floor. If your power level is very low, you have bad noise floor also.

In principle, in RF and microwave oscillators there are two kinds of noise:

AM noise, defined by oscillating device noise figure (NF is a representation of thermal noise) and often by low-frequency fluctuations of device current (1/f type).
In AM systems the AM noise has some importance but in other systems using FM and PM modulation, AM noise is usually suppressed by limiters.

FM or PM noise, mostly referred to as phase noise. This noise has a typical spectrum with a peak close to carrier, falling with a higher distance from it. All PM and FM systems are sensitive to it, and the higher the carrier frequency, the PM noise measured in dBc/Hz power density, grows.

Carrier frequency is generated in simple oscillators, often stabilized by quartz resonators. The high Q of such resonator offers a lower phase noise. To stabilize RF an microwave frequency, we use phase-locked and frequency-multiplied generators, with a low-frequency reference source. This reference has to be the best quality to get a reasonable phase noise.
Reference oscillators, VCXOs and OCXOs suffer by mechanical noise and noise from power supplies that significantly drives high the phase-noise spectrum.
Today there is an extensive literature, books and papers on the problem.

 

[young.microwave.eng]

hi
thanks for your replies
as I mentioned I'm going to design some stable oscillator and for that I choose DDS structure.my stable oscillator have two kind of outputs.1)constant frequency 2)variable output
for variable I choose DDS AD9910.
there is some simulator in analog device and it is possible to see the output of selected DDS
after simulating,I see that there are some harmonics in my desired band width when I'm sweeping the frequency
So I have to divide the output of DDS in to two distinct part and pass them through two filters.to divide the output I'm going to use active switches.I dont know If the active switch will degrade the phase noise or not??
because active switches always have high phase noises and I dont know how this high phase noise will affect my phase noise.
THIS IS MY PROBLEM.IF YOU HELP ME PREDICT THE EFFECT OF SWITCH ON PHASE NOISE OF MY OUTPUT SIGNAL,THEN THAT WILL BE OF YOUR KINDNESS
THANKS IN ADVANCE

 

[tony_lth]

So what is the target of your phase noise level before passing a switch?

 

[jiripolivka]

hi
thanks for your replies
as I mentioned I'm going to design some stable oscillator and for that I choose DDS structure.my stable oscillator have two kind of outputs.1)constant frequency 2)variable output
for variable I choose DDS AD9910.
there is some simulator in analog device and it is possible to see the output of selected DDS
after simulating,I see that there are some harmonics in my desired band width when I'm sweeping the frequency
So I have to divide the output of DDS in to two distinct part and pass them through two filters.to divide the output I'm going to use active switches.I dont know If the active switch will degrade the phase noise or not??
because active switches always have high phase noises and I dont know how this high phase noise will affect my phase noise.
THIS IS MY PROBLEM.IF YOU HELP ME PREDICT THE EFFECT OF SWITCH ON PHASE NOISE OF MY OUTPUT SIGNAL,THEN THAT WILL BE OF YOUR KINDNESS
THANKS IN ADVANCE

You caught me! I have no experience with DDS. Please refer to your AD device specifications; Analog Devices also have white papers on the PN problem. If your switch can be synced with the reference oscillator, there will be a minimum PN contribution.

 

[young.microwave.eng]

hi again
thanks for your reply
There is no problem with DDS.
the main question is:Imagine that there is a signal with phase noise for example -120dBc/HZ.The signal goes through some microwave part for example switch with noise figure=4 dB.then what is the phase noise of output signal???

 

[G4BCH]

Tony_lth has already told you, -174+NF(dB)-Pout(dB).
In this case Pout is you signal level. Assuming it is 0dBm then the contributed noise from your switch is -170dBm/Hz or -170dBc/Hz in this case.
Ignoring here that some of the added noise will be AM and some PM.
Compare that with your phase noise of -120dBC/Hz and you can easily see that it will have no measurable effect at all, it is 50dB below your oscillator.
Peter

 

[Mityan]

The NF shows the noise growth at the output relative to input. So it seems after passing unit with 4 dB NF the noise will be -116 dBc/Hz, won't it?

 

[G4BCH]

No.
Noise figure is related to the noise generated by a resistor at 290 kelvin kTB ie -174dBm/Hz.
Add the noise figure in dB to that and that becomes the noise power that is added to the signal passing through the system.
The signal to noise ratio is degraded by the added noise power not by the noise figure.
Peter

 

[Mityan]

Sorry, but I cannot agree with it. The noise figure is measured, when nothing is applied to the sustem input but resistor thermal noise. But it doesnt mean that while signal is going through the system, nothing is added but such thermal noise. NF shows the noise growth, so for signal S and noise N at the input of system with gain G and noise figure NF the output will be G*A + (G+NF)*N.

 

[WimRFP]

Imagine you have a 10 dBm signal that arrives at the output of the switch. When the phase noise spec is -120dBm/Hz, the noise is -110 dBm/Hz (at xx kHz from the carrier)..

Now the switch

F = 10 log(1+Pnadded/Pnresistor) For calculation of F, Pnresistor = kT [W/Hz] at 290 K this is -174 dBm/Hz.

When F = 4, the added noise (Pnadded) is 1.5*Pnresistor.

So the added noise is -174 + 10log1.5 = -172 dBm/Hz remember: oscillator phase noise was -110 dBm/Hz at 10 dBm carrier power

Ignoring that a part of this noise is amplitude noise (so actual added phase noise is even less), the noise contribution from the switch to the phase noise of the oscillator is not measurable.

So I agree with G4BCH

 

[jiripolivka]

I agree with G4BCH on that you tend to mix AM and PM noise. Thermal noise and phase noise are in quadrature, their momentum vectors are mutually normal.
PM noise in principle refers to phase fluctuations, not amplitude ones. Only the noise figure of the active device of an oscillator can affect the carrier phase fluctuation. Latter contribution of loss or thermal noise do not contribute to phase noise.
This is why the PN is measured in dBc/Hz as a function of spectral separation from a carrier. In any signal-generation chain, the PN is generated in the first (reference) oscillator, then only scaled up if the frequency is multiplied.

Recently the new sensitive PN measuring systems have detected certain very low level of PN contribution by amplifiers and switches. So it cannot be said there is NO contribution by nonlinear elements but such contribution is typically >30 dB below the oscillator PN.

 

[WimRFP]

@jiripolivka: I fully agree with you.

Theoretically an amplifier can contribute to the phase noise, even when F_amplifier = 0. low frequency noise (1/f or 1/f^2), or noise from power supply or bias circuits can modulate the gain and phase of an amplifier. Time varying phase results in phase modulation of the output signal, hence phase noise. In case of switches based on saturated components (FET switches) several 1/f noise phenomena are not present, so phase noise contribution from switches will be negligible.

I am happy with your "typically > 30 dB below oscillator PN statement".

 

[Dr.Drew]

The phase noise floor is equal to -174+NF(dB)-Pout(dB). So if you have bad NF, so you have bad noise floor. If your power level is very low, you have bad noise floor also.

Perhaps you forgot to sutract 3 dB to consider only a phase part of a wideband thermal noise kT.

 

[Mityan]

oscillator phase noise was -110 dBm/Hz at 10 dBm carrier power.

Consider we take the noise of the switch into account when it reaches 1/4 of oscillator phase noise - 6 dB below.
So 110-6-(-174) = 48 dB.
We dont care about switch noise figure until it is 48 dB. Or about 20000 times more than given 4 dB... That's just ridiculous
So why do we tend to minimize it if it is so negligible?
Read carefully about SNRin and SNRout: https://en.wikipedia.org/wiki/Noise_figure

 

[young.microwave.eng]

Tony_lth has already told you, -174+NF(dB)-Pout(dB).
In this case Pout is you signal level. Assuming it is 0dBm then the contributed noise from your switch is -170dBm/Hz or -170dBc/Hz in this case.
Ignoring here that some of the added noise will be AM and some PM.
Compare that with your phase noise of -120dBC/Hz and you can easily see that it will have no measurable effect at all, it is 50dB below your oscillator.
Peter

hi all
this formula is correct for offset frequencies that is far enough from the carrier.but for 1/f noise i dont think it is true
please see this file from wenzel and this approve me
https://www.wenzel.com/documents/hints.htm
I agree with you for offset frequencies that far enough from carrier but for closer offsets it is not true.Please think about closer offsets

 

[Dr.Drew]

hi all
this formula is correct for offset frequencies that is far enough from the carrier.but for 1/f noise i dont think it is true
please see this file from wenzel and this approve me
https://www.wenzel.com/documents/hints.htm
I agree with you for offset frequencies that far enough from carrier but for closer offsets it is not true.Please think about closer offsets

For proper calculation of phase noise induced by amplifier at close offsets one should use a flicker coefficient which is unknown in most cases (see Enrico Rubiola publications about phase noise in amplifiers). Ampllifier manufacturers don't measure this parameter because of small demand. Almost every sinusoidal signals have phase noise much intensive than one induced by amplifier.

 

[G4BCH]

If you look at the added phase noise of the amplifiers given in the link they are all way below your -120dBc/Hz so you would not see any degradation there. You dont give the operating frequency or offset I guess offset about 100Hz.
As for a switch I have never measured the added phase noise, I've always been more concerned about other affects that could degrade the signal integrity. I have generally used diode switches but have never seen any PN degradation using either FETs or diodes as switches in relatively low noise signal sources at high frequencies.
I think you will find there are more important things to worry about in designing a local oscillator than the added phase noise of any switches, isolation for one.

@Mityan
In this case it is negligible. The signal levels are so much higer than the noise added by the 4dB loss. If it were a low noise front end it would be a completely different story, there the input noise (sky noise) is very low and the switch loss and its added noise are significant and can not be ignored as the signal to noise ratio will be significantly impaired.

Peter