Paragon-X: Why is my SyncE Wander Transfer Test Failing?

Owned by Kevin Godfrey
Last updated: Jun 06, 2019
3 min read

If your wander transfer test is failing, you may be able to figure out what is going wrong by looking at the Paragon-X capture data.

 

What Data is Available?

When a wander transfer test is performed, Paragon-X will save one capture file for each frequency tested. So, when running only a single frequency, there will be one file; when running a table of frequencies, there will be one capture file for each row in your table.

Where are the Capture Files Stored?

All capture files are stored in the Paragon-X auto-save folder. See this article for information about the auto-save folder, where to find it and the naming convention used for capture files

You will need to have a rough idea of when you ran the test in order to find the correct files. Look in the auto-save folder for files created at around the time the test was run. The capture files you are looking for will have a filename ending with “_SE”.

If the test used a table of frequencies, then you will have a series of files with very similar creation times. The oldest of these corresponds to the first point in your table

How do I Analyse the Capture Data?

To analyse the data, you need to know how the measurement is performed and how a pass / fail result is calculated. See this article for details.

Load the capture from the frequency of interest into the CAT. The TIE/ESMC results tab will show you the TIE measured on the output of the DUT for that frequency.

For example:

The screenshot above shows the TIE measurement for a 2.104Hz frequency which was measured through a DUT with a stimulus amplitude of 0.25us.

From the above, you can see that the output pk-pk amplitude is 0.252us (125ns+127ns). The resulting gain is calculated from:

gain (dB) = 20.log10 (Amplout / Amplin)

So, in this case, we have:

gain (dB) = 20.log10 (0.252 / 0.25) = 0.069dB

At around 2Hz, this result passes the G.8262 limits.

The example below shows the result at 80Hz with 0.25us input amplitude:

In this case, the output pk-pk amplitude is 0.015ns giving a gain of -24.437dB. Although this would pass the G.8262 limits, it has clearly been affected by a transient when the frequency and amplitude have changed on the DUT input. The Paragon-X has a recovery time which is used to remove the effects of such transients but at 80Hz, this recovery time is set to 0s (see here for information about recovery times versus frequency).

To establish the actual gain for this point, the CAT can be used to remove the transient using the data analysis range:

Drag the Data Analysis Range lines to remove the transient period and then click Calculate. The statistics will be calculated over the selected range. In this case, the pk-pk output amplitude reduces to 8ns giving us a gain of -29.897dB.