PHD2 Drift Polar Alignment Tool for First Time Users
This video is for beginning astrophotographers who need really good polar alignment and can’t see Polaris from their shooting location.
The video ended up being almost an hour and a half long and I get that not everyone is going to sit through that, as educational and entertaining as watching a red line on your screen can be. So I posted my script for the video here as a blog post for those that would prefer to read rather than watch.
Why I made this video:
I had to be able to polar align without Polaris because of my need to take photos from my porch during Covid19 “stay the hell at home” period.
Drift Alignment was my only solution if I wanted to get some photos while stuck at home.
Problem was I had tried a number of times before and only had MASSIVE failure! Because I had no idea:
What the expected result of moving the mount controls would be in both in magnitude and reaction.
I was totally unaware of how long to expect this process to take in all, and how long to get a result from your adjustments.
This dual layer of uncertainty made it initially impossible to figure out why it wasn’t working.
Well now I had to make it work. So I looked on YouTube and found a video “Polar Alignment Without Polaris” in which the guy showed how you could get a good polar alignment using a combination of a smartphone app and the Drift Align routine in PHD2.
Finally, I had actually seen this work and that was huge but the problem was that this guy’s mount was actually pretty close to polar alignment already, but I didn’t have that luxury. I wanted to make it work from not very closely aligned at all.
So I looked at a couple more PHD2 drift align videos and picked up some tips, and the must successful tip was: In order to do this well and get predictable results you have to be systematic about your approach and methodology. A number of the things that you do with this method will actually work backwards from the way you expect if you change the way you do things and you’ll end up getting confused. So you need to be consistent about some choices you make in this procedure that are essentially arbitrary-you just need to be consistent about them.
Remember, drift alignment is very accurate so it stands to reason that at some point, tiny changes will have to be made to achieve that precision. This need for consistency is greatly aided by taking advantage of certain “machines” (both software and hardware) so I’ll detail my setup that will help you with consistency.
There’s the mount itself which is under control by a computer of some kind. The ASCOM method for talking to all our astro stuff is really good. It only runs on Windows although there is a movement afoot to make it work natively on a Mac/Linux called “ASCOM Alpaca”.
The guide camera looking through it’s own scope or the main scope via an “off axis guider”
I run three software programs while I’m getting this all together:
PHD2 of course, connected to the guide cam and mount via USB (cam) & ASCOM (mount).
A sky map program like Stellarium, Starry Night, Sly Safari or whatever that is capable of talking directly to the mount so it can show you where the telescope thinks it’s pointed.
Your main camera imaging program–Sequence Generator Pro or whatever:
Early steps:
Level Your mount
The early steps can be made less stressful by having a properly aligned finder scope on your scope besides the guide scope. You have to get the mount roughly aligned with the sky before you start drift align, so:
Start up your mount in the start position, Losmandy = CounterWeight Down but many mounts start with the weights down and the scope pointed at the N. Celestial Pole.
Make sure your mount knows it’s Long/Lat by entering the numbers manually if you have to, and give it the exact time to the second from your cell phone.
Tell the mount to GoTo a bright star that might be peeking through the dusk. Lately it’s been Sirius or Procyon for me. It won’t be in your main camera unless you’re really lucky
Look in your Finder scope and without using the mount’s hand control at all adjust the Azimuth (left/right) and Altitude (up/down) controls on your mount to move the star you tried to find into the finder and center on the crosshairs with a reasonable degree of accuracy but you don’t have to be super accurate because it’s all going to get updated in a minute.
If you’re really far off don’t be afraid to move the tripod legs first to get in the ballpark before reaching for the mount adjustments.
Check in your imaging program to see if the star is present. Your exposure will be ridiculously short, like .005 sec or .01 sec. You should also be roughly in focus at this point.
NOW use the hand control to slowly move the star to the center of your imaging cameras screen while in the “Frame and Focus” mode. Sequence Generator Pro has crosshairs, turn them on and get that star right in the center.
Go to your mount and now update your alignment on the hand controller. The mount should be tracking the sky at this point.
Now go to PHD2 and start looping exposures from your guide cam; hopefully the star will be in there somewhere. Short exposures again. If necessary adjust the screws on your guidescope to get the selected star in the center of the frame as exactly as you can (I can’t find any way to show crosshairs in PHD2 so do the best you can). You might have to turn up the brightness some to see the edge of your camera’s frame. There are huge benefits from having your guide cam pointed exactly the same as your main scope.
Finally, we’re ready to begin actual polar alignment.
Drift Align Concepts
So let’s get to it! Here are the basic ideas.
You’re going to visit a couple of special places in the sky and make a series of increasingly smaller and smaller adjustments to your mount’s pointing using the mount controls, not the hand controller.
Spot number one is the intersection of the Meridian and the Celestial Equator. We use this spot to adjust the mount’s East/West pointing, again called Azimuth. We will find a star near this point and, using the Drift Align tool in PHD2 will make a series of adjustments to the mounts E/W pointing until PHD2 shows us that we’re good by displaying a flat red line.
Then we will slew the scope (in RA only) to a spot reasonably low to the horizon in the East or West (whichever is more convenient and less visually cluttered) and adjust the mount’s altitude (up/down) control until PHD2 says we’re good.
Now we go BACK to the original spot and touch up the Azimuth because the Altitude adjustment we just made has wrecked our earlier setting by a some small amount. If you made any significant adjustments on the 2nd trip to the Azimuth position, that adjustment has deteriorated the Altitude setting a little so we’ll touch that up as well with a second run.
I’ve never had to do more than 2 pairs of adjustments to get really accurate guiding. We will then verify our results using the Guiding Assistant tool.
The fastest I’ve ever been able to do this is 40 minutes. The reason why this takes so long is because it can take up to 2 or 3 minutes of drifting/guiding to get your answer/result as you get close to the end of the process.
The actual process
If you haven’t already, launch your planetarium software and get it talking to the mount. It should show you being centered on whatever star you did your original GoTo to.
Use the settings of your sky map software to show the Celestial Equator line and the Meridian Line.
Consistency is everything here so you need to decide two things going in: what part of the area around the Cel Eq / Meridian line you’re going to work in for AZ adjustments and are you going to be using the East or the West ‘horizon’ for your Altitude
I always use the area just above the Equatorial line and right (West) of the Meridian line. I use west because you don’t have to worry about your star crossing the Meridian if it takes a long time to get dialed in at first.
Slew your scope to the Meridian/EQ line.
If you’ve chosen the West as your Altitude adjustment side then your scope should be in a configuration something like this: If you chose East then the counterwieght bar should be on the other side.
Select your star close (but not super close) to the intersection point. A couple degrees above and to the right (West) is what I pick. It doesn’t really matter you just have to do it the same way every time.
Call up PHD2, set your exposure for about 1 sec. Pick a good star, a good solid bright one but make sure that the peak value is consistently below 255 (saturation point) and make sure you are as in focus as you can get. FWHD values in the 3s or 4s. Start guiding. PHD2 will run through a short 3-5 minute calibration run of your mount where it moves it in a series of small adjustments to calibrate how much of a pulse moves the mount what distance.
In PHD 2 go to the “Settings” tab and change the color of the RA axis to the darkest color you can find which will essentially make it disappear; you are only interested in the Dec line for this task so you don’t need all that visual clutter. Remember where this is though, you’ll need it visible later.
Go to PHD2 Tools Menu and select Drift Align and a dialog box will come up. Take this moment to also open up some kind of note-taking software or if you’re feeling retro, set the wayback machine for 1982 and get a piece of paper and pencil out for note taking.
In the dialog box that came up click on the “Drift” button. PHD2 will now continue to guide in RA but will disable Dec.allowing the star to drift north or south. PHD2 will now start taking readings every time your cam takes a photo (1 sec) and start to draw a trend line which, unless you are the luckiest person on Earth will most likeley be going fairly steeply upward or downward after you’ve watched it for 15 seconds or so.
Now you need to adjust your mount’s E/W pointing to make the line being drawn be more flat. But which way? How much to adjust? This was the part that always confused me because I couldn’t get consistent results. First of all we need to determine what kind of adjustment on the amount affects the line in what way. So you just need to try something. In the bottom of the graph area of PHD2 there will be "Polar Alignment Error= x many pixels” which at this point is probably a number between 750 and 3000, which is pretty far off so a relatively large adjustment will be needed.
On the PHD2 dialog click the “Adjust”
A truly large adjustment in this case is a 270° turn of your azimuth knob. Make the adjustment; you should see your mount swivel a little. Note (on your pencil and paper) which knob you turned (East side of mount or West) and which direction you turned it (clockwise or CCW). Write words to the effect of “Turning the knob on the west side of the mount clockwise makes the red line go…(up/down)“
On the PHD2 dialog click the “Drift”. The graph display will clear and it will start a new data run. Wait at least 30 seconds before trying to draw any conclusions about what you see and then note the Polar Alignment Error number and line orientation. Is the number lower? Is the red line flatter? If the answer is ‘Yes” then you are turning the mount the right direction. Which way did the line move after your adjustment? If your line was too low before the adjustment and the line is now flatter write the word “up” after “makes the line go…” in your pencil/paper note. If the answer is No (line is not flatter then hit the “Adjust” PHD2 button and turn the opposite knob, twice as much (two x 270° turns) and hit Drift. Wait 30 seconds and look at your number and line orientation which should be improved from your original 1st time position.
In my particular situation I’m working with a Losmandy G-11 mount which allows you to move the mount either direction from either knob but for the sake of consistency I always turn the knob on the West side of the mount. So, if I turn the West side knob Clockwise, my mount swivels, and my red line goes up in the drift graph. Conversely when I turn the West knob counterclockwise, my red line goes down. Your mount might work differently and your knobs might do different things so you need to be observant about what’s happening.
Having a clear understanding of this situation and what happens to the red line when you turn your mount knobs a particular direction is THE single bit of knowledge you need to get this procedure to work without chasing your tail.
Ok so now you should be on the right track to getting a perfectly flat line.
We are going to make a series of increasingly smaller adjustments until we get the red line as flat as possible. From here on out we are not going to be making any more big knob turns. Once you’ve determined which adjustment makes the line go which direction hit "Drift” again and wait 15-30 seconds and make your line assessment. Hit “Adjust” and make your adjustment–err on the side of too small an adjustment until you get a handle on how much of a knob turn affects the line a particular amount.
Understand that as the line gets flatter you have to drift longer and longer to get a meaningful result out of your line because the line is the average of all the data points taken every time your guide cam takes a photo of the guide star. When you start getting your line to within 100 pixels or so after a minute of drift you should consider changing your guide cam exposure to 2 seconds from 1. Why? Atmospheric turbulence is a THING and when you shoot fast exposures you can possibly just chasing the seeing, so by taking exposures farther apart you make changes in seeing less a part of your line’s slope. Unfortunately this makes the process take longer.
Repeat this process until you start getting the line pretty flat. Note that once you get the line within 20 pixels of error the knob turns you’ll be making are so small that you’ll barely be sure that you’re even turning them at all. It’s also of benefit to keep ’sneaking up on it’ from the same direction rather than blowing past it and having to come back the other direction; the gears in this system aren’t all that precise and sneaking up on it takes the play in the gears out of the equation.
Here’s a thing about the trendline graph: the line is an average and is subject to distortion by “rogue events” that might come along at inopportune times like the beginning of a run. A ‘rogue event’ might be a gust of wind hitting the scope or a gecko climbing on your scope, whatever. So If you’re starting a drift run and you see some truly large deviation, especially early on, you can always hit the CLEAR button and start over.
At some point you’re going to be thinking, “this might be it, I might be done” so change your exposure time to three seconds and let the graph run all the way to the end unless it becomes patently obvious that it’s not as close to flat as you were thinking it was gonna be. If you can get the line to within 10 pixels of error over a long run at 3 second exposure length you can say ‘good enough’ for your Azimuth adjustment and hit the PHD2 “Altitude” button.
Altitude Adjustment
PHD2 will stop all guiding now and it’s time to pull up your sky map program and slew the scope to your next position for the first Altitude adjustment.
Slew the scope ONLY in Right Ascension until the left box (which shows the distance from the Meridian) is between 50 and 60 (plus or minus depending if you’re West or East). Make sure there’s a suitable guide star in the frame, the mount is settled, set your exposure time to 1 second and start looping on PHD2.
Typically your Altitude adjustment isn’t as far out as your Azimuth was so your first adjustment is NOT going to be a big crank in one direction of another. Since this is your first adjustment in Altitude, you don’t know what kind of adjustment is going to affect the red line in what way. So go back to that piece of paper and write, “Adjusting the mount HIGHER makes the red line go…” Now go through the process of determining which kind of adjustment causes the line to move which way and once you’ve determined it remember to fill that info in on your note. On my mount, while looking at a star in the West, if I turn the altitude knob CCW the red line goes up.
Repeat the process as you did with Azimuth, Drift/Adjust/Drift/Adjust with longer exposures, progressively smaller adjustments, longer run times until you’re within 5-10 pixels of being totally flat for an entire run of the graph.
In one of physics’ great injustices that adjustment you just made in Altitude has undone some of the work you did on Azimuth so you need to go back there and touch that up.
Slew the scope back up to the intersection of the Meridian and the Celestial Equator and make SURE that you pick a star in the same general area that you did last time; if you were up from the equator and right of the Meridian last time, do that again.
You should be pretty close already so when I make adjustments to the azimuth at this point I’m turning that knob about as small as I can turn it and still be sure that I actually DID turn it. Sometimes when I come back here my line is still flat and I consider myself ‘done’ but if you made significant adjustments in altitude you’ll probably have some work to do here.
Go back to the same area for a 2nd Altitude adjustment and you should be pretty close by now. If you’ve been meticulous in your approach you should be pretty damn close to perfect polar alignment which is good because this has probably taken well over an hour.
If you’ve got good “line performance” in both Azimuth and Altitude you can get a reading on your total polar alignment error by using a different PHD2 tool: the Guiding Assistant tool.
PHD2 Guiding Assistant Tool
At this point I slew the telescope back up to the Meridian point and go to the PHD2 Tools menu and get up the Guiding Assistant. I usually uncheck the “Measure Declination Backlash” box in the lower left which is a separate measurement from Polar Alignment which is what we’re concerned about here.
Hit “Start”. The instructions for this tool say to let it run for at least 2 minutes “for things to settle” but I usually let it go for 4 minutes to average out those ‘rogue events’ and get as solid an average as possible.
There are a lot of numbers on this screen but the main ones you need to pay attention to are the numbers in black in the upper right and the bottom number in the left column “Polar Alignment Error”. You want all these numbers to be as low as possible.
In the High Frequency Star Motion box, good values are 0.5 or under and the best I ever was able to get my Polar Alignment Error was 0.1 arc minutes.
Hope this helps and you can now add drift alignment to your polar alignment toolkit.
Carpe Noctem!
Bill the Sky Guy