The customer observed PLL Access Failure logs are seen in MX10008 with LC9600 (22.4R3.25), this Junos OS had fixed the PR/1715831 issue, however, the customer still observed the PLL access failure logs.

Related syslog messages - removed the customer router name:
Sep 26 18:27:50 router01 craftd[12867]: Major alarm set, FPC 2 PLL Access Failure
Sep 26 18:27:50 router01-fpc2 clockd[14873]: EMF_EVO_ALARM_SET: Alarm set: CG color=red, class=CHASSIS, reason=FPC 2 zl30642 PLL Access Failure
Sep 26 18:27:55 router01 alarmd[13780]: Alarm cleared: FPC id=1224737384, color=RED, class=CHASSIS, reason=FPC 2 PLL Access Failure
Sep 26 18:27:55 router01 craftd[12867]: Major alarm cleared, FPC 2 PLL Access Failure
Sep 26 18:27:55 router01-fpc2 clockd[14873]: EMF_EVO_ALARM_CLR: Alarm cleared: CG color=red, class=CHASSIS, reason=FPC 2 zl30642 PLL Access Failure
Sep 26 18:28:05 router01 alarmd[13780]: Alarm set: FPC id=1224737384, color=RED, class=CHASSIS, reason=FPC 2 PLL Access Failure
Sep 26 18:28:05 router01 craftd[12867]: Receive FX craftd set alarm message: color: 1 class: 100 object: 104 slot: 2 silent: 0 short_reason=FPC 2 PLL Access Failure long_reason=FPC 2 PLL Access Failure id=1224737384 reason=1224736768

The issue should be fixed in PR/1696511 - resolved in evo:23.1R1-EVO junos:22.4R3-J4.3

The CLI command - show clockd zarlink ref-info, or FPC shell command - "show clksync ref-info"  could trigger the issue. 
However, it is not limited to them. The internal daemons could also trigger the issue silently.
Not only ACX7908 platform, but the issue was also observed in the other platform, such as MX10008 with LC9600 card.