A Couple of Test Images

[avandonk]

I had a mate make up a restraint for the Pl16803 camera and optic train. The image is below. There is still slight flexure in the wood as the mount tracks. We are talking in microns here. I will get an adjustable version made out of aluminium and teflon pads. This mounts to the robust dovetail that holds the the whole image train.


NGC6188 X1.5 native size 7MB

http://d1355990.i49.quadrahosting.co...GC6188_NII.jpg

M16 and M17 mosaic 5MB

http://d1355990.i49.quadrahosting.co...16&M17_NII.jpg


It looks like the corner star issues are all down to flexure not the optic. It was far worse with my system because of the weight and size of the CFW.


Bert

[Mighty_oz (Marcus)]

Looking good there, wood always is a good alternative when testing, much cheaper
Images look great as well.

[alpal]

Hi Bert - some of those stars in the middle of the 1st frame
are not perfectly round.
Could you have some sort of guiding issue as well?

[avandonk]

Originally Posted by alpal

Hi Bert - some of those stars in the middle of the 1st frame
are not perfectly round.
Could you have some sort of guiding issue as well?

You are quite correct it is either guiding flex or image train flex or a combination of both. This distortion of stars was at an angle without the wood frame. It corresponded to the direction of gravity ie at the same angle as the latitude of the polar axis.

With the wood frame this distortion reduced and only showed in the vertical direction of the wood frame.

I was using a 100ED as guider on the side by side. I then mounted a 90mm MAK on top of the RH200 as a guider. It did not make much difference.

I suspect there is still movement of the optical train in the vertical direction of the wooden frame.

I am going to get a rigid aluminium frame made with eight or more adjustable bolts with teflon pads to hold the camera very firmly.


Here is an animated gif of two enlarged x3 crops from sequential 480s images that were guided by the MAK. 2MB

http://d1355990.i49.quadrahosting.co...2_06/480s_.gif

The same for 960s exposures.

http://d1355990.i49.quadrahosting.co...2_06/960s_.gif

One very good thing that has come out of this is that the distortion at the corners has disappeared. It was flexure of the image train relative to the optic not due to the optic. The factory alignment and collimation is spot on.

It is a very big ask to load an optic with this huge load of focuser, CFW and a very heavy camera and then expect a field that is far larger than the design criteria of the optic to stay perfect. If the camera was only 36x25mm with a lighter CFW rather than 38x38mm of the PL16803 and a very large CFW this would barely be noticeable.

I see this as a challenge to really push all the components I have chosen to work together by eliminating what is now just basically flexure. The combination of very fast F3 optics and a large image field is very prone to the slightest flexure.

Bert

[Paul Haese]

Hate to tell you this Bert but all the stars in the full size images are elongated in the same direction. Outer edges being more elongated than the center but not much more. That is more than likely an indication of flexure between the guide scope and the imaging scope. Are you sure you cannot use an OAG in this scope? There should be enough back focus to do this. From what I can see in your image of the setup it looks like the guide scope focusor is fully raked out too, which would be contributing to the problem. Not to mention cable management. Supporting the main camera is not going to work if the guide system is moving.

That said the field of view is great and the mosaic gives an indication of what you can do with the field sizes.

[avandonk]

Paul I hate to tell you this but the focuser on the 100ED is an Astro Physics focuser and it does not move with a 50g Lodestar.

Please read what I posted it was a 90mm MAK mounted on the RH200.

Only someone who is delusional would think that the stars recorded by a F3 optic are as perfect in the centre as the corners of a state of the art large sensor camera.

Advice is always gratefully received. It does help if it is well thought through.

Bert

[Paul Haese]

Hmmm, yes even Astro physics focusors can and do move Bert.

I am convinced though that if you sort the flexure related issue in the way described the very minor elongations in the corners or edge of field will be barely noticable.

I have considered this problem quite a lot and yes while an F3 is problematic it must surely be solvable but I do think not by the methods you are currently employing. The optical system and camera system you have bought should not need this method of control. Each are solidly constructed with screw thread design and therefore not likely to move. Mike's system is more likely to move than your system. I am more inclined to think your flexure is coming from the side by side bar system, refractor and the guide camera connections rather than the RHA and FLI gear. That is my considered option. And well thought through.

[avandonk]

Paul I am sorry that your your far greater knowledge is not available to me.

I mounted a dial indicator between the camera and dovetail plate. This dial indicator would show anything from +- 0.6mm movement to zero depending on the mounts position. It varied beautifully with orentation.

You just stated that an Astro Physics focuser will move and then in the same breath claim that a system you have no knowledge of will not move!

Thank you for your insight. I will take it on board.

Bert

[dvj (John)]

you could go with a much smaller guidescope, like a 50mm finder with a guider. Much like a I do with my FSQ.

[troypiggo (Troy)]

May I ask why you're choosing to do all of your testing using the NII filter? Would have thought most would choose Ha, which would be the dominant line in the targets you've been shooting?

[alpal]

Hi Bert -
is there any way you can use an OAG just for a test of your system?

Differential flexure must be the hardest problem to nail down &
an OAG would eliminate that from your list of possibilities.
Differential flexure can also be intermittent so that on one part of the sky
you get good results & in other parts you to have chuck out every subframe.

[Poita (Peter)]

If I was only looking at the image and didn't know the setup, I would have put big money on it being flexure between the guide scope and the imaging scope.

I know you have put an exceptional amount of effort into making sure that isn't the case, but is any form of OAG possible as a test to rule it out completely?

[avandonk]

Originally Posted by troypiggo

May I ask why you're choosing to do all of your testing using the NII filter? Would have thought most would choose Ha, which would be the dominant line in the targets you've been shooting?

I can get a colour image in about 120s with the stars just starting to saturate. What people have been calling HA is really a mixture of HA and NII. For example the Helix Nebula is about three times brighter in NII than HA.

I want to do deep NB and long exposures of about eight and sixteen minutes.

Bert

[avandonk]

Originally Posted by Poita

If I was only looking at the image and didn't know the setup, I would have put big money on it being flexure between the guide scope and the imaging scope.

I know you have put an exceptional amount of effort into making sure that isn't the case, but is any form of OAG possible as a test to rule it out completely?

With the lousy weather I have not made the definitive experiment where I take an image without guiding and image through both guide scopes with my GstarEX.

There is just no room in the image train for an OAG. I could drill into the rear bit of the Atlas focuser and install a small prism.

Bert

[troypiggo (Troy)]

Originally Posted by avandonk

I can get a colour image in about 120s with the stars just starting to saturate. What people have been calling HA is really a mixture of HA and NII. For example the Helix Nebula is about three times brighter in NII than HA.

I want to do deep NB and long exposures of about eight and sixteen minutes.

Bert

Yes, I'm aware that many Ha filters also overlap with NII and that yours doesn't because of its ultra narrow bandwidth. But the reason I asked is that most of the targets you've been shooting are Ha rich, not so much NII rich. ie you haven't been shooting planetary nebulae like Helix.

[Peter.M]

With a good polar alignment I would have thought the PMX would run for say 5 minutes without any guiding, wouldn't this rule out guidescope flex?

[alpal]

Bert

There is just no room in the image train for an OAG.
I could drill into the rear bit of the Atlas focuser and install a small prism.

What about if you uninstall the filter wheel?

Then you would have enough backfocus to run an OAG -
remember this is just for a test to eliminate differential flexure.


But:
The RH200 should have plenty of backfocus:
http://www.teleskop-express.de/shop/...strograph.html

The focuser provides 70mm of backfocus which is enough for most cameras, filterwheels, OAG and even AO

[avandonk]

The NII image of NGC 6188 would take about four hours with a FSQ106 to get equivalent signal to noise.


If I remove the filter wheel for some irrational reason it would prove nothing.


Bert

[gregbradley]

Use 2 autoguiders. One on your dovetail and one on the scope.

Set one to 45 second guide exposures and one for 1-4 seconds like normal.

Per SBIG the slow guide exposure guider would handle flex.

I don't think 70mm is enough for an OAG. They are about 40mm once you add in the adapter on each side of them.

The filter wheel is about 21mm and the Proline chip is about 33mm inside the body.

Greg.

[Poita (Peter)]

I'll be interested to see the results of the testing. I eventually gave up on a side by side system after pulling my hair out for six months. I know they can be made to work, but I couldn't manage it, I always ended up with problems that presented in the images that look similar to yours.
Eventually I made my own OAG that didn't eat up the space that the commercial ones do.

Something is definitely up, it will be interesting to find out what the culprit turns out to be.