PEC training Success on CGEM

[MarsG76]

Hello Astronomers,

It's been quite a while since I had time to play with astro toys... even my cooling project, even though it is somewhat cooling the DSLR, is still a project in progress...

It's been a while after the CGEM firmware has been updated and has reportedly fixed the broken PEC, I decided that it was time to play with PEC training and try and improve the amount of DSO detail I capture, particularly at 2000mm focal length.

As is natural with this hobby, it couldn't have been straight forward, no.. it had to have its problems and obstacles to overcome, really there was only one problem, but a "deep buried" one...

When I initiated PEC record, the mount just kept slewing toward the east on the RA axis without end until it hit the mount limit... simply put, the mount couldn't find the index point.

Looking at the TeamCelestron site, I found a post where a CGEM DX user has the same problem, and Derik, one of the firmware engineers, suggested that the mount had a faulty sensor or even a Motor Board fault... so at this point I thought that I might have to live without PEC or buy a new mount... but ultimately I kept thinking about what the issue could be...

I didn't want to let this go I thought that I'd swap the RA and DEC motors to see if this would make a difference but on pulling out the MC board and removing the RA motor, I very quickly realised that there is a extra ring or plate on the RA motor with a notch cut into it moving in between a U black... obviously this is the index sensor and swapping the motors would not work.

The other thing I noticed is that the index ring and U sensor were covered in grease so I cleaned the grease off the motor, gears and sensor using isopropyl alcohol... re greased the gear cogs and reassembled the mount... NOW PEC was working, initiated PEC record and it found the index mark within a second... wooohoo

The next clear night, last night, I setup NexRemote on the COM port, with a virtual COM port setup and connected PECTool directly to NexRemote.

I setup and calibrated the scope completely off the laptop, and once it was ready, I ran guiding using the ASCOM celestron driver connected to the NexRemote virtual port and commenced autoguiding on PHD2. I ran PEC training in PECTools and let it repeat for 14 runs.

I uploaded the averaged corrections from all 14 runs to the mount... downloaded the graph back from the mount successfully verifying that PEC program is in the CGEM memory.

Now I was ready to do a autoguiding accuracy test comparison with and without PEC but... as it should be expected... clouds!!!

I parked the scope and powered down the observatory, ready for the next clear night, ready for a detailed test and perhaps a imaging session with PEC running, and it hopefully making a noticeable difference to guiding and sub quality... to be continued.

 

Clear skies,

MG

[MarsG76]

I had a chance to play with PEC for a few nights. Trained, re-trained and averaged single night to various nights of training runs.... I compared the RMS guiding accuracy at various times with both PEC enabled and disabled, my results are as follows.

I tested PEC at 2032mm focal length using my 8" SCT on the CGEM mount with PEC averaged one night of PEC training runs, multiple nights worth of PEC training runs and with PEC off. I let it run for 15 minutes and read the average RMS.

I found that PHD2 reported the lowest RMS correcting with the PEC OFF!!!!!... around 0.5" RMS tonight.. mind you that the difference between the PEC enabled and disabled was only a matter of around 0.05"-0.1" between the tests... seems like it's hardly worth while.... 

How disappointing as currently it seems like that there is not only no improvement with PEC but it can actually make tracking (very slightly) worse??

 

Did anyone have a similar experience with PEC???

Attached are screen grabs on my PEC training averaged graphs. PEC training was performed by the Celestron ASCOM driver connected to NexRemote virtual COM port in PHD2 and PECTool connected directly to NexRemote. PHD2 was calibraed and allowed to guide the guide star as PECTool was recording the RA pulses.

Any input into what I might be doing wrong and hence the disappointing PEC result is welcome....

 

Clear skies,

 

MG 

 

[newbie alert]

I've not tried pec with my mount as there's a whole host of things wrong with it first and what I'd assume to be refining with pec would be the last thing to do..

Are you talking in pixels or arc mins on your RMS figures?

[Thommy]

I don't have any experience with PEC. But I've considered using it on my EQ8 to prevent some rapid movements in the gears. So just some general observations:

If you use PEC you would probably set the aggresivity of the guiding much lower, as rapid movements should be removed by the PEC - otherwise you would overshoot leading to a higher RMS figure.

If your gears are of good quality, and the grease is clean and free from debris, you would probably not expect much of a difference.

Just some thoughts ...

/Thommy

[hugo52]

Hello marsG76, I am trying this tool, on my CGEMII, do you have a PHD2 Guidelog, I would compair it with mine.

Regards Hugo

[MarsG76]

1 hour ago, hugo52 said:

Hello marsG76, I am trying this tool, on my CGEMII, do you have a PHD2 Guidelog, I would compair it with mine.

Regards Hugo

Unfortunately I didn't keep any logs, but I'll tell you that comparing with and without PEC didn't make any (if any) significant tracking improvements... and I did tests over a number of nights...

My current camera setup had a pixel resolution of 1.16"/pixel and my average RMS tracking is usually 0.7-0.8" and the best I have see was 0.38" on a great stable night, so the CGEM with PHD2 already track at sub pixel resolution, so PEC can't improve my imaging... 

BUT

After tracking with guiding and PEC enabled my laptop crashed and the autoguider became stuck on a pulse in one direction than sudent;y my RA side of my mount PCB failed... I attribute it to a state where autoguide port was getting a pulse in one direction and PEC tried to push it in the opposite direction, creating a electrical fault (current drain/short/excessive current due to fighting the directions) which might have fried my CGEM mount main board. So if you're already tracking at your imaging camera max res/pixel, think about if its needed and worth the risk... if you're already tracking at or sub pixel resolution, your detail in your images is now limited by the seeing rather than tracking.

 

[MarsG76]

On 09/05/2019 at 17:02, newbie alert said:

I've not tried pec with my mount as there's a whole host of things wrong with it first and what I'd assume to be refining with pec would be the last thing to do..

Are you talking in pixels or arc mins on your RMS figures?

Arc seconds

[MarsG76]

On 09/05/2019 at 18:02, Thommy said:

If your gears are of good quality, and the grease is clean and free from debris, you would probably not expect much of a difference.

Just some thoughts ...

/Thommy

And you're right.... there wasn't much improvement... atleast not with a camera that has a pixel size or around 5um...

[hugo52]

Thank you for sharing this information, Regards Hugo

[Starflyer]

Have you ticked the box to see visually the corrections that PHD2 is making?

With PEC off you should be seeing a lot more corrections than with PEC enabled.  This is an easy sanity check that PEC is actually being applied.

What algorithm are you using in PHD2?

[MarsG76]

18 hours ago, Starflyer said:

Have you ticked the box to see visually the corrections that PHD2 is making?

With PEC off you should be seeing a lot more corrections than with PEC enabled.  This is an easy sanity check that PEC is actually being applied.

What algorithm are you using in PHD2?

I haven't ticked any extra buttons that show visually the corrections outside the graph that get generated as the mount is guided... I use Resist Switch for both RA and DEC... but I suspect that not only does PEC make little to no difference on my current sensor pixel size resolution and that PEC can damage the PCB during a coincidental alignments of conditions.. this is what happened to my mount...