67 replies to this topic

[OzAndrewJ]

Gday Damo

I think I'm turning into a PEC addict

Nahhh, its just another diagnostic tool in the arsenal ( as is FFT analysis ).

If you can pro actively remove the bulk of the errors before guiding, it makes a the guiding a lot easier.

I always suggest people do a long duration plot with PEC off then on to get a baseline, then enjoy yourself.

At a later date, ,if something looks odd, just redo the plot and compare to earlier datums to get a hint of what might have changed.

 

I've run the new curve through PEMpro and If I understand this correctly I think I'm now getting 4.3 arcsecs pk-pk with the PEC on.

Yes and no, as it changes between each tooth, which is normal for the gearing Meade uses.

I replotted your data without it being overlayed by period as it gives a different view.

( you can see a scale at the top that shows the worm periods as a guide )

I see you still have a small high freq ripple in your data and that wont guide out at the frame rate you are using.

If you use the full range of the FFT window in the viewer, you will see this actually shows at 90x the worm period

and based on your gearing, this indicates the gear on the motor is eccentric.

Its only about 1arcsec pk-pk, but swapping motors might give a better result, but its a pain to do :-)

Just a game of chasing diminishing returns

 

Andrew Johansen Melbourne Australia

 

 

 

 

[Damo19]

Thanks Andrew,

 

The Pempro log viewer looks to be a very useful / powerful tool, especially when you know what you are doing

 

Below is a snapshot of the PHD guide graph while the PEC is active. I was imaging M81 (bode galaxy) at the time so Its a bit of a cheat as this galaxy is close to the celestial pole, but even so achieving a total RMS 0.4 arc-secs was not possible before without the help of the PEC. Not too shabby at all for an LX85  

 

 

 

[Menocchio]

Cheers again! After updating PEC for a few nights now, it has easily made at least an arcsecond of difference, and along with incorporating the multi-star guiding in PHD2's dev build, I'm rarely seeing my graph go beyond the 1" mark. I had no idea the LX-85 was even capable of this kind of precision. That one patchette of code has been a game changer lol. 

Edited by Menocchio, 27 April 2021 - 03:25 AM.

[OzAndrewJ]

Gday Menocchio

Good to hear.

If you have a reasonably good gearbox, then the PEC will be repeatable and if smooth as well

a PEC model can really help.

When its a moonlit night, whats always fun is to turn PEC off

and start a PHD log with guiding turned off ( via the brain )

Let it run for 2 -4 revs of the worm, then turn PEC ON and do another 2-4 revs.

Load it into the PEMPro viewer and overlap by period to see how consistent it is.

Andrew Johansen Melbourne Australia

[Andynator]

Anyone willing to do this mod for $$ if I mail them my Audiostar handset?

[SeaLint]

Hello all! 

 

So if I'm following the thread correctly it sounds like my LX85 PEC capable scope is not PEC capable out of the box? I'm new to GEM/tracking mounts and have read the thread a few times and still trying to figure out the fix. Is it a firmware update? If so where can I find the better firmware? I'm currently running my LX85 with the 6"ACF SCT. I'm getting into astro photography and would like to have the PEC working. 

 

Thanks!

 

-Clint

 

Hi Andrew,

 

I've applied the patch and run a dummy RA train with some pulse inputs and its looking good so far! Also the model name entered in Audiostar is "LX85 8 ACF" which is slightly different to what was reported in the log. Either way it seems to work fine from the test run results.

 

I'm all setup now so if I get lucky with the weather I'll do some proper PE training runs using PHD etc. Its going to be really interesting to see how much the PEC straightens things out.

 

Once again thanks for sorting this out   I think it would have been a really long wait for Meade to resolve this themselves given how many firmware updates they put out.

 

Cheers

 

Damo

[OzAndrewJ]

Gday Clint

So if I'm following the thread correctly it sounds like my LX85 PEC capable scope is not PEC capable out of the box?

Correct. It also has a lot of other bugs.
 

still trying to figure out the fix.

You need to update the hand box with patched A4S4 firmware.

( This fixes most bugs and adds a few enhancements )

The freeware patch loader and basic patches can be found here

http://www.stargps.ca/downloads.htm

I still havent got round to updating the patch, so you will need to add the earlier posted snippet

( ref post  #24 )

for a 6" SCT the patch edit would be

 

;//================================================================================================
;// Below here are Bug Fixes
[-------------   Bug Fixes   -------------]
S=UnChecked
> ----------------------------------------------------------
;//================================================================================================
[Allow PEC for LX85 6" ACF]
S=Checked
> Allows LX85 model to do PEC training
;// F8-B02E model selection
R 0D6C30  41 55     // change VM1 to LX85-6" ACF
R 0D6C33  7F 26     // change if equal goto xF8-B0B3, to "goto xF8-B05A" and test there jumps on

;//------------------------------------------------------------------------------------------------
[Fix Drive Train / Cal Sensor Resets ]           // *** Tested OK ***

 

For other people, the possible list of LX85 mounts and values is
 

#x50  "LX85 Mount"       
#x51  "LX85 5 Refractor" 
#x52  "LX85 6 MAK"       
#x53  "LX85 6 Reflector" 
#x54  "LX85 8 Reflector" 
#x55  "LX85 6 ACF"       
#x56  "LX85 8 ACF"       
#x57  "LX85 10 ACF"      
#x5D  "LX85 70 APO"      
#x5E  "LX85 80 APO"      
#x5F  "LX85 115 APO"

 

 

 

 

Andrew Johansen Melbourne Australia

Edited by OzAndrewJ, 31 May 2021 - 05:31 PM.

[rdc121974]

I am an LX85 owner and I am trying to implement the PEC capability. If I understand correctly you download the PatchAUA2S4v02.spf file and then copy and paste the commands from post 24 into the file in Notepad. You then save the file and upload into the hand controller. Does it matter where in the A2S4 file you copy and paste the information at? Not a coder but I do understand changing some basic commands in this fashion. I am also struggling sometimes with polar alignment and getting the proper information entered into my handset in order for it to recognize my location and how many degree's from the pole I am located. I am in Manakin-Sabot VA. Any tips or instructions from any of you on getting the most from this mount would be greatly appreciated. I ordered a Williams 81mm APO/ZWO294MC PR0 and a 178MC for this mount and I already have a 50mm guide scope with a ZWO120MM for guiding. I'm going to remove the 8in. SCT that came with the mount and practice with a smaller refractor for a while with a larger field of view. I love reading all of your posts. I have learned a great deal just from reading.

[OzAndrewJ]

Gday rdc121974

 

We uploaded a new patch version ( A4S4_v02) recently that has the LX85 patchette already incorporated

so no cutnpaste required.

You may need to edit the patch ( its a simple text file ) if you want to hardcode

the particular model

ie find this bit in the patch

// following pinches the VM1 slot to allow one LX85 to be picked
R 0D6C30  41 50  "LX85 Mount"
;R 0D6C30  41 51  "LX85 5 Refractor"
;R 0D6C30  41 52  "LX85 6 MAK"
;R 0D6C30  41 53  "LX85 6 Reflector"
;R 0D6C30  41 54  "LX85 8 Reflector"
;R 0D6C30  41 55  "LX85 6 ACF"
;R 0D6C30  41 56  "LX85 8 ACF"
;R 0D6C30  41 57  "LX85 10 ACF"
;R 0D6C30  41 5D  "LX85 70 APO"
;R 0D6C30  41 5E  "LX85 80 APO"
;R 0D6C30  41 5F  "LX85 115 APO"

 

The line in Red has no preceding semi colon so is what will be selected

If you want to select a specific model, you put a semicolon at the beginning of this line

and remove the semicolon from the line you want to use..

 

After that, just load the patch as per normal, but ensure the option is selected

 

getting the proper information entered into my handset in order for it to recognize my location and how many degree's from the pole I am located.

You can manually edit the site record in the handbox to store your real location,

( or, as you are patched, you can use my PEC editor to do that via a PC screen )

but dont understand the degrees from the pole bit???

You cant set that, it drops out of each align process you go through

 

Andrew Johansen Melbourne Australia

[rdc121974]

Thanks Andrew a polemaster helped solve my polar alignment issues. Where can I download A4S4_V02? Is it at the cjdskunkworks website?

[OzAndrewJ]

Gday rdc121974

The cjdskunkworks site is for the new ASCOM driver, and has nothing to do with firmware or patching.

The patches and the loader ( StarPatch ) can all be found on the StarPatch site.

http://www.stargps.ca/downloads.htm

You may need to manually download the A4S4 spf file and rom as the auto downloader

still picks the default A1F7 data

Andrew Johansen Melbourne Australia

[rdc121974]

Thanks for getting back to me. I figured that out last night and got the A4S4_v02 installed into the HC last night. StarGPS works so much better than the MEADE software for updating your firmware. Thank you for all your help. Praying for everyone in Australia right now. I hope you are not affected by some of the craziness over the Covid measures and protests. Thank you so much for all your help. You seem to know more about the LX85 mount than anyone I have spoken to so far. You sir are a veritable savant! 

 

Russell Coleman Manakin VA.

[atsidor]

Hi all,

I've followed along with this thread for PEC training on my own LX85. I think I have been successful (1 training run followed immediately by 2 update runs to get a nice average). I am now using MyScope to analyze and adjust the PEC model. I'm wondering what is the best approach for smoothing the PEC corrections. There is clearly large period variation in the worm that, I think, could be handled nicely if I do a lot of smoothing.

 

But then there are other shorter period variations of about 1-2 arcsec in amplitude over only a few offset periods. Do I do minimal smoothing and thus try to correct for those too? Or is that just not going to get me anywhere?

 

I have attached the two options (one with smoothing set to 1 and one set to 4) for example.

 

Of course I can test it out but just trying to get some guidance to hopefully avoid too much troubleshooting.

Attached Thumbnails

• 

[atsidor]

Attaching other example here because of size. This is more smoothing.

Attached Thumbnails

• 

Edited by atsidor, 08 October 2021 - 06:53 AM.

[OzAndrewJ]

Gday Atsidor

Normally, i would say smooth a bit harder, as the data in Meades isnt really repeatable across turns.

That said, you appear to have a "mechanical" problem in the gear meshing that you should

probably fix "mechanically", before doing PEC.

Open the file using the edit tab and then look at the FFT tab.

That will give you the period of the ripple and from that we can see what gear is not meshing properly.

( post the *.pec file here and i can look also )

My notes show the LX85 was a worm with a 52 tooth gear on it

then a 26 tooth gear on a 45:1 gearbox. That implies the fault is in the gearbox :-(

but need the FFT to confirm it as the nos dont add up when i count the ripples manually

 

As to using my app for the smoothing, what i have found best is not to use Meades "update"

You do a drive train then save the file

You then do a new train and save the file

repeat

This way you can see if any runs had bad data and can manually exclude them.

Meades firmware auto saves and processes every run, even if you abort a run

and you cant see that in the merged data.

[atsidor]

Ok I was trying to sort out the same thing. Here is the *.pec file attached. I am also taking a look at the FFT. 

 

Looks like a peak at 52 corresponding to the one gear and then a whole bunch between 2 and 20, which... I'm not sure.

Attached Files

•  pec_train2.zip   299bytes   3 downloads

Edited by atsidor, 08 October 2021 - 05:12 PM.

[OzAndrewJ]

Gday Atsidor

OK, definitely something inside the gearbox.

There is a tiny peak in the FFT at 52x the worm period and this is the tooth to tooth meshing

on the gear on the worm.

The rest of the FFT is not giving a clear error, but 20x is showing as the major error

and i assume that is a freq inside the gearbox, but no one has opened one up yet

so no idea whats in it. It doesnt look very robust tho.

 

Andrew Johansen Melbourne Australia

[atsidor]

Ok I guess a little out of luck then. I am probably not brave enough to open it up the gearbox. Going to try guiding using the "smoothed" version and see how it turns out. Maybe with luck the small oscillations will guide out.

 

Gday Atsidor

OK, definitely something inside the gearbox.

There is a tiny peak in the FFT at 52x the worm period and this is the tooth to tooth meshing

on the gear on the worm.

The rest of the FFT is not giving a clear error, but 20x is showing as the major error

and i assume that is a freq inside the gearbox, but no one has opened one up yet

so no idea whats in it. It doesnt look very robust tho.

 

Andrew Johansen Melbourne Australia