PHD2 Polar Alignment issues

[lazjen]

I use PHD2 for guiding but had not got around to seriously trying the polar alignment feature that it provides - drift align - until last night. I found a tutorial to try (https://sites.google.com/site/openphdguiding/phd2-drift-alignment) and over time I manage to get improvements - I think.

I could not get the red line to flatten very far at all, although the magenta circle became quite small. No matter what I did, the result was such that the RA was still jumping all over the place when guiding (+/- 2" easily).

I'm trying to work out why that was happening. I suspect the seeing has something to do with it, but trying with 1,3,5 sec guiding didn't seem to make much difference.

Would backlash cause this? What about the aggressiveness setting - I did have it at 100?

Finally, DEC seemed to be quite good compared to RA - hardly any movement.

My setup for this is as per my sig. I was at 2000mm - no reducer.

Thanks for any suggestions/help.

EDIT: I guess I should ask what information I could attempt to gather tonight that might assist in working this out? Is there anything I could try to eliminate some variables?

[codemonkey]

Did you ensure you were pointing at the correct parts of the sky when doing this? A mistake I made which caused me some frustration. It's very important to point to the correct locations.

[lazjen]

Yes, I'm pretty sure. Although I did stuff around for a long time on this, so perhaps I moved away from the optimum area.

[RobF]

Hi Chris, apologies if I'm stating the obvious, but you realise there is a different coloured line for Dec and RA? RA should show some noise, but also cycle over the worm reflecting the periodic error of your mount?

Dec gears have PE too, but generally once your aligned it shouldn't matter. As you get closer and closer need to drift longer to check it is correct.


Might be helpful to post up an image of what you're seeing if possible?
Others could then comment if seems normal for your gear.

[lazjen]

I should have captured some screenshots last night, but I was a bit too brain fried to remember. I'll capture stuff tonight if I don't get it sorted.

As for the different coloured lines - yes that applies for the general guiding, but if you look at the tutorial page I linked, it uses the "red" line for the slope in the drift align process. You're suppose to "flatten" the line in the process and I was able to do so up to a point.

Periodic error will be in there since I don't have PEC enabled. I'll have to look up the value, but the mess I was seeing was quite frequent. Maybe it was having more of an effect than I thought.

[lazjen]

I'm doing the drift align again now and I'm starting to think the periodic error is what I'm seeing.

189082

I did drop the aggressiveness down to 70 from 100. I can't remember why I had it on 100 as I normally try for less.

I'll try to do a small adjustment to see if I can reduce the error further.

However, the RA line looks periodic.

Also, I'm pointing west, so I'll have to check east later. I have a suspicion the scope balance is affecting thing too.

[lazjen]

There's definitely something wrong when I'm pointing east.

The first image was the last drift run in the west before flipping over. The second is the best east run so far - no azimuth adjustments made. All I've done so far is up the aggressiveness to 95 and balanced the weights towards the west - that seemed to help a bit. The variability though is a lot more in the east.

189093189094

[troypiggo]

When drift aligning, you completely ignore the RA errors. Just watch the red DEC trendline. Doesn't matter at all what the purple RA is doing until you've finished drift aligning, have calibrated, and are actually guiding. You will see it flatten right out then.

I was doing some testing the other night, and I found it much easier to follow with the graph displaying pixels instead of arcseconds. Go into the settings button of the graph there, and choose pixels. Might be just the way I'm wired, but I found it easier, and the graph wasn't as spikey.

[lazjen]

Ok, I still thought the other line was showing how much "effort" was being done. I didn't notice that much difference to the RA line when I went to the normal guiding.

I found rebalancing helped a lot - and it was to the east not west as I wrote previously.

Still, I have to say the PHD2 drift align procedure seems to give good results and relatively quickly too. Following the process is easy and the feedback quick.

[Rod771]

Hi Chris

I'm pretty sure PE cycles through every 8 minutes. Looking at your first graph your run appears to roughly finish around the 5 min mark? That would rule out PE as your major problem because you wouldn't see the RA graph rise and fall that many times in 5 mins due to PE.

I suspect seeing is your major problem, as it is for us all. :rolleyes: Try a higher exposure time of 2.5 to 3 sec and increase the RA min move so you're not sending guide commands every exposure. This may help settle the RA down a bit.

[bugeater]

I find getting the line nice and horizontal to be quite difficult too. It seems to change its trend periodically. Not sure if it is wind or something else.

But periodic error is a red herring - it can't be that as the line you are watching is the dec axis, not RA

[lazjen]

Thanks for the reply, Rod.

I've just come back from an unsuccessful dark site visit (weather, technical issues, human issues *sigh*), so I'll have to set up again in the observatory and go through the alignment process. I'll have a better go at sorting it out this time.

If seeing is the issue (it has to be part of the problem), wouldn't that also affect the Dec graph too? As in I should see the Dec vary as much too and therefore be able to use this as a test?

And I'll try the changes you suggest. Any idea what a good min move value would be?

[lazjen]

The pics I attached previously were during the Drift Align tool runs. It specifically uses the "red" Dec line each time as line that you need to make as horizontal as possible (no matter whether you are adjusting altitude or azimuth). Of course, that just makes things more confusing. :)

[Rod771]

Yeah , good point. One thing you can do if your using the latest version of PHD2 (2.5.0) is, after calibration run the guiding assistant which is on the Tools tab. What it does is disables guiding in both axis and measures the data then makes suggestions base on how the star is moving when unguided. This will give you an idea of how good or bad the seeing actually is. It also shows how close your PA is and, what I have found just tonight is, it helps you set your balance in RA. My RA graph had never been as good as my Dec, similar to you screen shot. After watching the RA graph rise and keep rising while the guiding assistant was enabled, I figured my balance was too far off. I just stopped the guide assistant, adjusted the counter weight, restarted the assistant and the RA graph leveled.

So along with seeing, balance could be affecting your RA guiding as it was mine.

Hope this helps. I know how it feels, stick with it. :thumbsup:

[lazjen]

Thanks Rod. What I like about PHD2 is the (in general) quick feedback it gives. I think I'm using 2.5.0, so I'll give the guide assistant a go as well.

Now I'm itching to get it all going again, but that's probably going to be Sunday night if I'm lucky. At least there's a few things to try.

[Camelopardalis]

This isn't a helpful comment particularly, but the drift alignment function gives me headaches :sadeyes:

Before getting any imaging done last night I checked my drift by using the DARV function in APT and by taking unguided exposures which resulted in round stars. However, PHD drift alignment showed a downward graph for Dec and guide assistant claimed my alignment was 6 arc minutes from the pole...

[bugeater]

It shouldn't be confusing. Dec drift is the indicator of not being perfectly polar aligned. Hence you adjust the appropriate axis to eliminate this drift. Which adjustment (alt or az) depends on which part of the sky you are pointing at. What makes it difficult is there seems to be other factors at play (but not PE :))

[lazjen]

Actually, I think the comment is (partially) warranted Dunk. I ran out of time before I had to pack everything up for my trip, so I never got a chance to try verifying that the process worked correctly.

Before I started using PHD2's tool, I had used Astrotortilla to do my previous alignments. While it worked ok, I wasn't completely happy with the results and the variability with repeat runs. So, on to PHD2.

I hope to have everything set up again tomorrow, so I'll repeat the PHD2 process again from scratch and report results, etc (assuming I sort out my other issues). I'll use the guide assistant tool Rod mentioned to help verify the results.

[lazjen]

PE can still be part of the problem, just like the seeing will be. It's just that I think we can rule out both as being the major component of the RA problem.

With my next set up I'll be able to leave it in place for a long time, so I'll be able to work through the issue(s) in less of a rush.

[Slawomir]

There could also be focuser/telescope flexing at various angles.

[lazjen]

I would assume flexing would be caused by problems with weight at these "various angles"? Wouldn't I also see the problem in Dec as well then (which I do not seem to have)?

And how realistic is it to have flexure in a RC10? Or a moonlite focuser?

I should clarify, that except for removing the 0.67 reducer (so it was at approx 1340mm), the set up was the same that I was using for taking 30 min subs before. I wanted to refine the PA further for the change to the 2000 mm, which I thought would be more sensitive to any inaccuracy I had.

[bugeater]

If we are talking drift alignment, then you shouldn't pay attention to the RA chart.
If we are talking guiding in the RA axis, then things like PE, seeing and wind will be involved.

I'm envious that you have a permanent setup. It'll certainly reduce the alignment headache once you get it right.

[Slawomir]

Fair enough. I might have gotten confused what the issue is in here. From the screenshots you have attached I saw a drift in Dec (red line) indicating either inaccurate alignment or flex under gravity (could be mirror, tubes, focuser...).

At 2000mm any flexture (something not being 100% rigid) will be more noticeable.

Just throwing some ideas into the basket....

[vlazg]

I also have trouble with PHD2 drift alignment tool.
What about EQ Align, doesn't seemed to be mentioned on the forum.
Haven't used it but it seems a simpler process, a bit more accurate than Alignmaster.

[lazjen]

With the permanent setup, the wind doesn't affect the mount at all, thankfully. And it got to the stage that if I couldn't have a (semi)permanent setup I think I would have had to pack it all in and give up. I just could not handle the setup/teardown all the time.

I've got everything set up again in the obs, so if tonight has any clear skies I'll start the process again.



No worries. I have used 2000mm in the past when I was using my DSLR, although I think the maximum guided images I took were 10 min. From memory, the subs were ok, but not great. I'd like to be able to get good images at 30 min, if possible.



The main reason I haven't tried it is that I don't have a camera it supports, I think. Just checked: Ok, it supports QHY5, hopefully that means QHY5Lii, so I might try that as well.