8 replies to this topic

[kathyastro]

I seem to have got dew on my guide scope.  (At least that's the only thing that makes sense.)  PHD2 reports "Star Lost", NINA figures it will carry on anyway, and maybe my images are only mediocre. Should be no big deal.

 

Instead, PHD2 goes nuts and starts driving the scope all over the sky, causing several frames to be total garbage.  Eventually, NINA did a meridian flip and everything restarted, but the mount was nowhere near where it was supposed to be, and the plate solve failed.  So the rest of the session was wasted.

 

So my questions are: why did PHD2 go nuts on me?  If it can't see stars why not just sit there doing nothing until they come back?  What settings can I change to avoid this?

 

 

[bobzeq25]

Do you use darks to avoid PhD2 from trying to guide on hot pixels?

Is your cable management good?

Are your cables good? Swapping out cables is kinda like turning your computer off and on to fix something.

[ButterFly]

You should post your logs.  It's likely that something else happened here.  PHD2 shouldn't be making corrections when it loses a star, by design.  Without a guide star, it has no way of knowing what the correction should be.  From the manual:

 

 

It's important to understand that PHD2 will not start moving the telescope trying to relocate the guide star.  It will simply continue to take exposures and look for a guide star to reappear within the bounds of the current tracking rectangle.   If the lost star condition persists for a considerable time, your mount will probably drift signiificantly off-target.  In that event, a later recovery of a guide star in the tracking region is likely to be a different star and your imaging target won't be properly centered.  If you're doing unattended imaging, the imaging application should handle this situation by recentering the frame after star recovery has occurred - this is not something that can be done by PHD2.  Isolated or intermittent lost-star events do not usually cause poor guiding or ruined images.  But extended periods of lost-star events can allow the mount to go too long without guiding corrections and the image quality may suffer.

 

It looks as though your mount started going nuts before there were any lost star warnings in that envelope just to the left of where the last star warnings begin.  The lost star warnings probably came from the selected guide star going off chip, then acquiring another that also went off chip, ... .

[ButterFly]

 

Instead, PHD2 goes nuts and starts driving the scope all over the sky, causing several frames to be total garbage.  Eventually, NINA did a meridian flip and everything restarted, but the mount was nowhere near where it was supposed to be, and the plate solve failed.  So the rest of the session was wasted.

PHD2 doesn't drive all around the sky.  It only issues guide commands.  Based on these data and nothing else, it seems like a loose clutch.  Of the frames that weren't "total garbage", could you get solutions on astrometry.net?  That can solve even "mostly" garbage.

[michael8554]

Hi Kathy

 

You haven't displayed the camera output on the graph, and both axis are very coarse, so it's not clear what is happening.

A GuideLog would be better.

 

When PHD2 detects "Star Lost" it stops guiding and carries on tracking.

And resumes guiding when a star reappears in the Search Region.

PHD2 was issuing many Fast Slew commands to re centre.

 

It appears that the star was being lost fleetingly, based on the "Minimum star HFD for Autofind = 6".

That to me means only stars larger than HFD=6 will be auto found, your star's HFD will typically be 3 to 4 when well focussed !

 

Later on there are no "Star Lost Messages", but PHD2 may be locking onto random hot pixels.

Perhaps Darks rather than Defect Map would be better.

Just guessing really, need the GuideLog.

[kathyastro]

Hi Kathy

 

You haven't displayed the camera output on the graph, and both axis are very coarse, so it's not clear what is happening.

A GuideLog would be better.

 

When PHD2 detects "Star Lost" it stops guiding and carries on tracking.

And resumes guiding when a star reappears in the Search Region.

PHD2 was issuing many Fast Slew commands to re centre.

 

It appears that the star was being lost fleetingly, based on the "Minimum star HFD for Autofind = 6".

That to me means only stars larger than HFD=6 will be auto found, your star's HFD will typically be 3 to 4 when well focussed !

 

Later on there are no "Star Lost Messages", but PHD2 may be locking onto random hot pixels.

Perhaps Darks rather than Defect Map would be better.

Just guessing really, need the GuideLog.

Here is a zoomed-in look at the graph.  Both the camera data and the guiding outputs are displayed on the graph.

 

 

If I set the minimum star size manually (I don't remember now), it would likely have been to avoid guiding on hot pixels. 

 

Whatever PHD was seeing, it thought it was moving and was trying unsuccessfully to follow.  In fact nothing appears to have been moving (I checked my mount for any kind of flexure) and what it was doing was driving the scope off-target.

 

I have seen it have hallucinations before.  One time, it spent several hours "successfully" guiding while the dome was closed.  I'll try shooting new darks for the guider and maybe fiddle with the minimum star size.

 

Here is the guiding log.  The relevant portion starts around 2024-08-13 00:22, which is frame 1553.  That is where the graph starts to get wiggly.  The serious deviations start at 00:27, frame 1626.

 

https://www.dropbox....t=no5umgv4&dl=0

[KTAZ]

Kathy, I'm a fan of Occam...cables or port extenders are my first suspect when I see what could be communication glitches between cameras, mounts, and computers.

 

Funny things happen when bits and bytes start to go missing...

Edited by KTAZ, 16 August 2024 - 10:09 AM.

[Forward Scatter]

I also think there may be a mount-side tracking issue caused either by cable drag, board issue (loose connection or flat out failure), low voltage/amperage to the mount, balance, bad cables, loose clutch/gears, and my repeated favorite of a Windows update messing with the PC time sync and location.

 

PHD2 is just reporting the issue and working overtime to correct the tracking errors.

 

As the dec and RA pulse are symmetrical, maybe the mount thinks it's in the southern hemisphere, aka 45^o S instead of 45^o N.

Edited by Forward Scatter, 16 August 2024 - 11:12 AM.

[michael8554]

From the GuideLog:

 

"The relevant portion starts around 2024-08-13 00:22, which is frame 1553."

I see no problems there.

 

At 03:50:44 frame 2022 the camera output dropped to zero and never returned, there was no intermittancy to the Star Loss.

 

PA error was 12arcmins, so probably the guide star drifted away.