Multi Boundary Clock Holdover Test

Resilience scenarios have always been a key focus of our annual testing event, designed to simulate network failures to validate the performance of PTP (Precision Time Protocol) during a source loss.
This test case evaluates the ability of boundary clocks to maintain acceptable time error during the loss of the Telecom Grandmaster (T-GM) as the PTP reference.
When the T-GM becomes unavailable, boundary clocks enter holdover mode while remaining locked to the SyncE frequency reference from the T-GM.
The goal is to evaluate the holdover performance of each individual Telecom Boundary Clock (T-BC) to ensure stable time synchronization and minimal drift without the primary time reference.

Microchip participated in this test with TimeProvider® 4500 as the Telecom Grandmaster (T-GM), providing PTP and SyncE to each T-BC. A second TimeProvider® 4500 with a rubidium internal oscillator was used as the reference clock for the two measurement tools, the Calnex Paragon-Neo PAM4 and the Keysight Time Sync Analyzer.

After all seven T-BCs were locked to the T-GM, Microchip configured the TimeProvider 4500 to advertise clock class 140, causing all the T-BCs to enter PTP holdover while still being locked on SyncE.
The measurement was left running for 1 hour. Calnex and Keysight measured MTIE, with the MTIE per device ranging from 9.75 nanoseconds (ns) to 42.734ns, with six out of seven devices below the 13ns mark.
All seven T-BCs were below the T-BC Class B holdover performance threshold at a constant temperature as defined in ITU-T G.8273.2, section 7.4.2.2, table 7-10/figure 7-1.