AZ EQ6 RA glitches

[almcl]

My until-now reliable AZ EQ6gt mount has developed a trend just lately of going wrong in RA.  Up to yesterday it has been occasional or very
infrequent but last night it happened about 5 times.

It seems that tracking either stops or speeds up.  On the PHD2 screen the stars appear to race off along the RA axis and a 1.5 or 2 sec
guiding exposure shows a streak rather than a dot.  Of course, the imaging sub that is underway at the time is just a blur.

I don't think PHD2 itself is to blame - it issues corrections immediately - but the mount doesn't respond or takes too long  (traces below of
two such events show that average oscillation prior to the event was around 0.6" or 0.8" but that when the problem occurs it got to 25" or
even 50" deviation before it was corrected).  

So now I am wondering where the fault lies?  Could it be the Lynx Astro EQdirect cable is failing in some way?  Or the electronics in the
mount itself, or possibly the RA tracking motor? The fact that the error appears exactly aligned with RA has put me off it being the camera
and OAG moving in the focusser, although I guess it could be that?

Any suggestions on how to narrow this down a bit before resorting to replacing bits in 'cheapest-first' order, or from anyone who has seen
this sort of bevhaviour or may have solved it would be welcome?

 

  

[symmetal]

Have you got the mount tracking 'East heavy' to keep the gears meshed and to avoid the RA possibly 'floating' across the backlash area and causing an effect similar to what you have.

Alan

[Laurin Dave]

Looks to me as if it stops tracking and then re-starts, how often does it do it?  Worm period is 480 secs so maybe its binding

Dave

[almcl]

Thanks for the thoughts both!

9 hours ago, symmetal said:

Have you got the mount tracking 'East heavy' to keep the gears meshed

Yes, we are 'east heavy'.

8 hours ago, Laurin Dave said:

Looks to me as if it stops tracking and then re-starts, how often does it do it?  Worm period is 480 secs so maybe its binding

That's what it looks like to me, too.  There's no regular pattern to it. Sometimes it will track for an hour or more with minimal error, and the absence of a pattern makes me think random electrical glitch rather than mechanical, but wondering how to narrow this down?

There is a peak in the PEC at about 480 seconds, but it is fairly small, around 1.8" and PHD2 seems well able to deal with this:

[blinky]

What’s the screenshot above from? As for your problem, is it maybe a bit of grit or something on the gear? Time for a strip down and degrease maybe...

[Dr_Ju_ju]

What do the PHD logs show ?

[almcl]

4 hours ago, blinky said:

What’s the screenshot above from? 

PHD log viewer (https://openphdguiding.org/phd2-log-viewer/) then right click the graph and select analyse selected , raw RA then the Frequency spectrum button.

4 hours ago, Dr_Ju_ju said:

What do the PHD logs show ?

In log viewer they show normal(ish) guiding for extended periods then a sudden jump (as in image above) with corrections being issuedand either the star is lost (1st image) or dragged back (2nd image):

 

The fact that the deviations show both above and below the line is puzzling me, too.

 

[Dr_Ju_ju]

How about the Debug logs in Documents\PHD2 ?

[almcl]

30 minutes ago, Dr_Ju_ju said:

How about the Debug logs in Documents\PHD2 ?

Julian

They are a bit massive (7Mb for the debug file), but for the event in the last image above, move requests were small decimals (0.27, 0.16).  Suddenly at 21.06.57 the RA move jumps an order of magnitude to 7.7 then in successive move requests it goes to 12, 22, 36, 22 before dropping back to smaller numbers and eventually to small decimals.  From this, I don't think EQMOD lost contact with the mount but rather that the mount didn't respond for a while?

There are no error messages that I can see so PHD2 seems to think all is well.

[Dr_Ju_ju]

If that's the case, then check your cables\connections, it may well be a 'dirty' connection where PHD2 is issuing a nudge, but the mount isn't getting it. Also check the guiding algorithms aren't too aggressive. 

[almcl]

I'll certainly have a go at the connections, but wonder if it's actually further downstream than the plug.  Perhaps I should look at the circuit board in the mount and check the cables there?

[Dr_Ju_ju]

I doubt that they would be at fault, but no harm in looking, BUT BE VERY CAREFULL, the connectors are small and surface mounted, so can VERY EASILY be ripped off the motherboard.... I know, because I've done it  !!

[almcl]

Thanks for the warning, hamfisted is my middle name!