This document provides an overview of clock synchronization for AS5xxx
Network Access Servers (NASs), and uses several examples to illustrate
There are no specific requirements for this document.
The information in this document is based on these software and
Cisco AS5xxx Network Access Servers.
Cisco 2691, Cisco 28xx, Cisco 3660 with MIX module installed, Cisco
37xx, and Cisco 38xx platforms.
For more information on document conventions, see the
Technical Tips Conventions.
Wide area digital transmission typically requires transmitters and
receivers to be properly synchronized. Local area asynchronous digital
transmission such as async RS-232 or Ethernet typically does not require the
internal clock of a receiver to be closely synchronized with that of the
transmitter. The receiver synchronizes anew with each frame sent. The timing of
an asynchronous receiver could differ from the timing of the transmitter by as
much as one part in a thousand with no loss of data.
In a time division multiplexing (TDM) network, all network components
must be synchronized with each other, or data may be lost. If a receiver runs
slower or faster than a transmitter, clock slips can occur. Each slip causes a
frame to be lost from, or added to, the data stream. The impact of slips
depends on the application that uses the TDM channel:
Digital applications such as synchronous Point-to-Point Protocol
(PPP) typically suffer the loss of a frame of data, and result in a
Voiceband modem sessions typically undergo a retrain, and result in a
transmission pause of approximately 15 seconds.
Voice calls typically suffer from an audible
When you connect AS5xxx NASs to TDM networks, it is important that you
carefully consider the clocking scheme you use. In the examples below, TDM
network components that are synchronized properly in one clocking domain are
shown in green. Other TDM components that are in a separate, unsynchronized
clocking domain are shown in orange, and packet network components are shown in
The NAS is connected to a single TDM switch, but is not synchronized to
that switch. The NAS can use the local oscillator (free running) or be
synchronized to some other source. There will be slips on the interface between
the NAS and the TDM switch. The solution is to configure the NAS to derive the
clock from the line.
The NAS is connected to two TDM switches. The two switches are in the
same TDM network, and are synchronized to each other. Configure the NAS to
derive clocking from one of the lines, and to fail over to the other line.
The NAS is connected to two TDM switches, which are in different TDM
networks. The networks are not directly clocked in common. However, they are
plesiochronous, and both use such accurate clocks that they are, for all
practical purposes, synchronized. As with Example
2, configure the NAS to derive clocking from one of the lines, and to
fail over to the other line.
The NAS is configured to derive clocking from the switch on the left.
The private branch exchange (PBX) uses a different source for synchronization,
and the orange span on the right will therefore experience slips. The solution
is to reconfigure the PBX to derive clocking from the line to the NAS.
In this case, the NAS is connected to two different TDM networks that
are not synchronized to each other. If the NAS derives clocking from the green
network, it will experience slips on the spans to the orange network. If it
derives clocking from the orange network, the spans to the green network will
slip. As long as the TDM networks cannot be synchronized, slips are inevitable
when they are interconnected. In this scenario, use two NASs and two separate
unsynchronized TDM networks, as shown in Example
Here, we have two unsynchronized TDM networks separated by a packet
network path. This prevents the occurrence of slips on the TDM spans. However,
if we build a circuit between the two networks with a Voice over IP (VoIP) link
through the packet network, synchronization problems can occur.
We have configured one NAS to source the clock (free-running). The
other is configured to derive clock from the line.
To configure the system to source clock, use the clock
source free-running controller command. To configure the system
to derive clock from the lines, use the clock source line
primary and clock source line
secondary commands. Use the clock source
internal command if you do not want to derive clock from a line
(for example, in Example 4, where the PBX
derives clock from the NAS).
(for versions earlier than Cisco IOS® Software
Release 12.2(11)T) or
(for Cisco IOS Software Release 12.2(11)T and later
versions) global commands to prioritize the clock sources.
Note: For information on the configuration commands for all other
platforms, refer to the respective user manuals.