A Clear Warning !

Chapter 5 of the Hewlett Packard Application Note N° 207, published on October 1976, starts with the following statement:

"Measurement of phase noise sidebands would be simple if frequency analyzers had dynamic ranges of 160 dB and 1 Hertz bandwidths useable in the GHz region. All measurements could then be made at the primary frequency of the source. However, equipment with this performance does not exist and so alternate techniques must be used."

Thirty years later, even with the technology advances we have witnessed, the above statement is still true. And even though modern equipment dedicated to phase noise measurement has made considerable progress, such instruments are still extremely expensive, and yet their performance is still similar the 3048. However, more modern instruments do boast better ease of use and the speed of measurement. But if you are reading this paragraph, you probably are not in a hurry, nor looking for an easy answer ;-)

HP 3048A Phase Noise - Basic Review

There are many techniques for measuring the phase noise from a source (or added by a device). How well each of these methods works depends on both the technique and the characteristics of what is being measured. In-depth analysis of these techniques is not the objective of this web site. Our goal is only practice-oriented. We just hope to help you by relaying our own experiences using the HP3048A.

Doing ourself this basic review, we found useful to build-up this chapter. It is nothing more than our own selection of references to the various available document in which you will find the best help to chose the right equipment and the right measurement method to achieve the best possible accuracy of the final result. In writing up advice, we found it best to simply reference the various excellent HP documents already available. They will help you choose the right instrument, and get the best possible accuracy.

You will find in the "Related Documentation" chapter all the necessary material to review basic noise concepts, which are important to remember before trying to do any serious phase noise measurement on a signal source. Once the basic techniques are understood, you can determine which are most applicable to the device you need to measure.

To help you in making a good choice, here is an abstract we made for ourselves, searching through this huge documentation list, highlights the most important points to take into account.

1. The Direct Spectrum Measurement Technique
2. The FM Discriminator Measurement Technique
3. The Two-Source, Phase Lock Loop Measurement Technique

The Direct Spectrum Measurement Technique

This is the simplest method used to measure phase noise of oscillators. The signal is directly input to a spectrum analyzer tuned to the oscillator frequency. The sideband noise power can be directly measured and compared to the carrier signal power. This method actually measures the total sideband noise power, AM plus Phase Noise. The measurement can be considered to be a phase noise measurement only if the AM noise is much less than the phase noise. The sensitivity of this measurement method is limited by the spectrum analyzer's internal Local Oscillator Noise. Another limit of this measurement approach is the analyzer's inability to track any signal drift during the measurement, thus limiting the close-to-carrier measurement capability.

Where to find more information on the Direct Spectrum Measurement Technique:

* The Application Note 270-2 "Automated Noise Sideband Measurements Using the HP 8568A Spectrum Analyzer"
* The HP 3048A Calibration Manual. The Direct Spectrum Technique description starts page 196 of the PDF document.
* The HP Phase Noise Seminar. The Direct Spectrum Technique description starts page 29 of the PDF document.

The available noise measurement range when using the Direct Spectrum Technique is shown below. The solid blue area is typically not available when using this technique. The boundaries tend to be slightly different for different spectrum analyzers and different frequency ranges.

Direct Spectrum Analysis is optimum for measuring of source with relatively high noise and low drift.

The FM Discriminator Measurement Technique

In this frequency-domain method of measurement, the signal under test is fed into a frequency discriminator and the output of the discriminator is monitored on a low-frequency spectrum analyzer. The best performance is obtained with a delay line / mixer combination as discriminator. The delay line converts frequency fluctuations of the source into phase fluctuations relative to the signal at the other port of the phase detector. The phase detector converts the phase fluctuations into voltage fluctuations for measurement and analysis by the spectrum analyzer.

Where to find more information on the FM Discriminator Measurement Technique:

* The HP 3048A Calibration Manual. The FM Discriminator Technique description starts page 219 of the PDF document.
* The HP Phase Noise Seminar. The FM Discriminator Technique description starts page 89 of the PDF document.
* The HP Product Note 11729C-2 is fully dedicated to the Phase Noise Characterization of microwave oscillators using the Frequency Discriminator method.

The overall noise floor of the FM Discriminator Measurement Technique will be the noise floor of the discriminator plus the noise of the measurement system. The noise floor of the discriminator dominates other contributors for close-to-carrier measurements. As shown on the graph below, changing the amount of delay will change the discriminator noise floor. But, while longer delays improve the noise floor, the eventual power loss in the delay line will exceed the source power available and cancel further improvement. The solid blue area is typically not available when using this technique.

FM Discriminator Technique main advantage resides in low noise floor, far-from-carrier analysis.

The Two-Source, Phase Lock Loop Measurement Technique

The basic measurement setup used for measuring phase noise with the PLL technique is shown in the diagram above. It needs two signal sources, a low noise reference, and the source under test. In this method the source under test is down converted to 0 Hertz or DC by mixing with the reference signal of the same frequency in a double-balanced mixer. The reference signal is set in phase quadrature (90 degrees out of phase) with the signal under test. When this condition of phase quadrature is met, the mixer acts as a phase detector, and the output of the mixer is proportional to the fluctuating phase difference between the inputs. Hence the SSB phase noise characteristics may be determined by examining the mixer output signal on a low frequency spectrum analyzer. The frequency of the noise displayed by the analyzer is equal to the offset from carrier.

Where to find more information on the Phase Detector Measurement Technique:

* The HP 3048A Calibration Manual. The PLL Measurement Technique description starts page 223 of the PDF document.
* The HP 3048A Operating Manual. Using the 3048A in the PLL Technique, starts page 40 of the PDF document.
* The HP Phase Noise Seminar. The PLL Measurement Technique description starts page 49 of the PDF document.
* The AN 283-3 and the AN 283-1 are mainly dedicated to the 8662/63 Synthesizers, and their use measuring Phase Noise.

The two-source measurement technique is by far the most sensitive method. It is the preferred one for measuring low noise, high stability oscillators with the HP 3048A. The graph below shows the typical noise floor of the system with no reference sources or downconverters.

In Summary, the PLL technique as the best overall system sensitivity and is generally considered the first choice among measurement techniques for sources.