What 1722.1 draft is app_1772_1_device compliant with?

[StephenTurner]

Hi Andy

The serial of the misbehaving board is 30011055. The one that seems to be OK is 30011049

Here are some observations:

set LabX switch to a priority of 254, so it should be master - correct?
power off both XMOS boards
power up #30011055
LabX switch shows it as a slave (so far so good)
power up #30011049
#30011049 comes up as slave (good), but now #30011055 switches to master (huh?)

Should I still get the debug logs?

regards
-steve

[Andy]

set LabX switch to a priority of 254, so it should be master - correct?

The XMOS endpoints are capable of being grandmaster and have a default priority1 value of 250 (recommended by the standard). The priority1 value of the switch must be less than 250 to force it to always be GM in this setup. By default the switch should have a priority1 of 246.

[StephenTurner]

Hi Andy

Looks like the endpoints are behaving themselves. It's the way I'm interpreting the info coming from the LabX switch.

Sorry for the fire drill.

Regards
Steve

[voodoosound]

hi stephen,

thats a correct behavior, since PTP builds an hierarchical tree clock.
if the switch is the grandmaster, it only has ClockMaster Ports.
if an endpoint is GM, the connected port changes his role to ClockSlave and all other switchports propagate the endpoints clock.