Phd2 orientation issues

[TheMan]

I started using my other computer to guide last night, I connected the drivers and everything looked smooth until I tried polar drift alignment feature. When I adjusted the altitude knob on the wedge the stars would go left or right. The stars would go up and down also if I moved the azimuth knobs. I could easily put up with that usually, but I have to move the stars in the opposite direction to the red line otherwise i gt further from being polar aligned! I use a knock off zwo asi 120mc and a skywatcher az-gti in ew mode. Any help is appreciated.

[Anthonyexmouth]

the direction the star moves isn't important. just follow the instructions. move one way it gets better and the other it gets worse,  it's not about up/down or left/right in relation to the red line its only about moving the Lat and Az  when told to . 

[almcl]

2 hours ago, TheMan said:

but I have to move the stars in the opposite direction to the red line otherwise i gt further from being polar aligned! 

I have had exactly the same experience: moving the star in the direction indicated increases the error.  Eventually, despite help from the author, who posts here, I gave up using it.  Apparently it requires several minutes of tracking to come up with an accurate correction, but even giving it 5 minutes didn't didn't improve matters.

[symmetal]

The direction the star moves when you adjust the azimuth or altitude bolts depends on the orientation of the guide camera in the guide scope so as Anthonyexmouth says don't worry about what direction it moves when you adjust the alignment bolts.

On your first drift test make a note of the angle of the red Dec line to the horizontal.  Then turn the bolt (az or alt as instructed) in one direction (at this point you don't know which direction will make it better so randomly choose one) until the guide star reaches the edge of the purple circle shown around the guide star, making note of the actual direction you turned the bolt. The direction the star actually moves in doesn't matter.

Then do another drift align. If the red Dec line angle is in the same direction but worse than before then you turned the bolt initially in the wrong direction so turn the bolt in the opposite direction to reach the purple circle edge again. This will be the opposite side of the circle than you moved to the first time.

If the Dec line is in the same direction but closer to horizontal then you moved the bolt in the right direction but not enough.

If the Dec line has changed direction then you moved the bolt in the right direction but too much. 

Repeat the drift align procedure until the Dec line is horizontal, or pretty close to horizontal.

Don't be too critical on your first try to get a horizontal line as the Az and Alt adjustments will tend to affect each other. Do an az adjustment until it's better than it was (not necessarily exact )then do an alt drift align and again adjust so it's better than before. Go back and forth making small adjustments to az and alt until both lines are pretty horizontal.

You just need to remember what direction you turned the bolts each time so you know what direction to turn on the next go to make it better. 

If you keep a note of what direction you turned a bolt to make the Dec line go in a known direction, the next time you do a drift align you don't have to do the initial random turn to find out.

Alan

[almcl]

That's interesting Alan, but the Polar Drift Align routine in PHD2 (which is what I thought the OP and I were referring to) doesn't work quite like that.  

There is no purple circle, just a line along which the chosen star must be moved by  alt/az bolt adjustment.  The direction can be determined quite quickly  as the star either moves towards its indicated target position or in the opposite direction.

The problem is that after this, a further run invariably reports a bigger polar alignment error than before, even with quite small adjustments.   This is quite a problem for those of us with OAGs (so no possibilty of using Sharpcap's polar align) and no east or west horizon for conventional  drift align. 

Kens' Polar Drift Align routine would be a blessing, but I can't get it to work here in the northern hemisphere.

 

[Anthonyexmouth]

11 minutes ago, almcl said:

That's interesting Alan, but the Polar Drift Align routine in PHD2 (which is what I thought the OP and I were referring to) doesn't work quite like that.  

There is no purple circle, just a line along which the chosen star must be moved by  alt/az bolt adjustment.  The direction can be determined quite quickly  as the star either move towards its indicated target position or in the opposite direction.

The problem is that after this, a further run invariably reports a bigger polar alignment error than before, even with quite small adjustments.   This is quite a problem for those of us with OAGs (so no possibilty of using Sharpcap's polar align) and no east or west horizon for conventional  drift align. 

Kens' Polar Drift Align routine would be a blessing, but I can't get it to work here in the northern hemisphere.

 

Hi, bit off topic, but what makes OAG and sharp cap incompatible? 

 

[almcl]

22 minutes ago, Anthonyexmouth said:

Hi, bit off topic, but what makes OAG and sharp cap incompatible? 

 

Sharpcap's field of view requirement can't cope with a 909 mm focal length. At least in version 2.9.

[Anthonyexmouth]

1 minute ago, almcl said:

Sharpcap's field of view requirement can't cope with a 909 mm focal length. At least in version 2.9.

ah yeah, i remember it saying it needed a certain FOV. Was sure i've used it with my mak90, but maybe im mistaken. 

[symmetal]

6 hours ago, almcl said:

That's interesting Alan, but the Polar Drift Align routine in PHD2 (which is what I thought the OP and I were referring to) doesn't work quite like that.  

There is no purple circle, just a line along which the chosen star must be moved by  alt/az bolt adjustment.  The direction can be determined quite quickly  as the star either moves towards its indicated target position or in the opposite direction.

Sorry almd and TheMan, I didn't notice the 'Polar' in your posts. My description refers to the 'standard' drift align feature in PHD2.  I haven't used the Polar version so can't help. I'll go away. 

Alan

[kens]

Make sure your mount is tracking when you do the alignment. This is the most common problem with PDA. I'd like to get to the bottom of these issues so if you can upload your Debug log that would be helpful. 

[TheMan]

30 minutes ago, kens said:

upload your Debug log

where about are the log files?

 

[almcl]

33 minutes ago, kens said:

I'd like to get to the bottom of these issues so if you can upload your Debug log that would be helpful. 

Next clear night we get, Kens, I'll give it a run and post the log.  Would you like it here or on the PHD forum? 

And thanks for taking the time.

[TheMan]

1 minute ago, almcl said:

Would you like it here or on the PHD forum? 

Here works fine. I don't have an account for PHD forum.

[TheMan]

PHD2 Debug Logs.zipI'm pretty sure theis is the right one.

[kens]

Thanks. At first I thought the problem was the hemisphere setting till I realised you are in Melbourne like me  So I know the tool is thoroughly tested in the southern hemisphere.

What is odd is that when you were ding the polar drift the mount was reporting a declination of close to zero rather than close to -90. That is not necessarily a problem as it could just be a sync issue but you do need to be pointing towards the pole. Also make sure the mount is tracking (not parked). The video tutorials here should be relevant since I made them from Melbourne. The star pattern near the pole is quite distinctive.

[almcl]

Ken

I think I cracked the problem tonight.

I've uploaded the guide and debug logs to the PHD server here:

https://openphdguiding.org/logs/dl/PHD2_logs_68mT.zip

but what I discovered is that while PDA is accurately measuring and reporting the error in PA, the red line that comes up on the screen is 180 degrees out in its orientation.  For my first three corrections I followed the instruction and moved the star towards the red circle.  On each subsequent run a greater error was reported.  This was my experience previously as well.  So after the third run, I tried moving the star in the opposite direction to that indicated.  In two goes I had the error down to ~1'.

I am not sure how, or even if, this is possible, but is there a way to make the red line display the opposite way round?

[kens]

I can make the red line go anywhere I want   But I do need to verify some things first. One is that the scope was tracking during he polar drift (at face value this would appear to be the case) and the other is that the PA error from an alternative method is consistent with the PDA result. 

Tracking is important because near the pole the RA drift when tracking is off is very small and mimics the PA drift very closely. That fact that you slewed from your calibration point to the pole would normally cause EQMOD to start tracking so that looks promising. Also, measuring the PA drift on your guided runs shows the drift was much less after your PDA run. This is not conclusive as it depends on where you are pointing in the sky but promising nevertheless.. A Sharpcap, KStars or SPA alignment would confirm the result.

Getting the direction right is mind bogglingly difficult due to the multitude of coordinate systems one has to deal with in the calculations. So its quite possible that the northern hemisphere direction is wrong. I'm confident about the southern hemisphere result is correct since I was able to confirm that myself.

Another possible problem is the use of an OAG. The prism reverses the image orientation which might affect the calculations. Basically, the speed of the drift determines how far out the PA is and the direction to the pole is at right angles to the drift. In the NH the drift is counterclockwise and in the SH it is clockwise (or maybe the other way - I can't recall). But in the mirror image caused by an OAG the direction could be reversed. I'll need to go over the maths and also try it out myself with an OAG if the clouds ever clear. 

If it turns out to be due to the OAG then I'll need to include an option for the user to indicate the image is reversed.

If you could verify the PA error after PDA that would be useful. Thanks for the data - it was quite clean

[almcl]

Thanks, Ken, that all makes good sense (and I take my hat off to anyone who can figure out the northern/southern hemisphere conundrum - it was months after my visit to NZ that I finally understood why Orion appears the other way up!)

The change in PA error was confirmed by the guiding stats in PHD log viewer - I know that's not independent corroboration - and I did wonder if the combination of Newtonian/OAG (3 reflections) might have an influence, so next time we have a clear night I'll bring my ST80 guidescope out of retirement and see if we can get PDA and Sharpcap to agree  - at least about direction of movement.  

The OP doesn't say what his equipment is, but in his first post he mentions having to move in the 'opposite direction'. As he is in the southern hemisphere, my money is on the mirror(s) being the source of the issue.

[TheMan]

9 hours ago, almcl said:

my money is on the mirror(s) being the source of the issue.

Sorry about not mentioning my equipment before. I use a lens as a guide scope and a knock off asi 120mc

[kens]

3 hours ago, TheMan said:

Sorry about not mentioning my equipment before. I use a lens as a guide scope and a knock off asi 120mc

Yours is a different problem. Were you pointing at the South Celestial Pole? According to the log you were pointing near the Celestial Equator (declination zero).

If that's the case then PDA won't work as you need to be pointing within 5 degrees of the pole.

[TheMan]

21 hours ago, kens said:

Were you pointing at the South Celestial Pole?

I didn't do a star alignment so PHD didn't know where it was pointing, it was indeed pointing right at it.

But I found a solution last night. I pointed it at the south pole, did the drift alignment and moved the stars in the opposite direction until I was about 30 arc minutes from being polar aligned. I then disconnected the mount and switched the hemisphere to north. It worked amazingly and was able to get within 2 arc minutes of being polar aligned. (In case you were wondering, yes I did check it over with hemisphere set to south). 

[kens]

Glad it worked for you but it sounds very odd. The only things I can think of that would cause that are some sort of mirroring of the image or the mount not tracking. The lens (I assume you mean a camera lens) should not do that but it is  possibility. Also, since you are using a knock-off camera there is the possibility that it is producing mirrored images (e.g. by returning the rows in reverse order).

Can you provide a screenshot or saved image from PHD2 centred around the pole? The asterism formed by BQ Oct, HD99685 and HD98784 is quite distinctive and will show me if the image is inverted.  If so there's more reason to add an option to PDA to allow for an image mirrored by either the lens, the driver or an OAG.

[TheMan]

3 hours ago, kens said:

mirroring of the image or the mount not tracking

This might be true, but I have been successfully been using PDA on two other computers using the exact same techniques and drivers