Aaaargh so frustrating! blurred stars with PHD guiding

[gasdoc]

Hi all,

Newbie imager here. Just got a EQ6R-Pro mount so set up last night trying to image M101 using my Sigma 150-600 lens attached to Nikon D7500 with 1.4 teleconvertor. 

Using zwo 120 with 200mm guidescope and PHD2.

Took about 3 hrs of subs (92 x 2 min @ ISO 1000) with flats, darks & bias and PHD reported errors as 0.13 so I thought I was going to get a half decent image (obviously not enough imaging time). However when reviewing the individual subs, every single one had blurred stars (see quickly processed image) - I've guided before with my star adventurer in RA using my WO 61 and never had stars this bad.

 

Is this due to poor PA? I checked the PA at the end and it had drifted from the guide circle by a small amount, but nothing significant and I thought that PHD2 would correct for this anyway?

 

Was wondering if anyone who understands PHD logs could have a look and suggest where I'm going wrong?

 

Many thanks

 

paul

PHD2_DebugLog_2020-04-11_201428.txt

[alan potts]

That is a shame and has spoilt a very nice image, I don't know your problem but there are some very clever and helpful people on here that I am sure will put you right, they did me, errrr, more than once.

Alan

[gasdoc]

Ha ha thanks Alan,

Yes I'm at the base of an all too familiar exponential curve - good fun learning but annoying when you have to wait all day to try again!

Paul

[almcl]

If you can post the guide log (rather than the debug log) we might be able to give better or more specific advice, but how are your imaging and guiding rigs attached to each other?  A picture of your set up might be useful, too.

[tooth_dr]

Just now, almcl said:

If you can post the guide log (rather than the debug log) we might be able to give better or more specific advice, but how are your imaging and guiding rigs attached to each other?  A picture of your set up might be useful, too.

I agreed, I think it could be differential flexure, if you've previously had good results.

[gasdoc]

Doh! Sorry - wrong file

Now attached.

The lens & camera are attached via a decent camera ball socket to the vixen plate which also has the guidescope on it. I tested it two nights ago with 2 min subs and the stars were pinpoint - which obviously made me think I had it all sorted so there must be something that I've done since then that has screwed it up. The conditions were better last night (less wind) so I can't even blame that.

Paul

PHD2_GuideLog_2020-04-11_201428.txt

[gasdoc]

[almcl]

At a quick look at that, there doesn't seem to be a focal length for the guidescope, so any statistics should be treated with caution (PHD2 doesn't know the arc sec per pixel, so can't report it).  It might be worth going through the new equipment profile with mount and camera connected (can be done during the day time)  and entering all the required data.

Not saying this is the cause of the problem, but eliminating any issues revealed may help.

Interesting mounting arrangements, and again this may not be the cause of the problem, but when I imaged with a 200 mm lens mounted on a single lens mount, differential flex was a constant problem.  I cured it with a second lens bracket, but that may not be an option in your case?

[gasdoc]

Yes I had issues with conflicts between EQMOD and starting PHD afterwards (guidescope view was pure white whatever I did) but after re-installing and ensuring I started EQMOD after PHD fixed this. Somehow I must have omitted the focal length of the guidance (but I’m sure I entered this).

Re: mounting camera - yes accepted this is not ideal but was really a test to see what an effective 1000mm FL would give me of galaxies without having to fork out for an enormous refractor.

 

I couldn’t find a way of mounting the camera securely to the vixen plate with the guiderscope attached - any suggestions very welcome though!!

 

paul

[almcl]

Mounting a lens with that long a focal length (I hadn't appreciated just how long you were going) is going to be interesting, I think.

There are a couple of threads about how people mounted their widefield setups here:

 

1 hour ago, gasdoc said:

 Somehow I must have omitted the focal length of the guidance (but I’m sure I entered this).

Doing it manually in PHD2 doesn't always 'stick' apparently, which is why the developers recommend starting a new equipment profile from scratch if anything changes.  There are also some calculations (below the user interface level) that don't get done with manual entry.

[gasdoc]

Many thanks for the advice - much appreciated!

It seems entirely plausible that its the differential flexure but that doesn't explain why it wasn't an issue the day before.

I'm going to try using my finderscope attached to my 6" SCT on a vixen mount tonight. Obviously the FL is even longer but on an EQ6R I would hope that this shouldn't be an issue.

 

One thought, could the lack of entering the FL of the guiderscope made the guiding inaccurate? Or was it just poor PA and I didn't realise this as I assumed the low error on PHD meant I was dialled in accurately?

So jobs for tonight:

1) Run sharpcap PA alignment routine

2) re-calibrate PHD now that I've entered the correct FL for the finderscope

3) try M101 again (clouds permitting) using the above arrangement

 

Best wishes

 

Paul

[almcl]

3 minutes ago, gasdoc said:

One thought, could the lack of entering the FL of the guiderscope made the guiding inaccurate? Or was it just poor PA and I didn't realise this as I assumed the low error on PHD meant I was dialled in accurately?

I don't think it will make a huge impact on guiding (calibration should have sorted out that) but it will affect the reported accuracy.  If you are going with the SCT, you might want to consider an OAG.  This helps get the guiding arcseconds per pixel ratio closer to the imaging one.  With your guidescope/long lens combo, you are imaging (if I've done the sums correctly!) at 0.87 "/px but guiding at 3.86 "/px a ratio of just over 4 to 1.  Much more than this (your SCT has1500 mm F/L, I think?) and the guiding will definitely suffer. 

[smr]

You've got a sturdy mount that's for sure. But just looking at the shot with the Sigma just looks like it undoes the sturdiness of the Mount. Doesn't look stable to me. 

You might be better off selling the Sigma for AP, there's a heck of a lot of lens elements the light is going through before it hits the sensor. I had the same lens but sold it to buy a refractor and then stuck the guidescope on top of that.

[tooth_dr]

2 hours ago, gasdoc said:

 I tested it two nights ago with 2 min subs and the stars were pinpoint

Maybe it was at a different time or a different target and gravity was working differently of the kit. I think you may struggle with this set up as 1000mm focal isn’t the easiest to guide well. 

[gasdoc]

Thanks everyone,

@smr I use the sigma a lot for my standard DSLR stuff (wildlife etc) so selling not an option

This was an attempt to see if I could get away without the added expense of a bigger refractor ( I am obviously lusting after an Esprit 100 but v difficult to justify to wife 😬 even after playing the "I need the stress relief during COVID" card [I'm nhs]).

 

So I guess I'll try with the SCT tonight, cross fingers, and see how that performs otherwise its penny saving time!

 

BW

 

paul

[smr]

Just now, gasdoc said:

Thanks everyone,

@smr I use the sigma a lot for my standard DSLR stuff (wildlife etc) so selling not an option

This was an attempt to see if I could get away without the added expense of a bigger refractor ( I am obviously lusting after an Esprit 100 but v difficult to justify to wife 😬 even after playing the "I need the stress relief during COVID" card [I'm nhs]).

 

So I guess I'll try with the SCT tonight, cross fingers, and see how that performs otherwise its penny saving time!

 

BW

 

paul

Good luck.

[alacant]

4 hours ago, gasdoc said:

EQ6R-Pro mount

Hi

You've a nice new mount but are using only a bit of it's capability.

Try connecting the mount directly to your computer, install ASCOM and EQASCOM, set the RA and DEC guide speeds to 0.9x and get both the guide telescope and the lens on the same dovetail without anything else in the way. 

Now make a new phd2 profile using the wizard, taking the default values but this time not forgetting to enter the guide telescope focal length. Recalibrate and take it from there. I think you're gonna be pleasantly surprised;)

Mechanical stuff:

RA:

PE: use the PPEC algorithm

 

 

 

 

 

 

DEC:

Adjust the backlash

 

 

 

 

 

 

HTH

 

Edited April 12, 2020 by alacant
consejos mecánica

[gasdoc]

@alacant many thanks - I presume those were my log files?

I’ll try all of this - seems Ppec algorithm is easy to activate In PHD and is backlash compensation a binary on/off or do I have to measure something and set accordingly? Will investigate tonight if skies are favourable 🤞

cheers

 

 paul

 

[alacant]

1 hour ago, gasdoc said:

my log files?

Yes:)

1 hour ago, gasdoc said:

backlash compensation a binary on/off

Hi. No, but first, you must adjust DEC physically at the mount. The worm to gear is either too loose or too tight. Adjust (a 2mm Allen) until you either take up the backlash or release the stiction. Try to use software backlash compensation as a last resort.

HTH

Edited April 12, 2020 by alacant
last

[gasdoc]

Hi all,

 

Had a quick chance to try again - see attached log file. Seems fine - didn't manually alter backlash as wanted to see whether other tweaks fixed the issue.

Haven't had a chance to look at the files once processed but raw subs looked much better.

 

I don't understand these graphs in any depth but my error was about 0.17 apparently - can't find that in the logfile though.

 

BW

 

Paul

PHD2_GuideLog_2020-04-12_115059.txt

[alacant]

3 hours ago, gasdoc said:

other tweaks fixed the issue

Hi

Looks like it. Despite the bad seeing/cloud/haze, your images should look very good. The guiding certainly was.

I'd still recommend EQMOD pulses however. As it is, you're gonna have to calibrate each time you move to a new target and automatic meridian flips will be more tricky.

Anyway, well done. Don't forget to post your image too:)

[ollypenrice]

It would be helpful to know the orientation of the image and its trails relative to RA and Dec. It's good practice to have your chip aligned orthogonally, meaning RA and Dec aligned with the edges. You can set this roughly by eye and then tweak it by taking a short sub while slewing slowly on one axis. This will produce trails at the angle your camera now is. Troubleshooting is then easier, identifying the problem axis if it is just one of the two, and you can add data to an image easily in the future.

I try to shoot orthongonally and when I dropped your image onto mine it fitted like this:

Unfortunately I can't be sure that my image really was orthogonal but, if it was, your trails don't arise from an error on a single axis. However, they are very systematic in being linear, so whatever was causing the problem was doing the same thing all the time.

Olly

[gasdoc]

@alacant thanks - I’m using EQMOD already but thought the pulse guiding was done from PHD? Why does EQMOD improve this? Sorry - any links to simple explanation would be good but I can google if not!

@ollypenrice I prefer your edit of my initial shot! I’ll read up about the orthagonal orientation (it’s a bit klingon to me at the minute!)

[alacant]

23 minutes ago, gasdoc said:

Why does EQMOD improve this?

https://openphdguiding.org/man-dev/Basic_use.htm

 

[ollypenrice]

1 hour ago, gasdoc said:

 

@ollypenrice I prefer your edit of my initial shot! I’ll read up about the orthagonal orientation (it’s a bit klingon to me at the minute!)

No, it's easy. It's basically the same as holding your daytime camera horizontal so it's aligned with the horizon. If you don't do that you can end up with a picture of The Leaning Horizon of Pisa if you're not careful. 😁 

Images which are orthogonal are simply ones which will be parallel with the earth's horizon when the object transits the meridian.

Olly