Drift-correct graph interpretation

[Miguel G]

Good morning,

Can anybody please explain how to interpret the graphic below from phd2logview? I've been searching but cannot find the documentation explaining it.

Regards

[Brian Valente]

Miguel

Please upload the guidelog. This graph is one of several tools that together explain what's going on

--
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/CABpRrO9VPPJpKjPGmsO6u0XCKBXbkuCqUKgwoNVp8G5xzCEoWg%40mail.gmail.com.

--

Brian 

Brian Valente

astronomy portfolio https://www.astrobin.com/users/bvalente/ 

[Miguel G]

Hi Brian,

Probably the log I'm attaching does not correspond to that screenshot, but it does show a similar graph; in fact depending on the time block the graphs are different and hence I would like to understand and interpret.

Thanks

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/CAJa45i4HeBLWbcA2%2BPD6nJoQX7t_tSZrfc3rEm9%3D4HJ3iGrn2g%40mail.gmail.com.

[Brian Valente]

Dec is Red, RA is blue

The screen cap you posted is the raw RA, so basically the guide corrections back out to approximate what unguided output might look like.it's one of two screen in the Analysis

 

the "arc" is normal drift across the sky due to a variety of factors (refraction, flexure, etc.). Large scale things like this are very easy to guide out. (on shorter guide runs you'll notice this is absent)

the "jaggy teeth" are higher frequency PE that needs further look using the Analyze Selected Frames (see below)

moving to the Frequency Analysis tab, it will show any periodic error by amplitude, which is helpful to see any obvious signs of PE. 

in your case you can see the primary period is approx 286 seconds and has a 5.8" peak-peak amplitude. (you can also see the large scale drift to the right) 

Right-clicking on your guiding results, you can then choose Analyze Selected Frames which is the residual (uncorrectable) erro

Here you can see that primary period error is not entirely corrected, and sticks out as an obvious place to look for improvements

Reviewing your settings, I see RA is a generic Hysteresis with relatively low aggression, so I would switch to PPEC with a period of 286.4 seconds, disable auto adjust period, and set the predictive aggression a bit higher like 70-80. That will likely clean this up quite a bit.

imo PHD Log viewer is really a combo tool, you need to look at both the selected frames as well as raw RA

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/CABpRrO-Btb-dR0ro36ULp-W0J4MVggTS3Ui-xgmj9CtHmEyGSg%40mail.gmail.com.

[Miguel G]

Hi Brian,

Thanks for the detailed analysis.

Sorry if I was not clear enough about my doubt. From the two graphs, the frequency analysis one is quite clear to me, but not the drift-corrected one.

You mention that the arc is normal drift across the sky due to a variety of factors, but what I don't really understand is what the graph means. The X axis is the time of the block you are analyzing, and the Y the error; right? Since the graph is drift corrected, does that mean that for example at time=1000s the error on RA was about 7" ? That's the point I don't get and most likely I'm missing something.

And now, by the way, you mention the frequency graph. Yes, the mount is an AM5N and the 286s is the period of rotation. I've read a lot of times that ppec is not useful with harmonic mounts due to the fact that they don't have a predictable periodic error, but I was surprised when I saw that peak. Does that mean that pec can be corrected on harmonic mounts? This mount is quite new and I have not had the chance to run some reliable tests, but yes I tried ppec and not noticed a significant improvement but I need to test again deeper, though.

Regards

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/CAJa45i5LEyciVoWfzoAs0Ge1Ni_%3D4v_j5JdechXvhnx_O%2Bci2g%40mail.gmail.com.

[Brian Valente]

>>>  Since the graph is drift corrected, does that mean that for example at time=1000s the error on RA was about 7" ? That's the point I don't get and most likely I'm missing something.

No, Remember this is with corrections backed out, so it describes what RA would be if there were no corrections at all. An approximation of the mount tracking without guiding

>>> Does that mean that pec can be corrected on harmonic mounts? This mount is quite new and I have not had the chance to run some reliable tests, but yes I tried ppec and not noticed a significant improvement but I need to test again deeper, though.

The data doesn't say anything about what mount type it is or whether it's appropriate for a certain mount type. It only shows a clear repeatable error it's measured at that period. I would try PPEC with the previous suggested settings and see what happens. I don't know what settings you used to test PPEC before.

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/CABpRrO9ypKC7MMw06dyzMPZj5N-B3ygkQosHBnoL9vkiYmMWvQ%40mail.gmail.com.

[Miguel G]

thanks Brian for the explanation. I would give a shot to ppec as suggested.

Regards

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/CAJa45i7AM-9Bpr0f9GXhR4A%2BofOvw_hvy7503wxe3ZGA4N1zYA%40mail.gmail.com.