I see a variance of around an arc minute or so between polar alignment measurements. Back-of-the-envelope calculations suggest that the variance is within the errors of mid-range mounts. (I'm using a CEM60 and it works very well for the loads I put on it.) It's not going to be perfect, and we shouldn't be obsessed with these measurements. With a 120mm refractor f7.5 imaging scope and an 80mm f4 guide scope, I've achieved 35" RMS on my CEM60 which is better than I expected. I can't get quite as good with a C11 + the same guidescope on the mount, but mount dynamics are quite a bit different for the C11 load - PHD2 reports I'm between 0.5 and 0.8 arcseconds RMS with the C11 which is pretty good.
Recent guidance from the PHD2 folks suggest you should not strive for a perfect polar mount as that can contribute to oscillation in DEC, and instead, a bit of polar misalignment helps to steady DEC . The amount of misalignment depends on the length of sub-exposure and plate scale which determines the amount of rotation you will see in the image. PHD2 tends to settle down at a couple of arcminutes off NCP, but that is likely because I polar align the image axis, not the guide scope which is not perfectly aligned with the image scope.
Bottom line is the EKOS polar alignment works very well,
Interesting. I noticed Trevor on AstroBackyard gets pretty good guiding but he only does polar alignment with a polar scope, so he would probably have a couple of arcmins of error. Maybe the guide scope PA can be avoided on mobile setups.
So in a nutshell, if you are guiding you do not need A perfect alignment, but if you're not guiding you actually do need near perfect to be able to do long exposures.
I was able to reach 5 min un-guided the other day without any apparent star trails. And After I got the guiding on "track", 10 min exposure.
Good point, and it reminds me I should be more careful in what I way. Slightly off PA may help guiding but the longer the exposure, the more important the PA is. Even with good guiding, longer exposures will result in image rotation about the guide star if the PA is not very good.
When I say "you do not need A perfect alignment", I do not mean by that a rubbish PA. It has to be as good as it gets without loosing sleep over it sorts of speak ...
Yes. You want to be reasonably close. The attached chart is the best results I've had with PHD2 on my CEM60 and I was about an arcminute off. Looking at the scatter plot, you can see that RA was worse than DEC (but still very good at 0.31").
There are websites that will compute the amount of rotation you will get for a given alignment and image plate scale, so you can get an estimate of how close you need to be for a given exposure length.