Paragon-X: Can I measure the time error of a DUT when switching between masters?

Owned by Kevin Godfrey
Last updated: Mar 05, 2019 by Gordon Muir (Deactivated)
1 min read

I see that Paragon-X has the ability to emulate two PTP Masters. Can I use this to measure the time error performance of my device when it switches between the two i.e. how it performs if BMCA is invoked due to one Master becoming unavailable?

Paragon-X's ability to emulate two masters is provided to allow testing of the BMCA (Best Master Clock Algorithm) of a DUT. Decoupling the two masters allows each to be configured, started and stopped individually, and so it is possible to use this to observe how a DUT behaves when its selected master becomes unavailable.

However, while the seed time used for both masters (ports) can be configured to be the same, the hardware implementation for port 1 and port 2 is different, and so there is no guarantee of alignment of PTP time on egress from the two ports. 

By design, most Paragon-X functionality uses port 1 as the PTP Master, and so the internal path delays for this port are well characterised and symmetric. The port 2 implementation does not require this same level of characterisation or symmetry, and so it is not valid to use the ‘two masters’ functionality to measure time error when using 2 masters.