M1 Oscilloscope Tools™: Phase Noise

Now with the press of one button M1 Oscilloscope Tools™ gives you all of the most important Phase Noise Measurements that systems costing ten times more than M1 provide AND gives the engineer new ways to analyze phase noise never before available.

Unlike Phase Noise systems that rely on spectrum analyzers or other frequency domain instruments, the M1 phase noise measurement group gives you more than just a mere phase noise number (though it does give you that). M1 shows you graphs of phase noise performance. For example: Phase Noise Spectrum, Instantaneous Phase Noise versus Time (fully sampled), Accumulated Phase Noise versus Time (fully sampled) and Frequency/Period versus Time (fully sampled). M1 gives you an ability to see the cycle-to-cycle variations in both Instantaneous Phase Noise AND Accumulated Phase Noise. This means M1 sees the transient nature of phase noise that is so critical to understanding THE PHASE NOISE NUMBER. With M1, not only will you get THE NUMBER, but you will have the tools that you need to understand that number and insights into how to reduce that number...

M1 can reach jitter noise floors as low as 330 fsec which is plenty good for most practical phase noise applications today. In addition, because all TimeViews and all FFTViews in M1 can be averaged, you can improve the signal-to-noise ratio by acquiring multiple acquisitions of the waveform. This will improve the signal-to-noise ratio and hence improve phase noise measurement noise floor also. NOTE: In the FFTVIEW averaging is actually called ACCUMULATE.

While frequency domain tools that cost much more than M1 and require you to buy new hardware, these tools tell the designer nothing about the real time cycle-to-cycle variations created by phase noise. Only M1 working on a real time scope, that you already own today, can give you this insight.

The press of one button puts up 4 views that most customers wanting a phase noise measurement will want to look at. Each of these four views provide measurements statistics (ieDUT's phase noise. the phase noise number), but they also help you easily visualize the real-life characteristics of your

The new phase noise measurement group started shipping with M1 v4.05 (early March, 2006).

FAQs:

What is the difference between the Phase Noise measurements in the frequency domain and in the time domain?

The time tag data of threshold crossing, created by M1, is equivalent to the accumulated phase of the clock period, also known as absolute phase. In the case of a 1 ns period clock, every 1 ns,Both amplitude versus frequency plots (whether converted from the time domain or taken in the time domain) will show the same spectral peaks in the same place with the same relative amplitude. If you're looking at a TimeView of Time Interval Error, the only difference is y-axis scaling. It's the EXACT same data, just scaled from time units into radian or degrees. In fact, in M1 you can toggle back and forth between the two axes units. represents one complete “cycle” or “revolution” and is equal to 2 pi radians. So M1; if you're looking at the FFT of Time Interval Error (which is what the current M1 FFT measurement is) there really is NO difference at all.

In addition, M1 has a FFTView of the “instantaneous deviation relative to the mean period” measurement. This analysis is equivalent to the instantaneous phase noise. Again, the vertical axes of the TimeView of cycle-to-cycle frequency view in M1 can be relabeled in radians or degrees. And in M1 you can view this information either in TimeView or FFTView (ie spectral view). But this is a different analysis that many designers will also find fascinating and very insightful. An analysis that you will not find in the frequency domain tools.

It should be noted that while there are equivalencies here, without the continuous record in time of the threshold crossing that M1 provides, conversion between the different representations of jitter is not possible. It is only M1 that will be able so show you all views of phase noise (instantaneous and accumulated, spectral and time domain, TimeView and HistoView). A frequency domain tool does not have access to the cycle-to-cycle threshold crossing information and as such will not be able to show all that M1 can show.

Can I measure the phase noise within a defined range of frequencies?

Yes. In the FFTView, M1 has a “Between Markers Only” measurement capability that allows you to measure just the jitter in a given frequency band. Press VIEW | Markers | Calculate Stats Only Between Markers | Drag the markers to your start and stop frequency.

Can the Noise Floor Calibration in M1 be used to improve the result of such measurement?

Not today. It is conceivable that the M1 Noise Calibration Procedure, currently used with SEEj RjDj Decomposition could be used to improve the RMS result. At this time we are not planning to add this capability.

What is the purpose of adding the Phase Noise measurements to M1?

Design Engineers want to use their scopes to make phase noise measurements. They do not want to purchase a dedicated phase noise test system. M1 can show them much more insight about phase noise than merely what is contained in the phase noise measurement group. M1 phase noise measurement are NOT just a number. It starts as a number, but the cycle-to-cycle TimeViews give designers real information about why it their phase noise is what it is.

What is the advantage of the Phase Noise measurements in M1?

The ability to see what's happening in a moment-to-moment view as opposed to just a single result number.