I'am having some major guiding problems whit my belt drive eq-5. Its driven by astroeq controller.
I have sorted some things out. I had difficulties whit 2.6.2 and my altair gpcam. But that was fixed when i copyed new altair drivers from sharpcap to phd2.
I have done some testing on my mount and found out that the manually guided north of phd2 is south on my eqmod ascom interface. Should it be like this? To clarify when i press manually north on phd2, dec axis is movin one direction. When i press north on eqmod ascom interface the dec axis goes to the other way.
Phd2 allwaya tells me that the calibration is done on west side of the pier, eaven when eqmod ascom says its on east side of the pier. Eny suggestkon how to fix this problem?
Hi Jukka. I've embedded some answers below:
-----Original Message-----
From: open-phd...@googlegroups.com
[mailto:open-phd...@googlegroups.com] On Behalf Of Jukka Paavilainen
Sent: Sunday, September 04, 2016 2:05 PM
To: Open PHD Guiding
Subject: [open-phd-guiding] Phd pier side and information about manual guide
direction
Hello.
Greetings fron finland.
I'am having some major guiding problems whit my belt drive eq-5. Its driven by astroeq controller.
I have sorted some things out. I had difficulties whit 2.6.2 and my altair gpcam. But that was fixed when i copyed new altair drivers from sharpcap to phd2.
I have done some testing on my mount and found out that the manually guided north of phd2 is south on my eqmod ascom interface. Should it be like this? To clarify when i press manually north on phd2, dec axis is movin one direction. When i press north on eqmod ascom interface the dec axis goes to the other way.
PHD2 doesn’t really know or care which way the axis rotates as long as it’s consistent. The PHD2 command is sent to the ASCOM driver to “move north”, for example, and it’s up to the driver to do the right thing. Most mounts and drivers move the axis in the same direction for declination regardless of side of pier. But some mounts/drivers choose to reverse the direction of rotation when the pier side changes. To handle this, PHD2 has a control in the ‘Guiding’ tab of the advanced setup dialog that lets you handle this situation:
If you find that guiding is going well on one side of the pier but then the declination guiding goes in the wrong direction after a meridian flip, you should check this box. You only have to do this once, just set it and it will be saved as part of your equipment profile.
Phd2 allwaya tells me that the calibration is done on west side of the pier, eaven when eqmod ascom says its on east side of the pier. Eny suggestkon how to fix this problem?
PHD2 simply reads the side of pier value from the ASCOM driver – that’s it. If you’re looking at the side of pier value in one of the calibration dialogs, that shows where the calibration was originally done, not how the scope is currently oriented. When PHD2 automatically adjusts the calibration data after a meridian flip, this UI field won’t change. What are you looking at that makes you think there’s a problem? If you’re convinced something is wrong, you can send us the debug log file and we can see what we’re getting back from your mount. Or if you’re having trouble with guiding after a meridian flip, send us both the guide log and the debug log and we’ll try to help you figure out what’s wrong.
Good luck,
Bruce
--
You received this message because you are subscribed to the Google Groups "Open PHD Guiding" group.
To unsubscribe from this group and stop receiving emails from it, send an email to open-phd-guidi...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Long story short(well it was not so sort):
I had problem that my dec guiding went all of a sudden like 8" wrong.
I had problem whit my mount. I tryed to get rid of all backslash by tightening wormwheel. But when it was tight my mount continued the previous corrected way for litle while. Like if phd corrected south after correcting north, the dec-axis moved a bit north before going south. My mesurement's were in sub millimeters
I got accurate way of mesuring that whit mesurement clock(dont know the english word for that)
I struglet whit that for week, before i understud it.
I was so keen to look my guide graph that i had the phd guide image in really small window. The notification about camera not working properly was set to 15sec. So it did not alert me. I realized that i got one frame in every 10-12sec. And that might be the reason to dec-axis moving so much all of a sudden.
It has been rainy in here so i have had no opportunity to test this.
Here are my guide graph and debug log's:
https://drive.google.com/folderview?id=0B24ECy12UBmLa0hqR3hydTJWSEE