offaxis guiding - has anyone actually used one of these

[h0ughy]

http://www.innovationsforesight.com/

would be used in the 2.3m dome with the 14" meade, dslr etc?

[Alchemy]

No but the main consideration would be back focus flexibility including field flatteners etc, I think it would work well with those GSO RC scopes as they have plenty of space out the back to the imaging equipment. Don't know how much room you've got there on the 14.

[gbeal]

No from me too, but a natty principle. I agree with Clive, back focus is one of the issues. Certainly for the bog-standard ~55mm I doubt it will work, but any of the R/C scopes I have seen seem to have back focus to burn.
Did you glean a price?
Gary

[h0ughy]

cough - about a grand $US

[gbeal]

Cough indeed, but mere pocket money to you Huff.
I'd be headed in the direction of the MOAG to be honest.
In my own case though, the Starlight Xpress OAG is a wonderful piece.
Gary

[bert]

a grand?

Buy a Monster moag. The monster moag is very thin, and I'll bet precise parts will make an adapter to fit direct to the back of the scope, depending on what focuser you are using.

I have the use of mmoag and I own a moag. Unless you are using sbig AOL, the moag is a waste of space. The mmoag could be quite handy for your setup, especially with a rotator.

The idea of using ir to guide on..... you will lose a lot of sensitivity.

Brett

[gregbradley]

MMOAG is also about $1000 + Adapters so it will end up more. But it is a pretty good piece of equipment but it has a lousy focuser setup.
With my ST402 you can't use the helical focuser anyway as it is too far out.
But even if you could how does one use a helical focuser on a rectangular guide camera? What, spin the whole camera around including all the cables? eehhh no.

That means you slide the guide cam in the eyepiece holder until its in focus and mark the spot. Pretty crude.

Apart from that it works well although you have to make sure you camera adapters aren't tilted from hitting the prism holder. It is adjustable so no big deal but it did take me a few days to work out where the tilt was coming from in images when I first used it.

Also it can be tricky to setup. If the MMOAG is even a small amount out of focus the guide stars completely disappear making it hard to work out where focus is. They only appear when you are quite close.
This made it frustrating for me when I first got it. So much so I assumed it did not work properly and did not use it for many months.

This unit though looks heavy and looks like it may cause flexure as that heavy camera and filter wheel sitting right above it - that's a lot of pressure.

Also reflecting the light to the camera is saying the guide camera is more important than the imaging camera and the imaging camera can get the weakended reflected light along with all the artifacts, possible distortions and dust on the mirrors whilst the guide camera gets beautiful views.

I wonder though if it may be more compact in those situations where backfocus is limited so its not all bad.

I laugh at their 2 x 1 minute exposures added to simulate a 38 minute exposure. Do they think we're stupid?? Why not simply post a 38 minute exposure or was that all eggy :lol:?

Greg.

[JohnH]

Hmm, there is prior art...

http://www.iceinspace.com.au/forum/showthread.php?t=29541&highlight=beam+splitter

[gaston]

Let me help if I could.

I am the IF's CTO and following this thread's questions and comments I would like to answer/precise some aspects of the on-axis guider ONAG for clarification.

1) Back focus:
The ONAG back focus (BF) is 66mm (2.6"), which is about the same than the MOAG. The ONAG has been designed for medium to long focal, such as SCTs, it has been used successfully with various scopes like the Hyperion from Starizona, or Takahashi refractors.
The ONAG comes in standard with 3 T-thread extension tubes (8, 16, 32mm), as well as a 2" and SCT adapters to attached to scope visual backs.
SCT scopes are designed for a specific imager focal plane distance from their visual back to reach nominal performances (Focal, F number,...). For instance, according from Celestron, a C11 EdgeHD at prime focus requires a BF to the imager focal plane of 146mm (9.24"). Of course each scope as its own requirement for that matter. The ONAG BF will be part of this optical path, you may have to add T extension tube(s) associated with the imager BF to be close to the optimal BF. The ONAG BF is not likely to be the limiting factor in such case.

2) Dichroic mirror
The ONAG dichroic mirror, or beam splitter, BS reflects the light to the imager, like a star diagonal would do, while the near infrared (NIR) goes through the BS.
At 45 degrees of tilt there are unavoidable optical aberrations for the guider (geometric and chromatic) due to the thickness of the BS (3.3mm). Although this would not be acceptable for imaging, it has no significant impact for guiding since tracking software uses centroid algorithms. Only nitrocellulose pellicle BS would have a minimum of distortions in transmission, but they are very expensive and sensitive to vibration, microphonic effects, as well as temperature shifts.
The BS reflection is near perfect and since there is no refraction involved there is no optical aberrations either. Dust on its surface is less an issue than dust on the imager window since the later is much closer to the CCD ship. In any case the flat frame processing will solve such problems.

3) Test images
The IF's website presents results for a 38 minutes guiding session. Each sub frame was 1 minute exposure bin 1x1. The last image was made by stacking, without any alignment or registration correction (Maxim DL), the first and last sub-frames, 38 minutes apart. This allows to clearly see and quantify the tracking error, if any, using a 3D star profile or centroid for instance. The profile will exhibit a double peak in case of tracking issues. Stacking the 32 sub-frames all together will make this observation less accurate, and more subjective. In this current case it would not matter, since both images (first and last stacked) or (32 sub frame stacked) are the same, the error is too small to be seen.

4) Near infrared (NIR>750nm) for guiding
CDD/CMOS unfiltered sensors have their maximum sensitivity near deep red and NIR. The star spectrum is function of its color temperature, often given by its spectral class. More than 75% of the main sequence stars belong to the M class (<3700K). Let's assume to simplify that the sensor sensitivity is uniform across the range from 400nm to 900nm, which would be pessimistic for the NIR contribution. The visible range would be defined from 400nm to 750nm, the NIR from 750nm to 900nm. For a class M there is still about 70% of energy in the NIR in comparison with the visible range energy.
Since the ONAG uses all the scope aperture (F number) it will typically collect more energy than a pick-up prism used in most off axis guider OAG systems.

5) IR blocking
The ONAG requires a IR blocking filter for the imager. Color camera, DSLR, have a built in IR/UV filter. For monochrome cameras, with filter wheels, the LRGB filters do the job, if not an IR blocking filter should be added. Of course the guider must be free of any IR blocking filter, which is the case for most monochrome cameras.

6) Flexure
Like OAG the ONAG solves the differential flexure issue, while providing up to 1.3 arc-degrees (2m focal length) FOV to locate a guide star, This is done by using the integrated X/Y stage for on and of axis exploration of the FOV. This feature alone increase dramatically the likelihood of finding a guide suitable star, versus OAG.
The ONAG has been designed to be light and to minimize the mount extra load, it weights only 770g (1.7 lb).

I hope the above information will help to understand/clarify some of the ONAG aspects evacuated in this thread.

[Marke]

Should have a look at one of these if after an OAG I can highly recommend it : http://www.iceinspace.com.au/forum/showthread.php?t=77439
There is a bigger one again OA6 for very large sensors and cheaper
than MMOAG

[g__day]

Seems like a great idea, two question spring to mind:

1. Are all typical guide cameras sensitive to the pass through wavelengths your gear provides? For example I use a Meade DSI II Pro mono camera - would that work fine to guide with your gear?

2. How does one achieve fine focus on the guide camera? Is it a lot easier as you're right on the main light path - so will using a Bhatinov mask give a better illuminated image on axis rather than off axis - and how do you fine tune your ONAG camera's position once your main camera is correctly focused?

Smart idea!

Manual is very helpful: http://www.innovationsforesight.com/Site/Tecnical_Information/ONAG_UserManual_V2_7.pdf

[gaston]

We have several customers using the DSI II pro monochrome for guiding without any problem. The DSI does not have any IR blocking filter.

Focusing is a three step process. First you need to select the right extension tube(s), if any, to reach focus with imager and guider (see the DBF table in the ONAG manual) for your set up.

Now you can focus your imager as usual, then the ONAG included guider focuser (compression ring) is used for focusing the guider (about 9mm travel). You can use your Bhatinov mask for that exactly like for your imager, I do the same with my set up.
The ONAG manual and website FAQs give more information on that matter with some images, I suggest you have a look.

One thing I learned from experience and customer feedback is the importance of the right tracking software setting. The ONAG means you are guiding with the same focal length than imaging, for instance at prime focus.
The guider FOV is smaller than using a guide scope, therefore if you have experience with a guide scope you will need some time to learn the right tracking software setting (see the manual for advices). However this is not different than using an OAG for that matter, excepted that the integrated X/Y stage gives you access to your scope FOV. You can search for a guide star on, and off axis as well.

[g__day]

Gaston,

Thanks - my set up is Celestron C9.25 -> Meade motor focuser -> Lumicon OAG -> Canon 400D and Meade DSI II (using PHD or MaximDL to guide). I find lots of coma on half of the guide screen and lovely round stars on the other half.

An invention like your is quite interesting - I will watch with a lot of interest reviews of how well your product works in the field.

Very innovative!

Matthew

[gaston]

Matthew,

Your set up should work find with the ONAG.
Guide star distortions should not be much a concern for most tracking software, including PHD, or Maxim DL.

With the ONAG, when available, you have to select "self guiding" in the software setting since there is no reflection involved in the guider anymore, unlike OAG.

There is an independent ONAG review coming soon on http://www.cloudynights.com/

[Merlin66]

I can add a little info.
For the last few years I've been using a modified Vixen flip mirror - mirror removed and replaced with a beamsplitter plate, to guide my spectroscopes.
Depending on the quality of the beamsplitter you can get some image distortion ( and multiple images) in the guide port. This is not a major issue - PHD can always find the centroid and maintain excellent guiding.
I also use Al's reticule V3.2 overlay to define a "virtual" slit in the guide camera...dropping the target star on the "slit" guarantees that the star lies exactly on the entrance slit. Majic.
HTH
(Details of the mod in "Astronomical Spectroscopy for Amateurs", p 220)

[gaston]

Interesting.

Do you use a cold mirror as well, same concept than the ONAG?

Tis confirms my experience, a distorted guide star does not impact the ability of the tracking software, as long as the deformation remains constant over time and position.
To avoid multi reflections and possible ghost images on the imager side the BS most have wide band high quality multi layer coatings, including an AR coating on its back.

Imaging through any type of window with that amount of tilt (45 degrees) will compromise the picture quality (distortions and chromatic aberrations). This is why AO units, such SBIG AO8, AOL, are limited to few degrees of tilt/tip minimizing such effects.
Reflecting the light to the imager is a much better solution, today's high quality BS are almost perfect, so do star diagonals.

What is the cut-off wavelength of your BS?

[Merlin66]

ideally for spectroscopy we still want to access all the available wavelengths.
I have various beamsplitter set-ups:
A standard microscope slide - good, reflects 4% transmits 92% -unfortunately the lack of AR coating on the rear surface causes a 4% loss, and double guide images.
Edmund Optics 30/70 beamsplitter - very good
Edmund Optics pellicle plate (4/94) - very good (no secondary guide images)
Surplus Shed 70/30 beamsplitter - good
Surplus Shed 50/50 cemented prism beamsplitter - very good, but extra loss of light to the entrance slit.
IMHO a beamsplitter with rear surface AR coating and a 20/80 split would be close to optimum. Bright star images for guiding ( on fainter stars >10mag) and good transmission across a broad wavelength to the spectroscope.

[gaston]

Very interesting indeed.

If I understood correctly you split the all light in two paths. Transmission for your spectroscopy and reflection for guiding.
There is no splitting function of the wavelength (Visible, NIR for instance). It is not a dichroic BS, just a classical BS.
If so, with the IMHO 20/80, you use only 20% (reflection) of the incoming light (all wavelengths) for guiding, yet you can do so with a >10mag star.
Am I correct?

What is your guide star exposure time?

On the other hand the ONAG splits the light in function of the wavelength, acting as dichroic filter. It is another time of splitting.
>95% of the visible (<750nm) is reflected toward the imager, while >90% of NIR (>750nm) is transmitted toward the guider.
Unfiltered CCD/CMOS sensors are quite sensitive in NIR, since for most stars there is large NIR contribution and because we work with the same F number than the scope we can easily track with 11mag stars.

I do not have a good experience with pellicle standard BS, they do not handle well large temperature swings over time, nor dust. My equipment is always outdoor (under a dome). In summer the temperature can reach 40C, in winter -20C in common. Over time the membrane becomes lose and you can see some microscopic wrinkles.
Good dichroic pellicle BS are very expensive too.
Do you have some experience over a long time, large temperature range for those?
Maybe you do not need a very good star image for spectroscopy, the requirements are different than for imaging I would imagine.

[bert]

Hi Gaston,

I'm interested in you ONAG and ao guiding. On a typical sbig tc237 guide chip, what percentage of the light is ir with an average star, and how does it affect guide star sensitivity and accordingly the guide integration times with the visible spectrum sent to the imager?

Brett

[Merlin66]

Basically yes.
With the 70/30 available options (there's currently no 80/20 on the market so it was my "wish list"), from memory about 3-4 sec guide integration for about 9mag. need to check my notes.
You may be correct about the pellicle - I have no long term experience...

Re the NIR v's stars....
The early stars (OBAFGK) emit predominatly from the UV into the green.
The later stars (MRN, carbon etc) peak from the red into the NIR.

As long as we can hold the maximum amount of starlight in a 20 micron slit, for as long as possible - spectroscopy is happy.

[gaston]

Thank you.

If I remember correctly about 76% of the main sequence stars are class M, R, N stars.
How large is your BS?

[Merlin66]

Typically the beamsplitter is placed about 85mm in front of the entrance slit. In an f7 beam ( the fastest for the spectroscope) this means a minimum size of 12.2mm; I use about 20mm x 35mm plates were possible.

[gaston]

Do you guide with the same star which is analyzed by the spectroscope?

[Merlin66]

Generally, yes.
The target star is on the optical axis.
(However, if there was a nearby field star which was brighter than the target star - I'd probably go for that! Why pass up on the opportunity!!)

[gaston]

Hi Brett,

Most stars are from class M (<3700K) , which in NIR have about 70% as much as energy than in visible. Since the ONAG allows the guider to access the all scope FVO, and associated F number, most of the time there is plenty of signal level for guiding in NIR.

I used the SBIG remote guiding head with an AO8 unit, which is based on the TC237 CCD ship as well.
The performance is a matter of scope FOV and F number. With my C11 @F/10 I can track with stars form 9 to 10mag. Since the ONAG addresses a very large FOV for guide star searching there is a very high probability to find a suitable one for the job.
I use the AO8 at 3 to 5 HZ because faster does not make much difference. Unless you are looking at very close double stars, fast AO correction rate is not a good idea.

The isoplanitc angle for which the guide star has still enough correlation with the target is quite small, few arc-second. Above this atmospheric turbulences can not be corrected with an AO. As a matter of fact doing so will decrease the over all performance by chasing the seeing from the guide star.

To make a long story short do not use AO faster then few HZ. AO units for amateur correct only the mount left over errors, and "fast" noise, not seeing.
At that rate the ONAG should do find for most target and set ups.

I hope this help to answer your questions.

[gaston]

Hi Matthew,

There is a review of the ONAG at: http://www.astromart.com/articles/article.asp?article_id=836


Gaston

[h0ughy]

Well I would like to post an update to this saga - Gaston has provided for me the adapter to allow the connection of the OPTEC 6.2 reducer which I have shown in the pics below. Thank you Gaston for that part – its like having the key to the Ferrari but no steering wheel;)

Given the one day of clear skies since it arrived i have not yet got to use it but it will shortly come.

I originally had some concerns over the grub screws not being long enough but being the complete tool that I am I needed to get the right tool ;). The allen key wasn’t quite right and wasn’t allowing me to nip it up, 10 dollars at Bunning’s got me a new kit and the right key (a ball key) which fitted the job precisely. Its nipped up and will be installed very soon.

I am looking forward to this – in desperation I turned to planetary imaging while I was waiting…… but that’s another story….

The next update will be the unit fitted and working – hopefully.

[g__day]

I've been dying for ages to see an indepth review by you on this gear. It's very high up on my stuff I want list!

[h0ughy]

soon Matthew Soon;) - can you order me a week of clear weather and no family issues:lol:;)

[rally]

I might be missing something here, but is there no concern about the lost spectral band from say 750-1100nm.

If you look at the spectral response curves of a typical CCD used for astro imaging eg KAF series, it will show that the chip continues well into the IR
At 800nm the Qe is still around 30% or better, at 900nm its maybe 20% and at 1100nm its dropped off to nearly zero
eg KAF168093 or KAF11002

Of course each CCD has its own curve and some are better and some worse.
DSLRs unless modded, dont see this part of the spectrum anyway so no loss there, but I am referring to chilled astro cameras

That means, at least to me, that the IR end of the spectrum is now being cut off by the OAG passthrough filter/mirror from around 750-800nm upwards
Visually that matters not, but for CCD sensing its a loss if you have any desire to capture that end of the spectrum.

So is it a case of there being nothing but stellar brightness that we are losing here and therefore it doesn't really matter, or are we in fact potentially losing 30% of the useful imaging spectrum ?

I understand how it works well for guiding, but it seems at the expense of IR image capture.

Like I said I might be missing something obvious.

Rally

[Merlin66]

I don't think you're missing much.
99% of the time a UV-IR cut filter is used (400-70nm) for astroshots to reduce the "bloat".
Working in the NIR and above is a completely different ball game...as I've found with the spectroscope!!!

[gaston]

Hello Rally,

This a good question.

As a matter of fact CCD/CMOS sensors are quite sensitive in IR and this is a problem for visible photography in general.
One shoot color cameras, or LRGB filters for mono-chrome cameras, cut of the UV and NIR light, leaving typical the visible range available from 350nm to 700nm.
The main reason being most optics are well designed and optimized for the visible range, at the expense of the NIR (and UV) unless you are willing to pay much more.
This is especially true for refractors, or reflectors with any corrector lens, like SCT Schmidt’s plate, and FR.
Not to mention that reflective and AR coatings are not designed for NIR either. Optical design is the art of compromising (as elsewhere I guess), if you want well corrected optics for a given wavelength range you would have to give up something elsewhere, like UV, and/or NIR performances.
Even black aluminum anodizing is most of the time transparent above 750nm, which is another significant issue, not very well known but quite common.
This is why the ONAG uses a special optical grade extended range black anodizing, used in aerospace applications, which is much more absorbent in visible (up to 5x) and remains so way above 1000nm. It is more expensive but useful for working in NIR. If you have a monochrome camera, without a built-in NIR filter, you could experience those effects, by taking pictures with and without a NIR filter, and compare the results (sharpness, focus, …). Try also to image a standard black anodized aluminum piece of equipment to see what do you get.
In the past most SLR good lenses used to have a special NIR ruler to correct focus for NIR imaging, this is gone with DSLR since they are built with UV/NIR filters.
Chromatic aberrations and other problems become more pronounced in the NIR for most scopes, leading to larger FWHM (fatter stars with halo) and sometime coma.
Even if they would perform well in NIR the problem remains to image visible + NIR together, most of the time the scope focal plane is significantly offset between visible and NIR. Having a good focus, sharp image, for the all range could be challenging.


Clear skies, Gaston

[h0ughy]

well it has been an extremely long journey but tonight I achieved two things - focus of the main dslr camera and the focus of the DMK618k. That took me all of about an hour of fiddling - but essential. Now for the heartache - it got real windy - so much so i have chucked it in - even though its clear. Unfortunately the scope doesn't like the wind ;) . all i basically got was some 30 second shots of no real importance. That aside i have now come to the realisation that i need a real sensitive camera and for that i am open for opinions and suggestions. the dmk cameras i have are not up to the task.

for those interested - i will continue to persevere with the setup, but my limiting factor is the subject of the image needs to have some star nearby to guide with - real close as the field of view is very small with the guide camera. This is a real step up in guiding for me and i am finding it rather challenging:question:

the key i think for me is to get a very sensitive guide camera:shrug::question:

[Merlin66]

I ended up with the Lodestar guide camera - for spectroscopy...very effective and better than my original QHY5 or the ATiK16ic replacement.

[allan gould]

David
I've just ordered a new guide camera, the Starlightxpress Lodestar. It was a toss up between that and the Sbig STi. I'll let you know how the lodestar performs and it may be the guider you are after. I'm also interested in the ONAG and may get one down the track.
Allan

[h0ughy]

looks like i might have to get one

[h0ughy]

i think the principle and quality of the ONAG is terriffic, but its the imaging equipment that is stopping me i think:question:

[Merlin66]

Hmm
Just noted the ONAG connection....could it be the camera(s) you have are not sensitive enough in the IR region???

[gaston]

Hello David,

I would suggest the SX lodestar as well. It is very sensitive with an extended NIR quantum efficient.
A very elegant, compact well-made design, I love mine.
Tests done to compare with the ORION starshoot (a CMOS sensor) have shown a bout 10x more signal for the lodestar versus the ORION (binned 2x2 to match the pixel sizes).
Also the lodestar is a 16 bits camera, not a 8 bits, with very small dark current even at ambient and powered on since a while.
It works perfectly with the ONAG. Have a look to the ONAG's NIR guiding page, the chip featured there is the lodestar one (Sony Exview series, ICX 429AL):

http://www.innovationsforesight.com/NIRGuiding.htm (http://www.innovationsforesight.com/NIRGuiding.htm)

In general the ONAG will have a NIR efficient around 0.6 to 1.0 magnitude less than the full spectrum with such chips.
For instance with my C11 at F/10 I usual can work with 9th magnitude stars at 1 second guider exposure. If you increase the exposure time to 2.5 s you get one more magnitude, or 10th.
Very low noise chip.

P.S: If you select this one you will need a C to T mount adapter since the lodestar uses a C mount thread system:

http://www.scopestuff.com/ss_c2tt.htm (http://www.scopestuff.com/ss_c2tt.htm)

Part #C2TT

Clear skies!

Gaston :)

[h0ughy]

Allan, you obviously shopped around - how much and where from? do all the cables and adapters come with it?

[h0ughy]

thanks for the info Gaston

[h0ughy]

yes i think you are right:thumbsup: stupid me:question:

[Poita]

The lodestar works like a champion for me as well.

[h0ughy]

in the meantime would the ATiK IC16 work better? the lodestar isnt available at the moment - have to wait a few weeks. its ruddy cloudy again.... geez this is frustraiting - it must be the worlds longest scope setup ever?

[Terry B]

For a comparison I have now got my STi. It seems super sensitive compared to my old QHY5. I have it on my new LISA spectrograph. SBIG seem to have caught up with supply of them now. It comes with CCDSoft also.

[Merlin66]

Re guide cameras I went from a QHY5 to a DSI II, then on to the ATiK16ic and finally the Lodestar. (BTW it comes with all the cables etc -"plug and play forever")
The Lodestar was the best. The ATiK16ic, being cooled, did a creditable job and is still used....

[g__day]

Merlin - which DSI II did you use - the mono or colour and normal or Pro version? I find the DSI II Mono Pro quite a nice guide camera.

[Merlin66]

Matthew,
Yes, it was the DSI II pro (mono) - OK but the cooled AtiK was better...
(BTW the DSI II pro has the same chip as the Lodestar! Amazing what a set of different electrons can do.)

[g__day]

h0ughty,

Could you please give us a review of your experiences with the ONAG now that you'd had it for three quarters of a year?

I would be very interested in hear how well the sum of your guiding gear works.

I would also love to see shots of what your guiding screen (PHD or MaximDL)? looks like - at focus and either side of focus. If you frame a shot of M42 for instance how well formed are the stars in the ONAG around the focal point?

Many thanks,

Matthew

[h0ughy]

Mattttttthewew,

the weather has been cr@p and there have been two major events where the 14" was retired to the warm room. but the weather didnt allow it back out. I am hoping to get this back up and running soon. one thing that is a problem is that i think the 14 is too large for the 2.3m dome slit. i promise i will do it