In real-world deployments, redundancy and recovery mechanisms are critical to maintaining accurate time synchronization if one of the time synchronization sources fails.
This test is designed to measure any changes in the time error at the output of each Telecom Boundary Clock when the primary Telecom Grandmaster (T-GM) fails, and a switch to the secondary T-GM occurs.

Two separate Microchip TimeProvider 4500 units were used as T-GM-A and T-GM-B, with all T-BCs being directly connected to both T-GMs. T-GM-A acted as the primary Grandmaster, as each Boundary Clock was configured with local priorities to prefer and lock to T-GM-A for both PTP and SyncE.
We disconnected the GNSS from T-GM-A, causing the BCs to switch to T-GM-B. After measuring for 1000s, including the failover, the measurement was restarted, and the GNSS connection to GM-A was reestablished, with all T-BCs locking back to T-GM-A on PTP and SyncE. The Calnex Paragon neo, Paragon neo-A, and the Keysight Time Sync Analyzer were used to measure the PTP and SyncE output of all the T-BCs.
We executed two separate test runs with different T-BCs but using the same procedure.
For both runs, the two-way time error for all devices was within the class-level accuracy limit of 6C (±65 ns), as per ITU-T G.8271, throughout the failover from T-GM-A to T-GM-B and the switchover back from T-GM-B to T-GM-A. For the MTIE, all devices were well within the threshold of the “Holdover Performance Const. Temp. T-BC/T-TSC Class A/B”, as specified in ITU-T G.8273.2, Table 7-10/Figure 7-1.

Table 75: Time Synchronization Source Failover - Devices under Test