In audio circuits, the phase noise voltage is uncorrelated and the voltage adds with the square root of the sum of the square. Since the noise is uncorrelated, if the two noise levels are the same, but random compared to each other (i.e., “uncorrelated”) the total phase noise voltage goes up by sqrt(2). If the two had been correlated, like two paths out of a phasing receiver on the desired sideband, then the voltages would be correlated and the voltage would be 2x rather than sqrt(2).
In a mixer things are different. All the products mix with each other. Thus the LO phase noise will mix with the phase noise of the signal and produce undesired products. Thus you can “hear a strong signal coming” as you tune closer to him for a combination of several things:
- The signal has poor phase noise and is strong enough that you can hear is the “in band” phase noise of the signal even when the main portion of the signal is not in the receiver passband.
- The signal is very clean, but your LO has poor phase noise such that the LO phase noise near the strong signal creates a mixing product that ends up in the receiver passband.
- The receiver LO is clean and the signal is clean, but the receiver is getting overloaded by the strong signal getting near the receiver pass band and distortion kicks in. This is a different effect than 1) or 2).
I guess you are worried about both 1) and 2) happening together. This is a superposition effect that would cause both noise products to be present at the same time. The source with the worse phase noise is the dominant factor.
Thus a transmitter with poor phase noise can overwhelm an excellent receiver with high dynamic range. The in band phase noise of the poor signal source can create a “real” in band signal that the “great” receiver will dutifully deliver.
| 13984
| 2017-06-07 10:37:36
| prysm678
| Re: is phase noise additive?
|
Dan Thank you that was very helpful. yes, I was mostly interested in items 1,2 of your list. regards fernan VE4FEB ---In emrfd@yahoogroups.com, wrote :
In audio circuits, the phase noise voltage is uncorrelated and the voltage adds with the square root of the sum of the square. Since the noise is uncorrelated, if the two noise levels are the same, but random compared to each other (i.e., “uncorrelated”) the total phase noise voltage goes up by sqrt(2). If the two had been correlated, like two paths out of a phasing receiver on the desired sideband, then the voltages would be correlated and the voltage would be 2x rather than sqrt(2). In a mixer things are different. All the products mix with each other. Thus the LO phase noise will mix with the phase noise of the signal and produce undesired products. Thus you can “hear a strong signal coming” as you tune closer to him for a combination of several things: - The signal has poor phase noise and is strong enough that you can hear is the “in band” phase noise of the signal even when the main portion of the signal is not in the receiver passband.
- The signal is very clean, but your LO has poor phase noise such that the LO phase noise near the strong signal creates a mixing product that ends up in the receiver passband.
- The receiver LO is clean and the signal is clean, but the receiver is getting overloaded by the strong signal getting near the receiver pass band and distortion kicks in. This is a different effect than 1) or 2).
I guess you are worried about both 1) and 2) happening together. This is a superposition effect that would cause both noise products to be present at the same time. The source with the worse phase noise is the dominant factor. Thus a transmitter with poor phase noise can overwhelm an excellent receiver with high dynamic range. The in band phase noise of the poor signal source can create a “real” in band signal that the “great” receiver will dutifully deliver.
| 13997
| 2017-06-12 09:49:31
| lmarion
| Re: is phase noise additive?
|
Google “is phase noise additive?” You will find far more detailed and informative information than can be handled here. Leroy | |