This is clearly not an ACP issue, but I'm posting here on the off chance that other power users might have some troubleshooting pointers.

I'm experiencing an identical and intermittent problem on two different telescopes when Scheduler runs the first autofocus of the night (at Sun -12).

Hardware - Paramounts via ASCOM TheSky driver - current version (both telescopes).
Focuser - FocusMax (latest version) on one of the telescopes. PWI (latest version) on the other.
ACPS - latest versions.

Briefly - Scheduler runs StartupObs, successfully connecting the telescope and camera, and setting the CCD temp. It then waits for Sun -12.

On an increasing number of occasions, the first autofocus will fail with:

10:32:58 Start slew to FocusArea...
10:33:04 (wait for slew to complete)
10:33:05 **Script Error (Tracking has been stopped)**
10:33:05 Source: System.Dynamic
10:33:05 Message: Object reference not set to an instance of an object.
10:33:06 [Source: System.Dynamic]
10:33:06 [Object reference not set to an instance of an object.]
10:33:06 [This is not an ACP problem]

Checking TheSky does in fact show a problem, with the telescope disconnected, requiring manual intervention. If I exit ACPS, reconnect the telescope in TheSky, home, park, start ACPS and enable the dispatcher again, the cycle will complete and the night will commence with an autofocus and proceed normally from then on - as if there was never a problem to begin with.

In other words, identical start state but it always works the second time.

I don't believe it is horizon limit related. TheSky's horizon is set to zero, ACP to 20 degrees, and the park position at ~ 22 degrees. FMx has no problem returning to the previous position, which is essentially the park position, after focus. Plus, it works the second time.

I'm wondering if there is a 'timing issue' that manifests itself after hours (or days) with telescope park position and slew limit so close together in altitude. It seems to sort it itself out once the telescope is reconnected, so perhaps a 'home on connect' might be warranted?

The other possibility is a comms error - these telescopes have long serial cable runs. But why this doesn't bother the telescope for the rest of the night is of note.

It's that not entirely uncommon "I need to hand hold things as they start up, but once running everything is fine" kind of experience.

Obviously there are a few things I can try, so I'm just posting here on the off chance that this type of issue has been seen before.

Any other ideas gratefully received.