Measuring Accurate Coaxial Cable Delay, Using Time-of-Flight with CX41 TDR | Quick Reference Guide

Accurate Cable Delay Compensation for GNSS/GPS antenna cable feeds and 1PPS distribution is required to deliver Precision Time to applications requiring nanosecond-level accuracy

2022-09-22 IP/SK

TDRs capable of measuring and displaying the true time-of-flight delay, with nanosecond resolution, like VeEX's CX41 TDR, are required for proper compensation of cable-induced delays, to assure 1PPS timing accuracy down to the few nanosecond limits required by precision timing applications. 

• The total Cable Delay measured at the antenna RF feed, can be programmed into the GNSS receiver, for it to correct its time calculation and output an accurate 1PPS time alignment pulse.
• The 1PPS outputs on Primary Reference Time Clocks (PRTC) can also be programmed with the expected delay of the 1PPS distribution cable, so the timing pulse is accurate at the far-end of the cable (compensate for cable delay).

The BNC version of the CX41 is recommended for this application, since adapters for popular SMA, TNC and N connectors should be easy to source.

1. Using the CX41 TDR to Measure Coaxial Cable Delay

1.1 Getting Started

A. Test Port - Available with BNC(f) or F(f) connector

B. LCD Screen

C. Power Indicator LED

     Green: Tester ON (charged or working on battery power)

     Orange: Tester ON and charging

     Red: Tester OFF and charging

     Off: Tester OFF, not charging

D. Navigation and Mode Buttons

E. Rechargeable Battery Compartment (back)

F. Power Button

G. Charging Port (Micro-B USB Connector)

1.2 Configuring the TDR for Time Measurements

(1) Press the Power button for at least 5 seconds to turn the tester ON. The LCD backlight will turn ON and the power LED should turn green or orange.

(2) Press the button to select Manual TDR (or Auto TDR) mode. Manual mode is recommended since it shows the TDR graph, and the cursor position may require some adjustments. (Auto mode only shows the time to the first event and whether it is an open or a short, which could be from a coupler and not necessarily the end of the cable.)

(3) Press the Backlight button for at least three seconds, to activate the Setup Menu mode. In this mode, the functions of the buttons are indicated by the green labels below each button.

(4) Use the () to scroll through all the available settings, such as Distance/Time Range, Velocity of Propagation VOP (%), Firmware Update, Impedance, Measurement Units, Backlight Timeout, Power Down Timeout, Cable Type, About the test set, etc.

(5) Within each of the parameter screens, use the () and () buttons to scroll through the different options available. Leave it on the desired setting.

(6) Set UNITS to nSEC (nanoseconds) for direct Time-of-Flight mode.

(7) Set IMPEDANCE to 50 Ohms (typical for coaxial GNSS RF antenna feed and 1PPS clock distribution cables) or set it to match the specification of the cable being tested.

(8) Depending on the cable length, one can set the expected measurement RANGE (nS). The cable delay range could be estimated as 5ns x Length(m) or 1.5ns x Length(ft). The TDR can also be set to AUTO mode, so it automatically sets the range based on the events found during measurement. 

(9) Once configured, press the () button to apply and store the current configuration.

(10) Press to return to the TDR measurement.

1.3 Cable Delay Measurement

(1) Confirm that the is set to Manual TDR or Auto TDR.

(2) Connect the cable under test to the tester. Make sure to discharge any static charges and follow proper safety procedures, when handling antenna feeds from the roof or long coaxial cables. 

(3) Press to initiate the measurement.

Note: If left in Auto TDR mode, the initial result will display the delay based on the automatic event finder. However, it is better to verify with the trace that the cursor is indeed positioned at the end of the cable. Press to display the trace.

(4) The tester will place the cursor near the first significant event and display the time of flight to that event. Use the and buttons to place the cursor next to the rising edge of the main pulse reflection, at the end of the graph. The delay reading will update to the new cursor position. For example:

  

• Press continuously to move the cursor faster to the right and also increase the display range (and to see if there may be other events outside of the current viewing range).
• Press continuously to move the cursor faster to the left and also decrease the display range.

(5) For better event visibility (e.g., longer cables), use the following button combinations (press them simultaneously) to adjust the Gain Control and vertical scale. By default, the tester is in automatic gain control (AGC) mode and adjusts its vertical zoom according to the highest peaks.

• Press + to increase the gain.
• Press + to decrease the gain.

(6) Press , at any time, to refresh the measurement.

(7) Make a note of the Cable Delay reading, for the records, and use it to program the Cable Delay Compensations for the GNSS antenna cable feed or Primary Reference Time Clock's (PRTC) 1PPS outputs.

(8) Press the button to turn the tester OFF.

The ultimate goal is to be able to properly identify the end of the cable and position the cursor correctly, to get the most accurate delay measurement, even though there may be other smaller events in the cable (connectors, small impedance mismatch, in-line devices, etc.)

1.4 Application Examples

2. Understanding Basic Event Signatures in a Trace 

The expected traces and events depend on the application, type of installation, and any other passive or active in-line elements installed along the cable. With that in mind, there are no itemized standard PASS/FAIL recommendations. The key is to understand the meaning of each of the events on the TDR trace and compare them with the way a particular installation is expected to look like. That is, looking for unexpected events.

2.1 Cable Run Only (ideally with its far end is open)

In the CX41, the first spike in the graph (trace) represents the initial test pulse or the beginning of the cable. It should not be considered as an event.

2.1.1 Open-Ended Cable (No Antenna or Termination)

This is the desired test scenario, with a clean coaxial cable run. The goal is to properly identify the end of the cable and position the cursor at the beginning of the far-end reflection, to measure the total cable delay.

or

Open cables generate a strong positive reflection. The reference sample above shows the typical sharp positive reflection from a bare, non-terminated, cable looks like. This is the expected scenario to measure its delay. The first screenshot is for the automated detection and the second (pressing the Mode button) is by manually positioning of the cursor at the beginning of the reflection.

2.1.2 Shorted Cable

or

An electrical short (or very low impedance termination) typically generates a negative reflection. Since shorts are not expected in any installation, they are considered a fault. However, users may introduce a temporary electrical short (0 Ohms) to visually confirm the end of a cable (making the pulse change from positive to negative), when in doubt of complex TDR trace interpretation (e.g., multiple unexpected events). 

2.1.3 Connectors (Connections)

Connectorized extensions can often be identified by a small positive reflection caused by the physical mating of the connectors and/or small changes in impedance between the two cables. Bad connections produce larger reflections. Connectors that are farther aways (towards the end of long cables) are more difficult to identify.

Note: The screenshot above shows the small positive reflection from a connector mating, followed by a larger reflection from the end of the cable. However, there is a negative reflection equally spaced to the first (in reference to the larger one). This is referred as a "ghost" event, caused by multiple reflections. Ghosts should be ignored, so check the expected length and delay of the cable and compare.

2.1.4 Terminated Cable

Ideally, a perfectly terminated cable should not have any reflections, making it impossible to determine its end. In practice, adding resistive terminators do not provide a perfect impedance match, but the reflections would be faint. Hence, it is not recommended testing cables that are terminated at the far end. However, the CX41 will automatically adjust its gain to the maximum, which on some short cable, it may still be able to show small impedance mismatches.

2.1.5 Branches or Taps

Having cables branching, or taps, coming of the main cable run are not recommended for timing applications (e.g., using a passive T-connectors to split the signal). Not only do they introduce impedance mismatches, but they can create undesirable signal reflections that affect the performance of the cable installation and ultimately the quality of the satellites' signals (C/No). Below is an ideal representation of a bridge tap and its typical signature trace:

In practice, a passive tap (T) makes TDR traces difficult to interpret, since they introduce two events: (a) low impedance event at the T position, that looks like a negative pulse from a short and (b) a high impedance at the end of each branch, that looks like a positive reflection from an open. The pulses bouncing back and forth could also add multiple secondary reflections, that translate on a noisy TDR trace.

(a)  and  (b)

The problem with the resulting TDR traces is that these two events from the branch will superimpose with the events in the main cable run, along with multiple back-and-forth reflections, so they are not easy to identify or troubleshoot. Although seeing a complex trace like the samples above would not help users to identify the end of the main cable, it is a clear indication that something is not right.

However, if a branch cable is properly terminated, the trace won't provide much information. The only event visible would be the low impedance caused by the splitter (T). It is also an indication that terminated branches are not as problematic as open ones.

2.2 In-Line Elements

2.2.1 Lightning Arrestors and Surge Protection

When installed correctly, these in-line passive devices should be almost invisible to TDRs, as they don't interfere with low power signals. If they are close to the TDR, one may be able to see its connector's reflection signature.

2.2.2 Amplifiers & Power Injectors

When installed correctly, active in-line devices may not be easy to identify, which is a good indication (e.g., no impedance mismatch). Sometimes the TDR trace could just identify the first connector, as the active device can be unidirectional and block the test pulses. In some cases, they may have a DC power bypass circuitry that may allow pulses to (partially) go through. However, the trace beyond the active device may not make much sense, due to distortion. Different devices, models or brands may have different signatures. 

TDR testing is only recommended on passive cable installations. If active devices, like low noise amplifiers (LNA), are in line, normally they are physically disconnected from the coaxial cable, then each coax segment is measured independently and added up. Note that once the amplifiers are connected back, they will add some extra delay, but one would need to look into their specifications (datasheet) to see if they disclose such value. 

2.3 Terminated Cables (Antenna Already Installed)

TDRs work better on open cables (non-terminated), due to the stronger and sharper positive reflections, which allow for better cursor positioning and more accurate results. However, one may not always have access to the bare antenna feed cable and may sometimes find the GNSS antenna already attached to the top-end of the cable. If the antenna is already in place, there may be two scenarios (results may vary depending on the type of antenna installed). It is not always feasible to access the roof or climb the tower/pole to temporarily disconnect the antenna, so here is some information to help identify those cases.

2.3.1 Antenna Connected at the Far End, But Not Powered

This could be the second-best case scenario, since some antennas may send a strong negative reflection (they act like a short), which can still be used to determine the cable delay. (Note: The first small reflection shown on the trace is due to a coupler at 7.6m or 25ft.)

2.3.2 Antenna Connected at the Far End and Powered

In theory, a perfect impedance termination (match) would have no reflection, since all the energy is transferred to the terminating device. That would make it impossible to identify the end of the cable. Fortunately, in some cases small impedance mismatch may create some reflections that could look like a branch, with a low impedance (negative pulse) at the front end, followed by a small positive peak. Whenever possible, it is highly recommended to remove the power feed to the antenna, to get a better reading.

2.4 Regular TDRs or Distance/Length Mode

The use of regular TDRs (or Cable Length mode) is not recommended for Precision Timing applications. Although, technically speaking, all TDRs measure time-of-flight, most of them use a Velocity of Propagation (VP%) from an internal "typical" look-up table, or value manually entered by the user, to convert the time-of-flight to a distance estimate. These TDRs only show users the resulting estimated distance readings (rounded up). In turn, users have to use the cable's datasheet to find out its nominal (typical) VP% and manually convert length back to delay (time), which constitutes a second estimation. That can result in not-so-accurate measurements. Especially when different types of cables are involved.  TDRs with direct reporting of the actual delay (time-of-flight), like the CX41 TDR are recommended, since they eliminate the multiple approximation errors.

2.5 Presence of Active Elements

• The direct cable delay measurement TDR technique may not work when there are active components in the GNSS antenna cable feed, such as amplifiers and active splitters, which are not bi-directional. They won't allow the test pulse to go through.
• As an alternative in those cases, it is recommended to measure the individual segments of coaxial cables and add up the typical delay specified for the amplifiers or any other active elements.
• The typical maximum length of a non-amplified antenna feed is around 80-100m.
• The presence of lightning arrestors is normally not a problem for TDRs.

2.6 Non-Coaxial Cables

• The use of balanced interfaces and cables could make cable delay measurements a bit more difficult. Especially in the case of differential interfaces. They may require active components for conversion and those add significant delays to the 1PPS pulses.

©2022-2023 VeEX Inc. All Rights Reserved

Related Articles

• Cable Delay Compensation & Time Error Measurement Accuracy
• How to Configure the Precision Timing GNSS Receiver?
• Why is the number of satellites in the GNSS 'Satellite View' smaller than the ones in the 'Signal C/No' table?
• How to Configure the Built-In Atomic Clock for Precision Timing & Frequency (Sync) Applications?
• Using a Test Set as a Clock Signal Generator (Reference Source)