First impression of Starlight eXpress AO unit

[Luke Bellani]

Hi Guys,
I have now had a chance to come to grips with my new SX AO-L 'adaptive optics' unit and thought I would post same initial results and impressions.

Since I had an Orion SteadyStar AO unit before this one and posted a report on that, I thought that some may be interested in my initial results.

I purchased the unit from OPT in the USA. I have dealt with them several times before and have always been very happy as I was with this purchase.

The unit comes in a nice plastic foam lined case and first impression on opening the case was good.

All the necessary cables and lots of adapters that would suite most gear.
I used the 42mm T adapters to attatch the unit to my scope and camera.

I have a QSI583WS CCD camera and so got the AO unit that included the off axis guider (OAG).

The build quality of both the AO unit and OAG unit is very good.

It was very easy to assemble, fit to the scope's focuser and fit the QSI camera and Lodestar guide camera.

So everything went together very easily and when mounted to the scope, the fit was very solid and without any kind movement.

Still happy at this stage.

Then it was time to get it going and so I loaded the CD that is marked "Software and User Manual" into the PC drive.

There was no software on the disk only the manual PDF file.

Looking on the Starlight Instruments web site didn't help as SX has about the worst web site I have ever seen and so I couldn't find any reference to AO unit software anywhere obvious.

I dropped them an email and the reply was that I have to use the software that came with my Lodestar guide camera or MaxImDL to control the AO unit.
Bad luck if I didn't have a Lodestar guide camera or MaxImDL software, but fortunately I had both.

The software that came with the LodeStar camera is IMHO absolute rubbish and crashes if you as much as sneeze. The manual is totally out of date as well. So I fired up MaxImDL.

I had read the help files on using the SXAO unit with MaxIm earlier and so I had a head start there and managed to get it talking quite easily.

So then I proceeded to focus my main camera and that was OK and then focus the guide camera.

Here there was a snag.
There wasn't enough adjustment to alow the guide camera to reach focus and so that was that for the first night.

The next day I went out and found a set of C mount extender tubes that fitted OK and allowed me to get the guide camera focused the next night out.

So that I could isolate the AO and guide camera operation, I used two computers both running MaxImDL.
One to capture the image from the main QSI583WS camera and the other to get images from the guide camera and control the AO unit.

Apparently MaxImDL is not a multi-threaded program and so stops guiding during download of the main camera images.
I don't like this as the QSI camera takes about 20 seconds for a full resolution to download and so I will continue to use two PCs.

I did the same thing with the Orion SteadyStar and never had any problems that way.

I really like MaxImDL and have used it to capture images from my QSI camera for a couple of years now without any real issues.
However guiding is another matter.

Now my experience so far is that you can have a blazing star like Sirius in the middle of the image frame, but MaxIm will still try to select a hot pixel as the guide star.

Guide star selection is automated and the user cannot manually select the guide star.
A really dumb program feature in my opinion.

This forces you to take a series of dark frames and create dark frame groups in MaxIm for the guide camera to use.
Darks for all guide camera exposure times need to be taken and processed so that hot pixels can be removed from the camera image. This takes quite some time to do as exposures vary from about 100ms to seconds.
This has to be done otherwise your totally stuffed because MaxIm will select any hot pixels it finds regardless of how may real stars are in the image.
Well that has been my experience so far and this is the main reason I have always used the free PHD software to guide with.
It totally kills MaxIm for ease of use IMHO and doesn't sseem to have any problems with hot pixels.

Well eventually I managed to get everything working right and got both the AO and mount calibrated with the MaxIm AO control interface and away we went and guess what, it all worked very well. At first.

After a few AO starts and stops for different reasons and nothing to do with the AO unit, MaxIm couldn't connect to the AO camera any more and returned an obscure error message that it couldn't find camera number 3.

Now there is no such thing as camera number 3 and so this was a bit confusing and this is where having read everything I could get my hands on about the AO unit operation and having had an AO unit before was a great help.

As it turns out, if the AO unit's optical glass window that is used to perform the image corrections and is controlled by 4 stepper motors exceeds its normal range of travel, it can get jammed.

So... When MaxIm tries to centre the windows movement following the Connect button press by the user, it can't do it and so after a while of trying, it gives up and returns that stupid camera 3 error message.

By using a small screw driver to unjam the stepper motors (as per the instructions when you find them), MaxIm is happy and away we go again.

This problem I put down to teething issues and shouldn't be a real problem now that I understands its cause.
I don't expect it to happen very often.

Now for a couple of comparisons with the Orion SteadyStar AO unit.

As far as the build quality is conserned, the SXAO unit is better quality.

The Orion unit is OK but the OAG is a real piece of junk and lets the rest of the unit down badly.
A real stupid way for Orion to try to save a few dollars.
I had to make some significant mods to it before it was really useable.

That said, the Orion SteadyStar always worked very very well and I never had any issues with jamming stepper motors.
Although this possibility is mentioned in the SteadyStar user manual, the quality of which kills the really poor user manual for the SXAO unit.

Now where the SteadyStar kills the SXAO unit is in its control software.
The most important bit me thinks.

The SteadyStar control program although still needing improvemet, is excellent and very easy to use. Much easier to select guide stars than with MaxImDL.

I used my SteadyStar AO unit with the Lodestar guide camera for about a year and never once had any issues with the program selecting hot pixels when bright stars were available in the image.

It also has a facility to take guide camera dark frames when needed (if all stars are faint) and this very easy to do as well.

I could go on some more but I think I have said enough for now. At least unitl I use the SXAO some more. If the weather ever clears that is.

By the way, I didn't sell the SteadyStar because of any issues or dissatisfaction. I just wanted to a larger format unit and decided to try the Starlight Instruments SXAO-L unit.

Overall my initial impression of the unit itself is good, but the MaxImDL control software while easy to get going, is not easy that easy to use due to the issues of selecting hot pixels instead of real stars.
Going through the lengthy process of taking and processing guider dark frames largely resolves the hot pixels issue. But not completely.

I believe that MaxImDL now supports the SteadyStar as well, but if its guide star selection method is still the same, then I would just stick to the Orion control program.

Cheers,
Luke

[gregbradley]

Oh man don't you love gear that doesn't work right out of the box!

Thanks very much for this excellent review. You had me laughing there several times!

I have an SX AO unit on my wish list but was a bit sceptical of SX's ability to produce a high quality product.

By the way I use a Lodestar also and with CCDsoft. CCDsoft will also chose a hot pixel over a star but not a bright star. I use autodark and it kind of works but because the Lodestar does not have a shutter, the autodark also mostly removes the guide stars as well! It still manages to locate the remnants of the guide star and work though. If the star is bright enough its not an issue. I tried library darks and whilst these work just taking an image with the Lodestar as soon as you click on autoguide it won't use them. Perhaps glitch in the CCDsoft code.

I agree SX website could do with an improvement but its really no worse than most European websites.

So if I got one I can expect a few issues but as long as they are surmountable then its probably still worth it.

Gee, where's the FLI AO and OAG unit eh? There's a real need for a high quality unit still.

Greg.

[Poita]

I manually select my guidestar in MaximDL
To do this, select the guide star by clicking on a star in the image, do not draw a box around it, that doesn't work for some reason. Just click on the star.
Do this when you are in the guide tab and have clicked on start, after it auto selects a star, just click on the one you actually want.
Then click on calibrate and click on start again.

I don't know if having an AO unit affects this?

[Poita]

If you want to get away with one PC, you might want to try running two copies of MaximDL on the one machine, run the guider one sandboxed and assign to to one of your cores.
Sandboxie might allow you to do this.
http://www.sandboxie.com/

[Poita]

From the MaximDL help page:
(see the part in green)

"http://www.cyanogen.com/help/maximdl/AO_7_Locate_Tab.htm"

Does this work with the SXAO to let you choose your guidestar?

There are also plugins available for advanced guiding, I haven't tried them though:
http://winfij.homeip.net/maximdl.html

[multiweb]

A great review Luke. :thumbsup: I have one unit too but haven't been using it as much as I would have liked because of the beautiful weather we've had in Sydney. I used PHD to guide through a finder with a QHY5 then had the AO on its own loop with the lodestar on a guidestar for the sole purpose of correcting micro jumps. This way I kept the AO guide star always within the range of the AO FOV and didn't have to bump the mount, because I couldn't got ST4 direct to my old Gemini at the time. For this I used the SX software that is very basic as you've pointed out but it did the job just for that. I agree MAXIM can be a pain to use.

[RickS]

Thanks for the interesting post, Luke.



According to posts by Doug George on the Maxim group this is intended behaviour. The USB traffic required to continue guiding has been known to cause artifacts in images downloaded from the main imaging camera. Personally, I'd prefer this to be an option that can be turned on if needed.



Yep, the automated guide star selection in Maxim sucks :D As Peter pointed out you can override this, but my experience was that while it works for guiding with the AO, it doesn't work when calibrating it. Maxim would let me select a guide star but as soon as I pushed the button to calibrate it would go and select a new one.

I made an attempt to compare OAG and AO guiding here:
http://www.iceinspace.com.au/forum/showthread.php?t=85338
I was hoping to perform some more tests, but between a long holiday and rubbish weather I haven't had a chance...

Cheers,
Rick.

[leinad]

Certainly can be a frustrating experience, especially if the mount suffers some backlash when doing a calibration routine.

Bad pixel maps will assist with the annoying auto star selection on hot pixels.
You cant select the star during the calibration routines as far as I know, it selects the best star it thinks has the right brightness level to keep track of. You can though select a star when you Locate, then Track.

Also you dont have to calibrate ao and drive on a star in your target fov; you can do this away from the target fov with a suited nice bright star, then slew back frame, locate then track.

With my VC200L at f/9 and the 583, I've always had issues with 1x1 binning, 2x2 binning works well with the AO tracking. Havent quite sorted that one out yet, but might be something to do with star size in guiders fov and flux.

There's a few perks alright with Maxim software and control; and takes a bit of patience and persistence, but once you get it working its a fab tool.

[Luke Bellani]

Hi Greg,
Thanks for you kind comments.

The SXAO unit quality is good and you wouldn't be disappointed with it.
It's the quality of the control software that is the problem in my opinion.
However, once the issues are understood, it does do a good job.

From what I have read so far, generally opinions of other users are positive.

Cheers,
Luke

[Luke Bellani]

Hi Peter,
Thanks for the tip.
I'll try it out tonight.
So far it looks like the sky will be visible. So far....

Cheers,
Luke

[leinad]

Luke,
The advanced guiding plugins and std guiding options dont work with AO guiding, as the AO has its own control window. They will only work with native guiding.

[Luke Bellani]

Hi Peter,
Thanks for the link to the plug-ins. I'll see how I go with a couple.

Also from the MaxIm help, where is states "If necessary, a dark frame will be taken and subtracted from the light frame." I think it means that If necessary, THE USER should take a dark frame to be subtracted from the light frame. I definately doesn't try to do it by itself.

Cheers,
Luke

[leinad]

That's right, well for native guiding really.. But if you've already taken a dark library, and have full calibration enabled in the AO Maxim Options it will use the darks to subtract from the lights. A bad pixel map is also recommended if the AO control is trying to select the hot pixels(real pain). This can be selected after you saved it in the Set Calibration.

[Luke Bellani]

Hi Mark,
Yes, I wondered about using the AO unit that way as well.
I always had good results with PHD once the settings were right and if it ensures that the AO motors don't go near their limits, then that would be a good thing.

Cheers,
Luke

[Luke Bellani]

Hi leinad,
Thanks for the tip about the plug-ins. Saved me some time.

I usually do my calibrations on a bright star away from my image target but MaxIm still wants to select hot pixels.
I have now created some dark frame groups and so will see how I go tonight.

Yep, MaxIm is good software but I wish they would allow manual selections in some of their tools.

Cheers,
Luke

[multiweb]

It works quite well keeping things separate. I guided with a QHY5/guider in PHD. I had the sx program running the AO and lodestar on a closed loop all the time and I was capturing with Neb2. Three separate process unrelated.

[leinad]

3 words, "bad pixel map".

[leinad]

errr... how exactly are you guiding with a qhy and AO/lodestar at the same time. :P

[marki]

To avoid hot pixels being chosen, open options in the guide tab and check simple auto dark, works everytime as long as you are not trying to ring your guide cams neck with high gain settings, better to increase the exposure time. To choose your own guide star simply click on the star you want to use. Remember Maxim does not like the guide star to be too close to another star. As for the guide cutout during download surely if your mount is in good order and properly polar aligned it should be capable of tracking a star unaided for the time it takes for a sub to download. If not you need to improve you mounts performance/setup.

Mark

[multiweb]

You guide with the QHY5 only and pick a guide star in the finder. Plenty of them. All the lodestar does is 'read' its own guide star from behind the AO glass plate to compensate for micro-errors. PHD will keep the whole rig doing standard guiding so basically the FOV of the AO/Lodestar doesn't need mount bumping because it's never going to change that much. Not outside of the mechanical limits of the AO plate anyway. Then it's up to you to capture with an external imaging camera and other program if you wish.

[RickS]

That's a neat idea, Marc. Have you tried dithering in Maxim just using the AO? Seems like it should work.

Cheers,
Rick.

[multiweb]

No I haven't. That was the only way I got it working at the time so it's just specific to my gear and not recommended as such. My old (ancient) gemini box didn't have the required auto-guider port wiring to work with ST4 output from the guider.

Using the AO in Maxim requires you to capture imaging subs in Maxim as well and I used nebulosity. If I could just guide in Maxim with the AO then I'd do it rather than use the SX software. But as basic as it is the SX software works on its own. Can you just connect the guider in Maxim without setting up an imaging camera? I don't think the selection drop down in Camera 1 has an option for no camera. Only Camera 2 which is the guider usually. :question:

[RickS]

Sorry, misunderstood what you were doing, Marc. I always thought it would be interesting to guide with PhD and capture with Maxim if there was a simple way to get dithering to work. I already have the AO working with everything under Maxim control.

Cheers,
Rick.

[Poita]

The plugins don't work with AO, but clicking on the star you wish to guide with does work.

[Geoff45]

I have to say my SX AO unit worked like a charm right out of the box, but I've only used it on one night. Fingers crossed that it will work this coming weekend. The hot pixel problem pretty much went away with dark calibration of the guider image.

Equipment: QSI540wsg. lodestar, MaximDL.

I found the following page very useful and much easier than the SX manual http://www.rarcher.net/sxao.html
Geoff