1. Generic hints

  • Whenever you find yourself in trouble while setting up the measurement – use Autoset:
    • Connect the signals,Choose measurement function/inputs,Choose Sample Count and Sample Interval,
    • Press Autoset button,

And it will find proper settings for most cases when measuring continuous signals.

  • Use Auto choice for settings items unless you understand the implications of selecting other option.
  • Settings → User Option → Recall Defaults will reset measurement settings to Defaults.
  • Save complex measurement configurations as Presets (Settings → Measurement Presets or dedicated icon on measurement screen). In this case you can easily recreate the same measurement setup if you need it later.
  • Make sure input circuits are setup appropriately:
    • Use Auto-trigger for signals above 100 Hz, otherwise make sure Absolute trigger level is set. appropriately (prefer using Autoset for low frequency signal – it will set trigger levels for you).Make sure input impedance is set correctly.Use only DC coupling for low frequency signals and rely on Autoset to setup proper trigger levels.Keep Preamp OFF, except for extremely low input signal levels (below 50 mVrms).Keep Attenuation 1x, except for signals with amplitudes above exceeding +/- 5V.
    • Keep Filter OFF, except for low frequency sine wave signals (below 100 kHz).
  • Settings → Advanced → Voltage Mode should be set to Normal, except for signals below 100 Hz. For signals below 100 Hz please use Autoset to let the counter select best voltage mode for you. Set Voltage mode explicitly only if Autoset fails to find appropriate instrument setup (e.g. for not continuous signals)

2. Measuring 1 PPS

Hints:

  • Set DC on inputs 1 PPS signals are connected to,
  • Select measurement function and inputs,
  • Set Sample Interval to 0,
  • Use Autoset or set Trigger Mode to Manual and set Trigger Level to the middle of 1 PPS signal voltage range.

Same is most of the time true for signals below 100 Hz.

3. Measuring single cycles or pulses

Hints:

  • Set DC on the inputs, which the signals are connected to,
  • Select:
    • Period Single for measuring single cycle frequency period, or
    • Time Interval Single for measuring intervals between events, or
    • Totalize for counting events, or
    • Any function from Pulse group for measuring pulse characteristics.

Please note: using other functions will not give reliable results for single cycles/pulses.

  • Set Trigger Mode to Manual and set Trigger Level to the middle of signal voltage range.

Please note: auto-trigger won’t work for single cycles/pulses.

4. Measuring Frequency/Period

See 6.2 Measuring 1 PPS if signal is below 100 Hz.

Hints:

  • The basic setting Sample Interval in the Measurement menu is central to all Frequency related measurement. This setting means the same as Measuring time, or Gate time, used by other counter manufacturers.

A long Sample Interval (Gate time) increases resolution (counting during a longer time) but decreases measurement speed. The Sample Interval is always a compromise between how many digits you want to read, and how fast you want to take your frequency samples. For normal bench use – 200 ms is a good choice, because it is hard for the eye to follow faster changes in the displayed value.

The CNT-104S will give 12-13 digits resolution with 1 s Sample Interval, 9 digits with 1 ms Sample interval, and 6 digits with 1 μs Sample Interval

  • Use AC Coupling because possible DC offset is normally undesirable.
  • Use Trigger Mode Auto and/or Autoset.
  • Use Preamp ON for signals with amplitudes below 200 mVpp.

Please note: amplifying the signal also amplifies the noise.

  • Sample Interval of 200 ms is a reasonable tradeoff between measurement speed and resolution on the bench.

5. Time measurements of continuous signals

Hints:

  • Use DC coupling.
  • High signal level, and Steep signal edges.

6. Jitter measurements

Statistics provides an easy method of determining the short term timing instability, (jitter) of pulse parameters.

Note: that the measured pulse parameter should be a single cycle value, whether it is period, or pulse width.

6.1. Single cycle jitter

Single cycle jitter made on random samples of single periods, is usually specified with its rms value, which is equal to the standard deviation based on single measurements. The Analyzer can then directly measure and display the rms jitter. Jitter can also be expressed as peak-to-peak value, which is also displayed in the Statistics screen.

6.2. Cycle-to-cycle jitter

Cycle-to-cycle jitter demands zero dead-time measurements without gaps and can be made on input signals with a jitter frequency of up to 20 MHz. There is currently no dedicated measurement function, but the raw data of a Period Average measurement, with Sample Interval of 0 or down to 50 ns, could be exported to e.g. Matlab or Excel for “number crunching” and analysis.

6.3. Wander measurements

Wander measurements, which is a “slow jitter” measurement with jitter frequencies <10 Hz is made by using the TIE function, which compares the accumulated period phase drift, with the ideal phase from an ideal clock.

6.4. Deterministic jitter

Deterministic jitter is revealed in the Distribution graph, which will show underlying noise sources in a clear way. For example a sine modulated noise source would give a bathtub shape, a pulse modulated noise source would give a twin peak shape, and a measurement of a source containing not one, but two, fundamental frequencies will be displayed as “double hump”.

7. Frequency Modulated Signals

A frequency modulated signal is a carrier wave signal (CW frequency = f0) that changes in frequency to values higher and lower than the frequency f0. It is the modulation signal that changes the frequency of the carrier wave.

The Analyzer can accurately measure:

  • f0 = Carrier frequency.
  • fdev = Frequency deviation = (fmax -fmin)/2. And via the timeline graph, you will also get a good indication of the modulation frequency fmod

7.1. Initial capture settings

The optimum settings is to find a balance between large enough sample intervals to achieve high resolution per individual frequency sample, max. 10% of the Frequency deviation.

And small enough Sample intervals to capture enough frequency samples per modulation cycle for good graph visibility.

A rule of thumb is that the number of samples per modulation cycle should be >10, for good graphical view of the modulation signal, and acceptable error of fmax and fmin

Example: 10 kHz modulation frequency (100 us modulation cycle) of a 200 MHz carrier with 200 kHz deviation (0.1% modulation).

Set Sample interval to 10 μs (10% of the modulation cycle). Set Sample Count (N) to 100 (to cover 10 modulation cycles).

Now every frequency sample will have a resolution of  (10 ps/10 μs) x 200 MHz = 200 Hz.

This resolution is 1000 times better than the frequency deviation.

Start measurement and view the Timeline graph, which will show10 modulation cycles, with 10 samples per modulation cycle.

To improve the graphical experience, lower Sample Interval to 1 μs (1% of the modulation cycle), and increase Sample Count to 1000.

Now each frequency sample has a resolution of (10ps/1μs) x 200 MHz = 2 kHz. Still with a lot of margin to deviation.

You may want to play around with the Sample Interval and Sample Count setting until you have found your optimum view of the FM signal (no of displayed mod. cycles).

7.2. Carrier Wave Frequency f0

To determine the carrier wave frequency, just look at fmean which is best approximation of f0.

Ideally the sum of all Sample Intervals should be selected to cover an integer number of modulation periods. This way the positive frequency deviations will compensate the negative deviations during the measurement.

Example: If the modulation frequency is 1 kHz, the Sample Interval 10 μs and N = 1000 will make the Analyzer measure exactly 10 complete modulation cycles. A bad combination of Sample Interval and N would worst case mean that exactly half a modulation cycle is uncompensated for, giving a max. error for a sine modulation of:

f0 – fmean = Δfmax / (sample int.) x N x fmod x π

For very accurate measurements of the carrier wave frequency f0, make an extra measurement session and set Sample Interval as close as possible to an integer number of modulation cycles, and increase the number of samples substantially. A worst case error of half a modulation cycle means far less in a million cycles compared to 10 cycles.

7.3. Frequency deviation fmax - f0

Read the max, min, and mean frequency values from statistics screen or beneath the graph and calculate fdev as either:

  • fmax – fmean
  • fmean – fmin
  • fp-p/2

These three values should be exactly the same for an ideal sine wave or square wave modulation.

7.4. Modulation frequency fmod

The modulation frequency is easiest found by visual estimate in the graph on screen by using Cursors. Place one cursor on the beginning of the first modulation cycle and the other – on the end of the last modulation cycle. If the end point time difference between cursors is T seconds and the exact integer number of modulation cycles between cursors is M, the modulation frequency is:

fmod = M/T

7.5. Errors in fmax, fmin, and fp-p

A too large Sample Interval compared to the modulation cycle time leads to an averaging error that will underestimate the true deviation. A Sample interval corresponding to 10% of the modulation cycle, or 36° of the modulation signal, leads to an error of approx. 1.5%.

If that error is not acceptable, decrease the Sample Interval to make more samples than 10 during the modulation cycle.

8. Frequency profiling

Profiling means measuring and plotting frequency variation versus time. Examples are measuring warm-up drift in signal sources over hours, measuring the linearity of a frequency sweep during seconds, VCO switching characteristics during milliseconds, or the frequency changes inside a “chirp radar” pulse during sub-microseconds.

The Analyzer can handle many profiling measurement situations with some limitations. In profiling applications, the Analyzer acts as a fast, high-resolution sampling front end, storing results in its internal memory. These results are later displayed on screen and/or transferred to the controller for analysis and graphical presentation.

You must distinguish between two different types of measurements called free-running and repetitive sampling.

8.1. Free-Running Measurements

Free-running measurements are performed over periods down to the sub-microseconds range, e.g., to measure initial drift of a signal generator or oscillator, to plot linearity of a sweep signal ramp, or to measure short-term stability down to microsecond averaging times. In these cases, measurements are performed at user-selected Sample Intervals, and performed as gap-free measurements in the range 50 ns to 1000 s.

Just start the block measurement and view the profile in the graph presentation mode.

8.2. Repetitive Sampling Profiling

The measurement setup just described will not work when the profiling demands less than 50 ns intervals between samples.

How to do a VCO step response profiling with 50 samples during a time of 1 us.

This measurement scenario means that you need to come to 20 ns between samples (50 points * 20 ns = 1 ms observation time).

You will need a repetitive input step signal, and you have to repeat your measurement 50 times, taking one new sample per cycle. And every new sample should be delayed 20 ns with respect to the previous one.

Profiling can theoretically be done manually, but the best would be to perform this series of measurements with a dedicated program on PC, controlling the Analyzer and collecting data from it remotely.

The following are required to setup a measurement:

A repetitive input signal (e.g., frequency output of VCO). An external SYNC signal (e.g., step voltage input to VCO). Use of start arming delay (20, 40, 60 ns, etc). See Figure 38 for a test setup diagram.

Figure 41. Setup for transient profiling of a VCO.

8.3. Vrms

When the waveform (e.g. sinusoidal, triangular, square) of the input signal is known, its crest factor, defined as the quotient (QCF) of the peak (Vp) and RMS (Vrms) values, can be used to set the constant K in the mathematical function K*X+L. The display will then show the actual Vrms value of the input signal, assuming that Vpp is the main parameter.

Example: A sine wave has a crest factor of 1.414 (√2), so the constant in the formula above will be 0.354.