Polar Alignment Discrepancy Between PHD2 and Sharpcap?

[mshe...@gmail.com]

Just curious. I'm getting a significant difference between PHD2 GA polar alignment error and Sharpcap polar alignment error. PHD2 shows an error of 40 arc-min while Sharpcap is under 10.

    Can folks provide me any guidance about the discordant results and how to resolve the discrepancy? Thank you and best regards.

Mike

[Bruce Waddington]

The measurements are being made differently.  PHD2 estimates polar alignment error based on the rate of drift in Declination.  That's an objective measurement but the cause of the drift could be something other than simple polar alignment error.  So if you have recent to doubt that it's due to polar alignment, you will need to find another cause: flexure in the mount payload, substantial imbalance of the OTA in Dec, hanging cables, a loose clutch, etc.  If you look at the unguided Dec behavior for extended times on each side of the meridian, that may point you in the right direction.  Since you didn't upload any data, that's about all I can tell you.

Regards,

Bruce

[mshe...@gmail.com]

Thanks! I should be running PA again later next week and should have more data. Best regards.

Mike

[Tommy M]

I usually have a similar discrepancy.  However, when SharpCap shows an improvement, so does PHD2 GA.  I had a PA error of 6, I polar aligned again with SharpCap and it went down to 2.2 (and the improvement showed in my guiding graph).  That's all I care about.  Unlike Bruce, I have no idea how the magic works, only that it does!

Tommy

--
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.
To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/327d5d9b-a9ac-46eb-a696-9e33372c94f4n%40googlegroups.com.

[alex reinhardt]

I don't want to hijack this but I have a similar discrepancy with PDH2 drift numbers and NINA 3 pt alignment.  Tonight, my PA is about 3' according to NINA and PHD2 drift PA is anywhere from 7' to 15' depending on the time period selected; total guide RMS error is about 0.5" today.  Based on your comments Bruce, I will certainly look at my cables as there does seem to be a bit of difference before/after flip (the scope is pier mounted and pretty solid and I just checked the balance again today).

Bruce, question then, is the PHD2 PA estimate accuracy influenced by where in the sky you are pointed?  Today, I'm imaging around LBN1228 which is only about 10 deg from Polaris.  Would that change the accuracy of the estimate vs pointing south?  Reason I ask is when I'm imaging towards the south, PHD2 and NINA PA are usually pretty close...or maybe it's just my rig/cables.  Thanks!

Alex

[Tim Stevenson]

Alex, I just finished setting up an EQ6R mount on a pier and worked my polar alignment.  I used NINA first to get my close and SharpCap to fine tune the PA <1'.  Then while gathering some PEC measurements in PHD2 using Guiding Assistant I noticed that the PA changed over time.  At first it was ~2' and then after 10-20min it settled to be <1'.  Note: I ran GA for such a long time to get several worm gear periods recorded.  Therefore, I'm in agreement with Bruce, look at unguided Dec behavior for extended times on each side of the meridian to make the comparison.

Tim

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/48de37bc-5313-44aa-b188-1e8655ae32b1n%40googlegroups.com.

[Bruce Waddington]

PHD2 is only estimating polar alignment error by measuring the rate of drift in declination.  It assumes that *all* of the drift is caused by polar alignment error because there is no way to estimate the contribution from other sources.  Or at least there is no easy way to factor those things out.  The common contributors to drift are small movements in the payload riding on the mount - slightly loose fittings, sagging (flexure) of the OTA or various adapters, compression of the soft materials used in mounting rings or on the ends of thumb-screws, dangling cables, cable drags and pulls, the list goes on.  Obviously, all of this is dependent on pointing position because that determines how gravity exerts its pull on all the different components of the payload. To visualize this, you have to come to terms with the tiny amounts of movement that the guide camera can see.  Most of the common guide cameras now have pixel sizes smaller than 5 microns, some much smaller, and most people have no hands-on familiarity with measurements at that scale.  If you look at the smallest interval on a typical metric ruler, that will be 1 millimeter - and that interval corresponds to 1000 microns.  Very fine leads for a mechanical pencil are likely to be 700 microns thick; a human hair is typically more than 50 microns thick.  These are truly tiny distances.  So it's pretty hard to imagine that all the typical gear riding on a telescope mount can't contribute to unwanted movements at this scale.  And if this isn't enough, atmospheric refraction will also cause apparent drift as the altitude of the target changes during the course of the night.

The important thing here is to not obsess about drift unless it's clearly causing problems with your final results.  Drift is the easiest kind of problem to guide out so there's no point in worrying about a problem that in all likelihood is not limiting your results.  The reason we talk about it all is that large amounts of drift may cause problems with PHD2 calibration and may create image rotation artifacts in your final images.  To avoid that, we generally tell people to get a polar alignment error below 10 arc-min as a general rule of thumb.  And clearly, there is little point in comparing alignment estimates from two difference measurement approaches when neither one is likely to be very relevant.

Bruce