EQ6-R Pro star-cross fail?

[discardedastro]

Hi all,

I'm trying to figure out what's going on with my guiding - while I'm seeing <0.8" RMS in PHD2 I still see some subs come out with a lot of lateral blur. After a particularly bad one I decided to try a star-cross test in PHD2, which produced the following image (~200s SC test + 180s exposure after).

So something's clearly not terribly happy here - or at least, it doesn't look it.

I figured the results were due to a mechanical failure - counterweights not balanced or something. However, I re-ran the test after checking everything and noticed the south adjustment step just pretty much instantly finish, with each step taking no time at all. So now I'm thinking software, but not sure.

I'm using PHD2 with INDI and the EQmod driver. I'm sat here with an exposure running and the reported RMS guiding error is 0.45", fairly balanced RA/DE. I'm using the predictive RA algorithm and resistshift dec.

Any ideas on how to go about working this out? Example of a 900s sub showing that blur:

[michael8554]

Hi

In a successful Star Cross Test the star should end up in the middle of the cross.

In your test the mount has moved west

moved east to the middle

moved east again to the right end of the cross

moved west back to the middle

moved north to the top of the cross

so far so good, but the final moves are missing - did you stop the test here ?

Your image of the Whirlpool Galaxy has elongation in Dec.

Is your Dec RMS Guiding Error 0.45" or thereabouts too?

In that case you have differential flexture of your guidescope.

PHD2 is guiding the guidecam very well, but the imaging scope and/or imaging camera are moving in Dec with respect to the guidescope.

But if Dec RMS is worse, you have Dec Backlash to contend with.

If you did do all the moves in the Star Cross Test, that would explain why the mount hasn't moved south after reaching the top of the cross !

Michael

 

[discardedastro]

6 minutes ago, michael8554 said:

 

so far so good, but the final moves are missing - did you stop the test here ?

Your image of the Whirlpool Galaxy has elongation in Dec.

Is your Dec RMS Guiding Error 0.45" or thereabouts too?

In that case you have differential flexture of your guidescope. 

PHD2 is guiding the guidecam very well, but the imaging scope and/or imaging camera are moving in Dec with respect to the guidescope.

But if Dec RMS is worse, you have Dec Backlash to contend with. 

Yep - so I didn't stop the test, but PHD "raced through" the south correction steps. I was using the defaults (24x1000ms, if memory serves) and W/E/N all took about 25 seconds as you'd expect but PHD2 just appeared to skip the actual moves for S.

The backlash calibration is odd; normally when I run a backlash calibration the results are fairly sane (~620ms pulse to clear the backlash) but I did have the guiding assistant tool fail to check the Dec backlash on this night - it had succeeded earlier in the night so didn't fret about it.

My guiding error is indeed about the same in RA and Dec from memory - haven't seen wild variance, at least, the graphs both look as noisy as each other normally. I'll dig the logs out tomorrow and post them up. I hadn't considered differential flexure at all, though I've heard about it. I've got the guidescope mounted on the tube rings with an ADM dovetail set and Primaluce Lab rings, thought I'd be able to avoid any problems there...

[michael8554]

6 minutes ago, discardedastro said:

I was using the defaults (24x1000ms,

The instructions I've read say single steps of 5000mS at 1X Sidereal, so 8 moves in all ?

But shouldn't matter, more than one way to crack an egg.

Not sure what you mean by "raced through" the south correction steps ? You paused after each step to allow it to move before applying the next, but you saw no south movement ?

I'd still say Dec backlash, so will be interesting to see your guide figures for Dec.

Adjustable guide rings with soft-tipped screws will have micron sized flex, which will translate to arcsecs in your images.

But I'd be more inclined to believe the mount is too well balanced, so is wobbling north and south by the dec backlash amount - so balance it to make it take up the backlash.

Michael

 

[discardedastro]

On 12/04/2019 at 00:04, michael8554 said:

The instructions I've read say single steps of 5000mS at 1X Sidereal, so 8 moves in all ?

But shouldn't matter, more than one way to crack an egg.

Not sure what you mean by "raced through" the south correction steps ? You paused after each step to allow it to move before applying the next, but you saw no south movement ?

I'd still say Dec backlash, so will be interesting to see your guide figures for Dec.

Adjustable guide rings with soft-tipped screws will have micron sized flex, which will translate to arcsecs in your images.

But I'd be more inclined to believe the mount is too well balanced, so is wobbling north and south by the dec backlash amount - so balance it to make it take up the backlash.

Michael

 

Ah, I'm using PHD2's automated star-cross test which gave me different moves. It sat on each direction doing all the steps, but got to South and just went straight to "completed". I'll give a manual star-cross via the manual guiding input in INDI a try, and see if that produces the same results. I think that's just software, given the other moves seem fine.

I am using adjustable guide rings, so that tallies up. Being a Newt, I'm fairly sure that I'm not *that* well balanced, but I'll have a play around and make sure the load's a little bit on the south side.

I think my options for fixing this are going to be to replace the guidescope with an OAG or replace the tube rings for the guidescope and see if that solves it; both are about the same cost so I'm inclined to go OAG particularly as that should produce a much better arcsecond resolution on my guide camera (ASI120MC - would be 0.77" on the OAG rather than 2.97" on the guidescope). Definitely planning to swap the guide camera for something more sensitive/higher resolution/mono in any case, having now done the maths on my guide resolution!

[michael8554]

7 hours ago, discardedastro said:

ASI120MC - would be 0.77" on the OAG rather than 2.97" on the guidescope

2.97 down to 0.77arcsecs ? That's going the wrong way, assuming you're imaging at 1 to 2 arcsecs/pixel,  IIRC.

Didn't make South moves? Guiding set to North Only ?

Michael 

 

[discardedastro]

5 hours ago, michael8554 said:

2.97 down to 0.77arcsecs ? That's going the wrong way, assuming you're imaging at 1 to 2 arcsecs/pixel,  IIRC.

Didn't make South moves? Guiding set to North Only ?

Michael 

 

Guiding set to N+S. I'm imaging at 0.5"/px.

[discardedastro]

Attached a few guiding logs. I'm generally seeing <0.55" on RA and <0.45" on Dec. I did notice, while checking the scope over yesterday, one of the guidescope screws had come loose at the rear of the tube - so I suspect that's what caused this becoming so noticeable in my last night. I'm going to trim the plastic on the screws back to the bare minimum to limit the compressive error potential, add a bit of Loctite, and screw them in firmly so the whole system is under tension.

Hopefully that'll mitigate this stuff enough for now while I research a better fix, which is looking like adding the ZWO OAG to the front of the filter wheel and possibly upgrading the guide camera; that, or getting some better guide rings that don't rely on the delrin screws, but with the various potential source of optical error in the train I'm leaning towards an OAG.

PHD2_GuideLog_2019-03-28_192707.txt PHD2_GuideLog_2019-03-29_201628.txt PHD2_GuideLog_2019-04-10_200847.txt

[discardedastro]

Right - I've taken the tips almost entirely off the screws. The threads are pretty abysmal in the Primaluce Lab 80mm rings - there is a lot of play in the bolts when they're even slightly out of tension. So my attempt at fixing this with no budget revolves mostly on a bunch of Loctite on the bottom two screws - I still need to be able to remove the scope so I've left the top screw free on both sides.

I also went around the scope's grub screws and nipped everything up, which has helped reduce play in the coarse focus tube and helical focuser (still an annoying quantity of backlash there, but it's only a cheap scope and it isn't like it needs adjusting for focus often).

So hopefully this'll significantly reduce my problems - will have to wait for some clear skies and see how I go.