Ekos calibration is a mess.

[Steenamaroo]

Hi all,

Hoping someone might be able to help a new-comer to guiding.

I have a 200p (downsizing shortly) on Vixen GP with EQstar EQ5,
and QHY5L-ii on 50mm Svbony guide scope.

I have ST4 option available, or pulse via serial.

I have set the EQstar unit in guiding mode but no matter what I try in Ekos the calibration either completes with questionable pattern and guides badly or,
much more likely, doesn't complete, failing at the last stage.

Most often I see drift....recovery....drift.....dot dot dot...no recovery. Doesn't even look like an attempt is made.

Does this make sense to anyone?

The EQStar EQ5 offers 0.5/0.75/1 guiding rate. I have tried all three, setting it at the eq5 control (for both axis) and in ekos, also making sure to adjust the proportional gain in ekos to match.
 

I briefly tried Linguider app to rule out a software issue but it has the same results.

Many thanks in advance.

Edited October 26, 2020 by Steenamaroo

[Steenamaroo]

I should probably add that I have had good clear, strong, guide stars using 1 or 1.5 second exposure, stable enough seeing,
and I'm testing in an area sheltered from wind. Polar alignment is decent, tracking is good, and the motors all work fine via manual control.

[MarkAR]

Pulse via serial is the better option.

Not familiar with the EQstar, is your profile set to EQMOD ?

[Steenamaroo]

Hi Mark,
At this stage I'd take any option that works by, yes, ideally I'll be using serial.

Yes, the mount is set as Skywatcher EQMod and, once connected, shows as EQ6 mount which is what I think it's meant to do.

[Steenamaroo]

7 hours ago, MarkAR said:

Not familiar with the EQstar, is your profile set to EQMOD ?

I'm starting to wonder about that. Has anyone here successfully guided with this kit?

Edited October 27, 2020 by Steenamaroo

[Steenamaroo]

Well, it seems the main issue here was polar alignment.
I used Ekos software alignment, made the corrections, and things are much better.
The calibration graph still looks a mess but I'm not concerned now as I'm guiding to within 2 arcseconds drift when there's no breeze.

[Danield06]

On 26/10/2020 at 21:21, Steenamaroo said:

I'm starting to wonder about that. Has anyone here successfully guided with this kit?

I guide with my HEQ5 pro from .7" to .85" RMS regularly. good nights I can get down to .55" to .70" RMS. It guides extremely well for me. 

What are ALL of your settings for guiding? What algorithm? Backlash measurement on? Etc...let me know everything or post a screenshot of the guide tab and all of the guide menu options form the "options" button on the bottom right.

 

Your calibration plot needs to be orthogonal or at least close to it. I am assuming it is not. Can you post what it looks like? I would also set exposures to something like 2.5". IMO 1" is too fast and you'll be 'chasing the seeing" essentially.

 

 

[Steenamaroo]

Hi @Danield06
Thanks for the reply.

The plot is not even nearly orthogonal but I suspected it should be. I had been trying to find images or videos of other people's calibrations but struggled so thanks for confirming this.
I'll take a short video of the calibration plot next time I get a clear view of the pole.

I'm still very much doing trial and error so I haven't settled on an algorithm, but had been trying Smart, SEP, and SEP (multi).

To be honest, the stage I'm at, changes in image filter, algorithm, and exposure, aren't making a tremendous difference.
I can calibrate as described and guide reasonably well but I definitely still feel like there's something not quite right and I'm not getting the best from it.

I'll attach screenshots and more info, as you asked, at the next opportunity.

"Backlash measurement on"
DEC backlash is measured during the calibration process - Is that what you mean?

I had been deliberately picking a very bright star to make things easier for myself, so 1 second was adequate, but perhaps I'd be better testing with stars which require slightly longer exposure,
as that's a more realistic real-use setup.

Thanks again.

 

 

Edited October 31, 2020 by Steenamaroo

[Danield06]

Yes, I would definitely change the exposure to 2.5". I wouldn't change from SEP Multistar yet. Reason being is that it is the recommended default and works extremely well when you get everything going correctly(which you will). I just helped another user out over on cloudy nights forums and we got it all sorted out. Went from a 2" + RMS to now regularly below 1" RMS, we will figure it out! Additionally go to indilib.org . Jasem(Username Knro), the creator of Indi and Ekos/Kstars is regularly replying to peoples issues as well as his team of contributors and moderators. Specifically, Hy(thats the username) is very active and he is responsible for the guiding tab, implementing new algorithms and settings etc. Just want to clarify, Jasem and Hy are two different people)

 

Somehow we will figure this out so do not worry. Plenty of people successfully guide and almost all issues I or others have had are somehow user/equipment related, not pertaining to the software. 

 

Edit: Have you done any type of mods to your mount? Rowan Belt mod? I had to tidy up my belt mod and my calibration plot when from a scatter plot to nearly orthogonal. Just wanted to put that out there incase you have done so.

Guarantee once we get your calibration to be orthogonal that your guiding will be perfect. 

Edited October 31, 2020 by Danield06

[Steenamaroo]

Thank you for all that info. I really appreciate it.
The EQ5 kit doesn't seem to be particularly well known but it's a belt-driven stepper motor kit.

I think my mount RA gearing could be tightened up a little - It's not what you'd call loose but DEC is tighter, so I'll have a look at that today.
The EQ5 kit motor mounting brackets are heavy enough but they don't lock snugly against an edge - You're effectively securing two sheets of metal together with a single nut + bolt so it's not unreasonable to think there could be some minute movement there.

That's another thing I'll look at today.

 

I had suspected the weight, and sail-like characteristics, of my 200p might have been contributing to the problem so I'm mounted up with a new (to me) 130pds now.
Didn't actually make a difference under nice conditions but I expect it will be making my life easier when there is a very light breeze or whatever.

Thank you again.

[Steenamaroo]

Got a break in the clouds...Here are some screen shots showing a typical guiding plot.

I know there will be recommendations and preferred settings and, of course, I want to hear them,
but please keep in mind I've tried every combination of exposure duration, image filters, and guiding algorithm, possible.

Mount was polar aligned by eye, then by the alignment tool in ekos. It's damn close.

I've tightened up everything I can think of tightening up - Belt pulleys are lined up nice and straight, and the belts are clear of the edges.
Everything's tight as a drum, and the guide camera cable is cable tied to the main scope mount ring, so there's no possibility that there's any pull on the camera or anything like that.

 

Edited November 2, 2020 by Steenamaroo

[michael8554]

If the display shown bottom right in your last image is analogous to a PHD2 display, then the cluster of blue Dec dots instead of a straight line suggests you have forgotten to move the mount north until the star moves before Calibrating, to take up Dec Backlash.

Instead the Dec Calibration starts with clearing the backlash instead of moving the mount, which compromises the Calibration.

Also PHD2 likes a 12 step Calibration, yours are only 5 steps or so, which may be normal for Ekos, I wouldn't know, so maybe try a lower Pulse Length ?

Even so, with good PA, Dec should only need infrequent corrections, so your 2 to 3 arcsec guiding is probably down to the guide settings and balance.

I understand you can use PHD2 with Ekos ?

Michael

[Danield06]

4 hours ago, michael8554 said:

If the display shown bottom right in your last image is analogous to a PHD2 display, then the cluster of blue Dec dots instead of a straight line suggests you have forgotten to move the mount north until the star moves before Calibrating, to take up Dec Backlash.

Instead the Dec Calibration starts with clearing the backlash instead of moving the mount, which compromises the Calibration.

Also PHD2 likes a 12 step Calibration, yours are only 5 steps or so, which may be normal for Ekos, I wouldn't know, so maybe try a lower Pulse Length ?

Even so, with good PA, Dec should only need infrequent corrections, so your 2 to 3 arcsec guiding is probably down to the guide settings and balance.

I understand you can use PHD2 with Ekos ?

Michael

Yes you can use PHD2 but they shouldn't have to. The internal guider works very well. As you are suggesting, it is most likely an issue with settings. I had sent a PM to OP. They need to increase their iterations, and change a few settings. I use the Ekos internal guider and I can see a few things that should be change before moving onto a new issue. 

This messages seems like it is attacking you, but I promise in no way am I doing so lol. I am just suggesting that if there is an issue with guiding using the internal guider then it is a user issues, not a result of the software.

Edited November 4, 2020 by Danield06

[Steenamaroo]

Thank you both for replies. I'm taking it all on board but this strikes me the most.

"you have forgotten to move the mount north until the star moves before Calibrating"

 

Forgotten...First I've heard of it. Is this something I need to do in ekos?
I've read quite a lot of support documentation but never saw this.

I have had very good RA in/out at times but dev is always a mess - Could this be the main reason?
 

 

Another thing I discovered is that disabling dithering entirely seems to help quite a lot.
I'm going to try tonight, weather permitting, with the settings suggested in PM ( Thank you @Danield06) and dither on/off.
Choosing the best from that test I'll then experiment with reducing guiding rate slightly.

[alacant]

Hi

A few observations and hints.

- Dismatle DEC, clean, grease and take up the backlash.

- Balance DEC to perfection: 10-plus kg on an eq5 demands it

- Take the guiding box down to its default of 32; in dense star fields there's chance of an adjacent star infiltrating.

- Use the guide algorithm SEP-Multistar rather than SEP. 

- You can use PHD2 with EKOS but with hy's new GPG and multistar, it's become second best. 

- Try pulses of 400ms

13 hours ago, michael8554 said:

you have forgotten to move the mount north

EKOS does this for you. The OP already has it checked.

HTH

 

Edited November 4, 2020 by alacant

[Steenamaroo]

Thank you for the clarification and suggestions @alacant
Balancing is something I've found difficult - With clutches off both axis are quite firm and it takes a fair push to move either of them.
They're not ridiculously tight feeling, or anything, but the it would take a heck of an imbalance for them to rotate as a result of it.

As such there's a lot of guess work going on when trying to balance the scope/mount.

Is this abnormal? Should clutches-off movement be less firm?

Edited November 4, 2020 by Steenamaroo

[Davey-T]

3 minutes ago, Steenamaroo said:

As such there's a lot of guess work going on when trying to balance the scope/mount.

You can use those  digital luggage scales to pull it both ways to check resistance.

Dave

[Steenamaroo]

[alacant]

43 minutes ago, Steenamaroo said:

it would take a heck of an imbalance for them to rotate

OK, so you need to slacken the cone bearing at the base of each axis. After you've cleaned and regreased of course;)

Both axes need to spin freely.

Cheers

[Steenamaroo]

I suppose it's a question of how freely but that seems like conflicting information there.
Releasing clutches and moving by hand isn't difficult, by any stretch, but to make a 130pds with canon 500d rotate DEC due to imbalance I'd have to shift the dovetail bar about an inch one way or the other.
Does that description help to clarify at all?

[alacant]

 

Even the slightest imbalance should make the axes rotate. 

For autoguuiding, the mount must be mechanically perfect. It takes only half an hour or so to make it so. Well worth it for the amount of variables you have removed when you need to diagnose any errors.

Cheers

[Steenamaroo]

1 hour ago, alacant said:

OK, so you need to slacken the cone bearing at the base of each axis.

Alright, I guess I need to figure out how to do this, then. Thank you.

[alacant]

11 minutes ago, Steenamaroo said:

figure out

There's a guide here I think:

http://www.astronomyshed.co.uk/forum/viewtopic.php?f=34&t=623

 

Edited November 4, 2020 by alacant

[Steenamaroo]

Thanks for that - That's this evening sorted! 😀

[michael8554]

4 hours ago, alacant said:

 

18 hours ago, michael8554 said:

you have forgotten to move the mount north

EKOS does this for you. The OP already has it checked.

HTH

Yes he has checked Dec, and it is moving, but the first 4 moves appear to be clearing backlash, not Calibrating.

Michael