#APCC - V1.9 Tracking Error #APCC


Roland Christen
 

Try running the same test with the model turned off. See what the drift looks like.

Roland



-----Original Message-----
From: Rouz <rbidshahri@...>
To: main@ap-gto.groups.io
Sent: Thu, Aug 26, 2021 4:05 pm
Subject: Re: [ap-gto] #APCC - V1.9 Tracking Error

Hi Ray,

I'm not completely certain but I did 2 tests back to back without changing anything, about an hour apart. Did not touch anything physically.



Run 1:
 Model Dec spacing 1 degree - Ra spacing 5 degrees  -   PEM ON for APPM and capture   =  Had drift

Run2:
Model Dec spacing 1 degree - Ra spacing 4 degrees  -   PEM OFF for APPM and capture  =  No drift


Again PEC file was incorrect.


With the much closer spacing, the drift was much less than previous attempts but still very clear with PEM on.  I tried all possible combinations or dec arc/refraction/model corrections on and off.

Also, I don't know how practical it is to make a model with only 1x4 degrees apart per point. The whole sky would need thousands points. I'm hoping its not that spacing but the corrupt PEC.


--
Roland Christen
Astro-Physics


Rouz
 

I can try that next time.

-Rouz


Rouz
 

Roland,

I did try with PEM on and all corrections off, including the model.

Only got 1 x 300s image and it was quite poor.


Roland Christen
 

Up until now we only have the what, but not the why. In order to find the why, I suggest using PHD or MaximDL to generate a guiding chart with the guide corrections turned off. Run the guide software for 15 to 20 minutes to generate both an RA and Dec graph showing both the drift and the cyclical nature of the tracking. Do it with PEM off and PEM on. Once you have the graph, you will instantly know exactly what your basic equipment, polar alignment, etc are contributing to the actual tracking rate. Until you have a clear picture of the tracking rates in RA and Dec you are grasping for straws.

The whole point of the previous exercise was to see how the system tracked by taking a series of 300 sec exposures. But since you cut it off at one exposure, we have learned essentially nothing. So, instead of repeating your exposures, the only thing I can suggest is to do the above. At least it will establish a baseline.

Roland


-----Original Message-----
From: Rouz <rbidshahri@...>
To: main@ap-gto.groups.io
Sent: Fri, Aug 27, 2021 12:48 pm
Subject: Re: [ap-gto] #APCC - V1.9 Tracking Error

Roland,

I did try with PEM on and all corrections off, including the model.

Only got 1 x 300s image and it was quite poor.

--
Roland Christen
Astro-Physics


Roland Christen
 

By the way, you don't need a separate guide camera to do this simple test. Just use your main imaging camera to be a guide camera in your guider software. Once you have a 15 to 20 minute guide chart (with no corrections sent to the mount) the shape of the RA and Dec tracking curves, with and without PEM, will give all the information needed to figure out what is going on. No modeling during this test!

Roland

-----Original Message-----
From: chris1011@...
To: main@ap-gto.groups.io <main@ap-gto.groups.io>
Sent: Fri, Aug 27, 2021 1:38 pm
Subject: Re: [ap-gto] #APCC - V1.9 Tracking Error

Up until now we only have the what, but not the why. In order to find the why, I suggest using PHD or MaximDL to generate a guiding chart with the guide corrections turned off. Run the guide software for 15 to 20 minutes to generate both an RA and Dec graph showing both the drift and the cyclical nature of the tracking. Do it with PEM off and PEM on. Once you have the graph, you will instantly know exactly what your basic equipment, polar alignment, etc are contributing to the actual tracking rate. Until you have a clear picture of the tracking rates in RA and Dec you are grasping for straws.

The whole point of the previous exercise was to see how the system tracked by taking a series of 300 sec exposures. But since you cut it off at one exposure, we have learned essentially nothing. So, instead of repeating your exposures, the only thing I can suggest is to do the above. At least it will establish a baseline.

Roland


-----Original Message-----
From: Rouz <rbidshahri@...>
To: main@ap-gto.groups.io
Sent: Fri, Aug 27, 2021 12:48 pm
Subject: Re: [ap-gto] #APCC - V1.9 Tracking Error

Roland,

I did try with PEM on and all corrections off, including the model.

Only got 1 x 300s image and it was quite poor.

--
Roland Christen
Astro-Physics


Rouz
 

Hi Roland,

Got it, will do that next. 
Was waiting for clear weather, the single exposure is from a few nights ago.

I'm familiar with PHD and will use that as you have stated below with the main scope (I have no guiders).

I could run it through the "guiding assistant" that works out the error rates and polar misalignment as well.

Thanks,

Rouz



 

Hi Rouz

Can I suggest you do a baseline guiding by following these steps:

In addition to fulfilling Roland's needs, it will also ensure you have a proper baseline setup and avoid common configuration mistakes.

you will also want to do the optional step 8B 

I am a regular contributor on the PHD forums and this is what we use and recommend for this kind of thing
Brian


On Fri, Aug 27, 2021 at 1:45 PM Rouz <rbidshahri@...> wrote:
Hi Roland,

Got it, will do that next. 
Was waiting for clear weather, the single exposure is from a few nights ago.

I'm familiar with PHD and will use that as you have stated below with the main scope (I have no guiders).

I could run it through the "guiding assistant" that works out the error rates and polar misalignment as well.

Thanks,

Rouz




--
Brian 



Brian Valente


Roland Christen
 

The only thing I would leave out is the guiding part at the end. What we want to see is the shape of the drift with PEM ON and PEM OFF. Once we do that, I can use that info to determine if something is wrong in the PE correction data that's in the memory.

Roland

-----Original Message-----
From: Brian Valente <bvalente@...>
To: main@ap-gto.groups.io
Sent: Fri, Aug 27, 2021 3:57 pm
Subject: Re: [ap-gto] #APCC - V1.9 Tracking Error

Hi Rouz

Can I suggest you do a baseline guiding by following these steps:

In addition to fulfilling Roland's needs, it will also ensure you have a proper baseline setup and avoid common configuration mistakes.

you will also want to do the optional step 8B 

I am a regular contributor on the PHD forums and this is what we use and recommend for this kind of thing
Brian


On Fri, Aug 27, 2021 at 1:45 PM Rouz <rbidshahri@...> wrote:
Hi Roland,

Got it, will do that next. 
Was waiting for clear weather, the single exposure is from a few nights ago.

I'm familiar with PHD and will use that as you have stated below with the main scope (I have no guiders).

I could run it through the "guiding assistant" that works out the error rates and polar misalignment as well.

Thanks,

Rouz




--
Brian 



Brian Valente

--
Roland Christen
Astro-Physics


Rouz
 
Edited

Ok,

Uploaded PEC from a few days ago that looked good, read from the mount, curve is there.

Fine polar alignment.

Calibrated PHD with main OTA and camera.

Ran guiding assistant for 1000s with PEM OFF.

Ra P-t-P error was 4.3 arc-seconds.

Pa is good at 0.4 minutes.


Rouz
 
Edited

Now with the previous PEC curve and PEM ON.

Ra error improved slightly to 3.56 arc-sec peak to peak.

I suppose I should run PEMPRO again and try for a better PEC curve.

-Rouz


Rouz
 

Made a fresh PEC curve 


Rouz
 

Then re-ran PHD with the new curve.

The RA doesn't drift now like it did previously. I see a couple of arc-seconds of plus/minus movement but still better than before.


Ray Gralak
 

Hi Rouz,

Ra error improved slightly to 3.56 arc-sec peak to peak.
Just eye-balling the screenshot, that value does not look correct.

So, try loading your phd2 log into the PEMPro Log Viewer utility, which is designed to measure periodic error in PEMPro and PHD2 logs. If you don't have it, you can download it from the link below. I am the software's author, and it is completely free:

https://www.siriusimaging.com/downloads/PEMProLogViewerLatest.exe

-Ray


-----Original Message-----
From: main@ap-gto.groups.io [mailto:main@ap-gto.groups.io] On Behalf Of Rouz
Sent: Saturday, August 28, 2021 12:24 AM
To: main@ap-gto.groups.io
Subject: Re: [ap-gto] #APCC - V1.9 Tracking Error

[Edited Message Follows]

Now with the previous PEC curve and PEM ON.

Ra error improved slightly to 3.56 arc-sec peak to peak.

I suppose I should run PEMPRO again and try for a better PEC curve.

-Rouz


Ray Gralak
 

Rouz,

Then re-ran PHD with the new curve.

The RA doesn't drift now like it did previously. I see a couple of arc-seconds of plus/minus movement but still better
than before.
Again, just eye-balling your screenshot, it looks like PE is now worse. Maybe you need to invert the curve?

If you created a new PEC curve and uploaded it to the mount that might take 40+ minutes. So, unless you slewed RA back to near where you did your earlier measurement, the RA drift will likely measure differently. You cannot reliably compare drift between two logs unless the hour angle and declination are similar.

That said, as expected, there is an insignificant change in drift between PEM on and off, which indicates that PEM does not cause residual drift.

-Ray


-----Original Message-----
From: main@ap-gto.groups.io [mailto:main@ap-gto.groups.io] On Behalf Of Rouz
Sent: Saturday, August 28, 2021 2:40 AM
To: main@ap-gto.groups.io
Subject: Re: [ap-gto] #APCC - V1.9 Tracking Error

Then re-ran PHD with the new curve.

The RA doesn't drift now like it did previously. I see a couple of arc-seconds of plus/minus movement but still better
than before.


Roland Christen
 

That looks perfect. No need for a new PEM curve. The PEM compensation is doing exactly as it should, it is removing the slight 2 arc sec wiggle in the RA tracking rate.

Bottom line, the mount runs very accurately and the only reason for the remaining drift is due to slight polar misalignment, telescope flex, atmospheric refraction, etc. In order to remove this drift, you either use modeling with APCC or guiding. So the next thing is to make a new model and try this same test with PEM ON again. Theoretically the model will compensate for the drift in the two axes whether PEM is on or off, but it only eliminates the drift and not the 2 arc sec periodic error that you see in your first run with PEM off.

I hope this makes sense to you. There should be no interaction between the model and PEM. They are as separate as the Moon is from Mars. There is no reason why having PEM on would cause the drift to come back, unless somehow when you turned PEM on, you might have inadvertently also turned off the model. Ray might be able to illuminate this further.

So for a final test, try making a model and doing this same PHD test again. With the model active and PEM turned on, you should get a flat line for both RA and Dec for at least 10 minutes, if not more.

Roland


-----Original Message-----
From: Rouz <rbidshahri@...>
To: main@ap-gto.groups.io
Sent: Sat, Aug 28, 2021 2:24 am
Subject: Re: [ap-gto] #APCC - V1.9 Tracking Error

Now with the previous PEC curve and PEM ON.

Ra error improved slightly to 3.56 arc-sec peak to peak

I suppose I should run PEMPRO again and try for a better PEC curve?

-Rouz

--
Roland Christen
Astro-Physics


Roland Christen
 

You should have left the old curve in there. This new PE correction is NOT correcting for periodic error like the last one. It may have made it worse.

Please think about what the PE correction does and does not do:
1) PE correction eliminates the slow periodic wiggles in the RA tracking
2) PE correction does NOT remove the drift, nor does it produce drift.
3) PE correction does absolutely nothing to the Dec tracking or Dec drift. Dec is stationary and does not move.

The only reason drift is there or not is due to polar alignment, atmospheric refraction and other effects.
I'm afraid you are chasing red herrings down a blind alley because of fundamental misunderstanding of what PE correction does (and perhaps what modeling is supposed to do?) It is a complex subject and I cannot really cover everything in a few sentences. All I can do is to try leading you slowly down the proper path, but if you go off on a side adventure then it becomes difficult to get back to the right path. Unfortunately your side adventure was to make a new PE curve without waiting for my input of the tests that I suggested.

Roland



-----Original Message-----
From: Rouz <rbidshahri@...>
To: main@ap-gto.groups.io
Sent: Sat, Aug 28, 2021 4:39 am
Subject: Re: [ap-gto] #APCC - V1.9 Tracking Error

Then re-ran PHD with the new curve.

The RA doesn't drift now like it did previously. I see a couple of arc-seconds of plus/minus movement but still better than before.

--
Roland Christen
Astro-Physics


Roland Christen
 


I suppose I should run PEMPRO again and try for a better PEC curve.
Your PE curve is PERFECT! DON'T change it.

Rolando


-----Original Message-----
From: Rouz <rbidshahri@...>
To: main@ap-gto.groups.io
Sent: Sat, Aug 28, 2021 2:24 am
Subject: Re: [ap-gto] #APCC - V1.9 Tracking Error

[Edited Message Follows]
Now with the previous PEC curve and PEM ON.


Ra error improved slightly to 3.56 arc-sec peak to peak.

I suppose I should run PEMPRO again and try for a better PEC curve.

-Rouz

--
Roland Christen
Astro-Physics


Rouz
 

Hi Ray,

I believe I had it downloaded, will take a look at it. Thank you for the link (I was aware you have created quite a few software and contributed much to the community).


-Rouz


Rouz
 

Hi Roland,

Having had some sleep now (severely deprived) I see what you mean about the previous curve being better. A flatter curve to work with and then I use the model to correct for PA and other factors not worm gear related.

I'll reload the previous curve and re-test with PEM and make a new new model.


Also Ray:
My previous assumption was that the PEC file was corrupt and was throwing off the model numbers leading to the incorrect model and therefore causing drift.



 


Roland Christen
 

Periodic error can be describes as speed bumps at regular intervals on the road. The PE curve measures those and applies the opposite movement so that the ride is smooth (and your head doesn't hit the roof at every bump). PE doesn't compensate for the curve in the road.

Drift is the slow curve in the road which has to be compensated by steering the car (autoguiding) or by putting software into the steering mechanism which measures the curve and then applies the appropriate amount of steering to keep the car on the road. Periodic error has no effect on how the model steers the vehicle (in this case the mount) to eliminate drift.

Rolando

-----Original Message-----
From: Rouz <rbidshahri@...>
To: main@ap-gto.groups.io
Sent: Sat, Aug 28, 2021 1:51 pm
Subject: Re: [ap-gto] #APCC - V1.9 Tracking Error

Hi Roland,

Having had some sleep now (severely deprived) I see what you mean about the previous curve being better. A flatter curve to work with and then I use the model to correct for PA and other factors not worm gear related.

I'll reload the previous curve and re-test with PEM and make a new new model.


Also Ray:
My previous assumption was that the PEC file was corrupt and was throwing off the model numbers leading to the incorrect model and therefore causing drift.



 


--
Roland Christen
Astro-Physics