Trace Noise

M9370A / M9371A / M9372A / M9374A / M9375A

previous page next page

Noise Reduction Techniques

Random electrical noise which shows up in the analyzer receiver chain can reduce measurement accuracy. The following features help reduce trace noise and the noise floor which can lead to better dynamic range and more accurate measurements.

See Also

Group Delay

Increase Dynamic Range

Other topics about Optimizing Measurements

Averaging

Averaging is a feature that reduces the effects of random noise on a measurement. The analyzer computes each data point based on the average of several  measurements. You determine the number of measurements by setting the Average factor. The higher the average factor, the greater the amount of noise reduction.

Effects of Sweep Average

Both Averaging and IF Bandwidth can be used for the same benefit of general noise reduction. For minimizing very low noise, Averaging is more effective than reducing IF bandwidth. Generally, Averaging takes slightly longer than IF bandwidth reduction to lower noise, especially if many averages are required. Also, changing the IF bandwidth after calibration results in uncertain accuracy.

How to Set Averaging

Using front-panel hardkey [softkey] buttons

Using Menus

  1. Press Avg

  2. then [Averaging]

  1. Click Response

  2. then Avg

  3. then Average

Average dialog box help

Average ON   Check to enable Averaging.

Average Factor   Specifies the number of measurements that are averaged. Range of 1 to 65536 (2^16).

Average Type

Sweep  Each data point is based on the average of the same data point measured over consecutive sweeps. When the number of sweeps = Average Factor, the averaging continues following the Sweep Averaging formula.

(Sweep) Restart   Begins a new set of measurements that are used for the average. Applies only to Sweep averaging - NOT Point.

Point   Each data point is measured the number of times specified by the Average Factor, and then averaged, before going to the next data point.  

  • On subsequent sweeps, averaging is automatically restarted by measuring each data point again the number of times specified by the Average Factor.

  • Because measurements occur quickly in the background, the Average Counter is NOT updated.

Notes

  • An Average Counter appears on the screen when Sweep averaging is selected, displaying the number of sweeps that has been averaged. The effect on the signal trace can be viewed as the Average Factor increases. This can assist in the selection of the optimum number of sweep averages. The Average Counter is NOT updated for Point averaging.

  • Channel-wide scope- Averaging is enabled and the factor is set for all measurements in a channel. The Average counter is displayed for each channel.

  • Calibration - Because averaging is a mathematical process that occurs after the raw measurement is made, averaging can be turned ON before or after calibration without invalidating the error correction terms. If averaging is ON before calibration, the measurement of calibration standards are averaged measurements. More time is needed to perform the calibration, but there will be less noise in the resulting error correction terms. Subsequent corrected measurements will also have less noise error. In addition, noise is further reduced by turning Averaging ON after calibration. See the data processing map.

  • Triggering is implemented separately from Averaging. For example, setting averaging factor to 4 has NO effect on the number of triggers that are required to achieve 4 sweeps or 4 data points.

  • Unratioed measurements - Although averaging unratioed (single receiver) measurements is allowed, you may see unexpected results.

  • The noise floor does not drop when averaging unratioed measurements as on ratioed measurements.

  • Phase results may tend toward 0.  This is because phase measurements are relative by nature. Measuring absolute phase with a single receiver appears random. Averaging random positive and negative numbers will tend toward 0.

Sweep Averaging Formula

NewAvg = (NewData/n) + [OldAvg*(n-1/n)] 'where n = average factor

From the formula, you can see that data from the first n sweeps continues to be included in the results of subsequent sweeps. Its effect is increasingly smaller but never diminishes to zero.  For example, with n = 5, the average of the 5 sweeps is displayed. On the 6th sweep, you see 4/5 the average of the first 5 sweeps plus 1/5 the new sweep.

The effects of older data can be eliminated by clicking Restart.

Learn more about Averaging (scroll up)

IF Bandwidth

The received signal is converted from its source frequency to a lower intermediate frequency (IF). The bandwidth of the IF bandpass filter is adjustable from 40 kHz (for most PNA models) down to a minimum of 1 Hz.

Reducing the IF receiver bandwidth reduces the effect of random noise on a measurement. Each tenfold reduction in IF bandwidth lowers the noise floor by 10 dB. However, narrower IF bandwidths cause longer sweep times.

  • Channel - IF bandwidth can be set independently for each channel

  • Segment sweep - IF bandwidth can be set independently for each segment of segment sweep.

  • Calibration - Changing the IF bandwidth after calibration will cause a 'C-delta' correction level, which means that calibration accuracy is uncertain.

Effect of Reducing IF Bandwidth

 

How to set IF Bandwidth

Using front-panel
hardkey [softkey] buttons

Using Menus

  1. Press Avg

  2. then [IF Bandwidth]

  1. Click Response

  2. then Avg

  3. then IF Bandwidth

IF Bandwidth dialog box help

IF Bandwidth  Specifies the IF (receiver) bandwidth. The value of IF bandwidth is selected by scrolling through the values available in the IF bandwidth text box. The IF BW is set independently for each channel.

The following is a list of selectable IF Bandwidths:

10 | 20 | 30 | 50 | 100 | 200 | 300 | 500 | 1k | 2k | 3k | 5k | 10k | 20k | 30k | 50k | 100k | 300k | 600k | 1.2M

 

Learn about IF Bandwidth (scroll up)

Trace Smoothing

Trace smoothing averages a number of adjacent data points to smooth the displayed trace. The number of adjacent data points that get averaged together is also known as the smoothing aperture. You can specify aperture as either the number of data points or the percentage of the x-axis span.

Trace Smoothing reduces the peak-to-peak noise values on broadband measured data. It smooths trace noise and does not increase measurement time significantly.

Because Trace Smoothing follows Format in the data processing map, the formatted data is smoothed. Smoothing is automatically turned off if the format is Polar or Smith Chart.

Learn more about Data Format Types.

See the data processing map.

Tips:

  • Start with a high number of display points and reduce until you are confident that the trace is not giving misleading results.

  • Do not use smoothing for high-resonance devices, or devices with wide trace variations. It may introduce misleading information.

  • Smoothing is set independently for each trace.

Effects of Smoothing on a Trace

 

How to set Trace Smoothing

Using front-panel
hardkey [softkey] buttons

Using Menus

  1. Press Avg

  2. then [Smoothing]

  1. Click Response

  2. then Avg

  3. then Smoothing

Smoothing dialog box help

Smoothing ON  When checked, applies smoothing to the displayed trace.

Percent of Span  Specify percent of the swept stimulus span to smooth. For example, for a trace that contains 100 data points, and specify a percent of span = 11%, then the number of data points that are averaged is 11.

Points  Specify the number of adjacent data points to average.

Learn about Trace Smoothing (scroll up)

 

previous page start next page