×

INDI Library v1.9.9 Released (30 Nov 2022)

Bi-monthly INDI Library released with new drivers and bug fixes.

EQMOD/EQ6-R regression/fault?

  • Posts: 12
  • Thank you received: 0
Hi,
Tearing my hair out over this one!
So I have run a rig with Ekos/KStars for a good long while now without issue and know my way around pretty well. My setup is two Pis, one running INDIserver on the telescope and one in the warm running KStars, Ekos, and PHD2. Both are Pi 4s, 8G.
The setup is a 200PDS and a EQ-6R Pro, early version. Two ZWO cams for image and guide, and off-axis guiding. Sesto senso focusmotor. Power is plentiful via a Nevada linear PSU, and slewing is smooth and consistent. There's been no mechanical change to the setup in years, aside from a successful mount rebuild which has been working fine.

I bumped to INDI 1.9.7 and KStars and Ekos seem very confused about what's going on. With the scope booted up from cold, with no alignment/mount config, pointing roughly at polaris in the default park position, plate solving correctly finds the location and KStars shows the mount position as it is on the mount position overlay.

However, Ekos is showing the error in alignment as huge - >200,000 arcseconds - on more or less every solve. I can't discern a pattern. Even when pointing at the same location and platesolving twice I get huge errors. Slewing by right-click on a target in KStars points the mount incorrectly in real terms and Ekos still shows error but in slew to target mode doesn't actually move the mount. The result is that I can't point the scope.

I have wiped all the config, started from scratch twice, checked everything mechanically, and verified the fundamentals (GPS position is right and being used everywhere I can see, time is NTP-synchronised on both devices to a local grandmaster and UK pool, nothing's under huge load). PHD2 polar align reckons about 7 arcminutes of PA error, which isn't great but should be perfectly fine for the basics. I've tried with both the Align and Alignment subsystems, and the closest I can find to an error message anywhere is "[ALIGNMENT] Failed TransformTelescopeToCelestial" a bunch in logs.

At my wits end when it comes to further diagnostics - any suggestions for approaches on how to figure out what might be going on? Debug logs attached.
2 months 3 weeks ago #88083
Attachments:

Please Log in or Create an account to join the conversation.

  • Posts: 225
  • Thank you received: 16
I know you said that you checked the time... but the error amount seems about right for a one-hour time error. Are you certain that your GMT offset is correct?

Good luck!

Ron
Mounts: Sky-Watcher EQ6-R Pro, Meade LX85
OTAs: Celestron 8" Edge HD w/HyperStar, Meade 80mm APO Triplet Refractor
Cameras: ASI533MC Pro, ASI183MC Pro, ASI224MC, ASI120MC-S, ZWO ASI290MM
Computing: Raspberry Pi 4 with Stellarmate OS, ASIAir Plus, MacBook Pro
The following user(s) said Thank You: James
2 months 3 weeks ago #88085

Please Log in or Create an account to join the conversation.

  • Posts: 47
  • Thank you received: 6
Join the club... I upgraded to the most recent stuff tonight (11/12/2022 - using Stellarmate OS) and I have had NOTHING but issues with my EQ6-R Pro.
In fact, it's so bad I'm moving my captures onto my external storage and am going to go back to the original version that I had, as I KNOW that works with my setup... It's rather frustrating as it appears that regular licensed StellarMate OS users are actually the "BETA" testers for it, even when not using the "cutting edge" BETA stuff.
I am SO tempted to dump these RPi4's off on someone else and simply invest in two ZWO ASI Airs. At least THEY seem to work reliably and don't keep taking a crap all over themselves with every update. I lost over 5 hours of capture time tonight dorking around with this stuff. And for what I've spent on these RPi's I could have easily purchased the ASI Airs. The only reason I didn't was I liked the idea of being able to use other cameras down the line.
Hell, I've even thought about going to a Intel NUC, but I figure I'd have the same issues with it as I do with StellarMate since they both would use the same core software.
And yes, my time is set correctly... and it recognizes that we are in DST here in my EQMOD config screens for the mount.
A beginner trying to find their way.
Astro What? - my home away from home.
2 months 3 weeks ago #88087

Please Log in or Create an account to join the conversation.

  • Posts: 225
  • Thank you received: 16
Tracy,

Sorry for your issues. I also use an EQ6R Pro and am using the latest build, but not experiencing this issue.

One thing you mentioned (and may have misspoke) is that you are "in DST here". I don't know where you are located, but we are not in Daylight Savings Time here in North Carolina.

Good luck!

Ron
Mounts: Sky-Watcher EQ6-R Pro, Meade LX85
OTAs: Celestron 8" Edge HD w/HyperStar, Meade 80mm APO Triplet Refractor
Cameras: ASI533MC Pro, ASI183MC Pro, ASI224MC, ASI120MC-S, ZWO ASI290MM
Computing: Raspberry Pi 4 with Stellarmate OS, ASIAir Plus, MacBook Pro
Last edit: 2 months 3 weeks ago by Ron Clanton.
2 months 3 weeks ago #88088

Please Log in or Create an account to join the conversation.

  • Posts: 47
  • Thank you received: 6
TZ offset is -6 now, whereas before the time change it was -5... so the TZ is set correctly, even though I may have typed DST we are in CST - the offset is correct though.
I mistyped as I was also working in PI trying to figure out how to do color on narrow band before Stellarmate OS took a mighty dump on itself after a meridian flip.
A beginner trying to find their way.
Astro What? - my home away from home.
Last edit: 2 months 3 weeks ago by Tracy Perry.
2 months 3 weeks ago #88091

Please Log in or Create an account to join the conversation.

  • Posts: 12
  • Thank you received: 0

Replied by James on topic EQMOD/EQ6-R regression/fault?


So I'm one of those boring system administrator types who insists that all their servers and other headless devices run in UTC; both Pis are set to Etc/UTC as a timezone and were showing good NTP health, and manual checks showed them synchronised to what I expected. There is a GPS/gpsd on the telescope Pi for location data, which is a little redundant given it's a stationary setup, but it isn't used as a source for system time (e.g. by chrony, as PPS/GPS ref, etc). I'm also in the UK, so in GMT anyway. I also had this issue with local time as BST, a month or so ago.

I will double-check, but I am very confident it isn't this. Good thought though!
2 months 3 weeks ago #88096

Please Log in or Create an account to join the conversation.

An update is pushed to StellarMate & INDI PPA Stable channels. Please check.
Jasem Mutlaq
Support INDI & Ekos; Get StellarMate Astrophotography Gadget.
How to Submit Logs when you have problems?
Add your observatory info
2 months 2 weeks ago #88103

Please Log in or Create an account to join the conversation.

  • Posts: 228
  • Thank you received: 58

Replied by nou on topic EQMOD/EQ6-R regression/fault?

I had once similar problems like these with my EQ6-R. I solved them like this.
1. slew to some point away from pole
2. plate solve with action set to do nothing just to get align error.
3. manually adjusted scope position without moving mount to remove align error
4. repeat from step two to get until I got error to half a deg.

After this I get much better initial align and got rid instances when mount refused to move to correct position.
Last edit: 2 months 2 weeks ago by nou.
2 months 2 weeks ago #88119

Please Log in or Create an account to join the conversation.

  • Posts: 12
  • Thank you received: 0

Replied by James on topic EQMOD/EQ6-R regression/fault?

OK, so I rebuilt everything from scratch on the INDI side. Fresh Raspberry Pi OS install (Lite 64-bit) and Astroberry repositories set up.
I set up chrony for NTP sync with the UK pool and local servers, and set the gpsd plugin to use the system time so there was no confusion from the GPS. All the clocks in KStars, local system time on both machines, etc all synchronised to UTC and using the same timezones (Etc/UTC).

I used default configurations for everything otherwise in INDI etc.

I left the scope in the home position, pointed at Polaris, and powered up. I then solved a few times and got wildly differing answers. I then moved the scope (slewed from software) and tried again, and again had wildly differing answers while the image remained perfectly still (160 degrees error).

File Attachment:

File Name: alignfail_...1-15.txt
File Size:1 KB


Interestingly the KStars EQMod Mount position indicator is perfectly correct throughout all this.

I then repeated the Polar Alignment routine and got to below an arcminute, just to be sure of that, and tried using the internal solver and got similar answers.

So, no luck so far.
2 months 2 weeks ago #88171
Attachments:

Please Log in or Create an account to join the conversation.

  • Posts: 225
  • Thank you received: 16
That's really weird. I've never had this problem with my similar setup. However, I use StellarMate... so things may not be exactly the same.

One thought... Have you tried taking the GPS and NTP Sync out of the equation? Those are two components that I don't use. I just setup the location manually and get the time from my home network. I'm thinking that there may be some issue with their interaction???

Ron
Mounts: Sky-Watcher EQ6-R Pro, Meade LX85
OTAs: Celestron 8" Edge HD w/HyperStar, Meade 80mm APO Triplet Refractor
Cameras: ASI533MC Pro, ASI183MC Pro, ASI224MC, ASI120MC-S, ZWO ASI290MM
Computing: Raspberry Pi 4 with Stellarmate OS, ASIAir Plus, MacBook Pro
2 months 2 weeks ago #88173

Please Log in or Create an account to join the conversation.

  • Posts: 12
  • Thank you received: 0

Replied by James on topic EQMOD/EQ6-R regression/fault?

I took GPS out of the equation, both by setting it to use system time in the gpsd INDI driver settings, and by turning the gpsd driver off entirely.

NTP sync is essential to keeping everything to UTC since nothing has real-time clocks, and is working correctly. Time really isn't an issue here! Location from GPS was correct and the site management tab of the EQMOD driver had correct information.
2 months 2 weeks ago #88175

Please Log in or Create an account to join the conversation.

  • Posts: 225
  • Thank you received: 16
Okay... and KStars shows the correct location and coordinates in the lower left corner of the display?

If that checks out... I give up! Someone from the support team will need to weigh in.

Good luck!

Ron
Mounts: Sky-Watcher EQ6-R Pro, Meade LX85
OTAs: Celestron 8" Edge HD w/HyperStar, Meade 80mm APO Triplet Refractor
Cameras: ASI533MC Pro, ASI183MC Pro, ASI224MC, ASI120MC-S, ZWO ASI290MM
Computing: Raspberry Pi 4 with Stellarmate OS, ASIAir Plus, MacBook Pro
2 months 2 weeks ago #88176

Please Log in or Create an account to join the conversation.

Time to create page: 0.742 seconds