here's a typical sequence of events, let's pick the first one at 18:43 (green arrow below)
after a period of typical guiding, the mount returned control sooner than expected:
18:43:54.835 00.001 12808 MoveAxis(W, 6, ABG)
18:43:54.835 00.000 12808 Guiding Dir = 3, Dur = 6
18:43:54.835 00.000 12808 IsGuiding returns 0
18:43:54.835 00.000 12808 PulseGuide returned control before completion, sleep 16
this continues with the Dec
18:43:54.860 00.000 12808 MoveAxis(N, 126, ABG)
18:43:54.860 00.000 12808 Guiding Dir = 0, Dur = 126
18:43:54.861 00.001 12808 IsGuiding returns 0
18:43:54.861 00.000 12808 PulseGuide returned control before completion, sleep 136
after this, the mount appears hung, reporting it's still trying to complete a move, which goes on until you restart the mount,
18:44:03.377 00.007 12808 IsGuiding returns 1
18:44:03.377 00.000 12808 Still moving
(many dozens of similar messages like this follow)
at this point any subsequent guiding is effectively disabled, as you can see:
The problem is not within PHD, it's somewhere within the software or the mount connection to the software. I would check the logs for GSS and see what it may tell you about this problem.
The behavior seems pretty consistent and repeating, and restarting seems to resolve it, so that suggests to me a software issue rather than hardware/cabling. If it's specific to certain angles, it could be a cable snag or other issue (the gss logs would hopefully uncover this).
On a minor point, consider creating and using a dark library for your guide camera
Brian