Automating focus

[Paul Haese]

Ok, I am going to automate my focus on my scopes. I am planning on using robfocus units on both the Tak and the feathertouch focusors. Software is focus max and maxim. I plan on using ACP for automation.

My questions so far are these.

1. Do I need ACP before I can autofocus? Am I right in thinking that maxim combined with focus max can automate focus without my imput?

2. Do I need ACP immediately to get focus to swing onto a GV2 star and then back to the imaging target? I think this is a yes but wanted to know what others think.

3. What autofocus systems do other use? This seems the best system and I am trying to take the advice of Mark, who has been 100% correct with other automation systems in the past.

The main idea is that I want to be able to set this up and then not have to check focus every hour myself. Its all part of the automation process and leading to the end point. Data collection uninterrupted by SNAFU.

[bert]

Focusmax works stand alone, but you cannot get maxim to periodically call it in an imaging session. You maybe able to with a script, but this is beyond my capability.

I personally use ccdautopilot. I did try acp, but I found the GUI of ccdap5 a lot easier, on a clear night I have 2 complete observatories running with autopilot very successfully. The backup service of autopilot is excellent as well, it's usually tick the options rather than writing a program to do an imaging run like in acp.

My 2c.

Brett

[Paul Haese]

Ah all interesting Brett. I was originally going to go with CCDautopilot but was advised that getting a guide star after a flip was problematic. That also includes a rotation 180 degrees from previous setting. Are you using a rotator Brett and have you had this problem?

[bert]

Yes, I use a Pyxis on the eq6 rig. And I am looking for another for the remote rig.

And no problems with guide stars after meridian flip, or ever. Ccdap 4 I found to be difficult to use with a rotator. In version5 all the difficulties I had have been addressed, and the issues I had in 4 were due more to my laziness With the the pyxis software rather than the program.

John smith the author is fantastic with his back up service.

Brett

[Bassnut]

Aquirestar in Focusmax (with DL and platesolving) will automatically slew to the nearest star of a set mag (say 5), focus, and slew back.

[bert]

Aquire star is not used with an automation program, the automation program has it's own focus star selection routine. Ccdap's focus star selection is called skystar. In ccdap you select the mag range of the guide star, and it selects a guide star automatically at the mag and altitude selected.

Brett

[Paul Haese]

This is all good guys, CCDAP is cheaper too. What operating system are you guys using? I am using Win 7, I remember last time some things are glichy in various OS.

This will be all with a view to eventually going fully auto with the obs as well and imaging from the City on nights I cannot get down to the obs.

[Paul Haese]

Just read through the maxim manual on setup of autofocus and that all seems pretty straight forward. Working out the 1/2 flux might be interesting.

[bert]

I run win 7 64bit in both obs. As of about about 6 months ago, I think win 7 overtook xp as the best observatory os. Avoid vista like the plague.

Brett

[frolinmod]

My gosh Paul, that's the understatement of the year. :):):)

How do you guys who use ACP (and with the scheduler even more, ouch!) personally justify not just the initial payment, but the yearly maintenance payments too? I know Bob's gotta eat too, but wow.

[bert]

Hey, Im not trying to starve Bob. I bought pinpoint!

Brett

[Moon]

Paul,
My advice is to try and get Focusmax up and running first, in case you aren't already using it. It takes a while to get it running smoothly. If you follow this guide step by step, you'll be ok:
http://www.flemingastrophotography.com/focusmax.html
Also make sure you get the latest verison from the focusmax yahoo group (don't use the web download version - it's out of date)

After you have that running, then the next step would be ACP or CCAP.
James

[Paul Haese]

Thanks guys.

Ernie, yes the shock of the purchase price of ACP did seem quite high. I also bought Pinpoint which is nothing short of great.

Brett, that is good news about win 7. I have the starter pack loaded on my netbook which does the imaging run. I will need to get a USB hub to run all the etc equipment I plan on running.

James I will use that link (thank you) and I will check out the yahoo group for the latest version. :thumbsup: Then I will get either of those two. I am glad to hear that CCDap will work well too.

[frolinmod]

Wow, did you buy an ACP license? Which one? Did you buy the scheduler license too? That would seem like heaven to me. :thumbsup: At least in theory. Practice would seem to be quite a bit more, ah, complicated. If you bought those, it'll be interesting to see how well you get along with implementing them productively. :2thumbs:

Whoa, you're actually able to use Windows 7 Starter Edition for anything? I thought at least Home Edition was required to use Windows 7 for any useful purpose. I'll be darned. :confuse3:

[Paul Haese]

No way, not spent that money on a license. Not decided, but cost is a significant factor.

[jase]

As Fred and James note, you can autofocus with Maxim and FocusMax, but its not at the level of complete automation of which I believe you are trying to achieve. You still need something to trigger an autofocus routine and thats where observatory automation software comes in. James is spot on, you need to ensure FocusMax works flawlessly before overlaying any automation tools. If it doesn't work in this manner, automation is next to useless.

Do you have FocusMax performing the AcquireStar routine today where it platesolves and slews to an appropriate magnitude star and commence the focus run, then slew back to the original position? If not, get this sorted.



No. As noted above, FocusMax will do all that. It will platesolve to determine its current position and determine 3 or more candidate focus stars. It will then slew (and center if desired) to the first suitable star based on configured parameters such as altitude and magnitude. The focus routine then runs, once complete it will slew back to the original position. G2V star? The star doesn't need to be G2V. Usually something between mag 4 and 7 will work fine. Don't want it to be too dim as focus exposures will be too long, too bright and the stellar profile will not be calculated correctly. Most observatory automation software will simply pass off the decision to FocusMax as to of which star to use focus to FocusMax.



FocusMax is the defacto standard for autofocus systems. MaximDL has elementary focusing algorithms as does CCDSoft but they don't compare to the functionality achieved in FocusMax.



Yep. Easily done with observatory automation software. You are on the right track, just take small steps to begin with. When ready establish filter offsets to introduce even greater image acquisition efficiency with your focusing.

If its of any relevance, I use ACP Scheduler with ACP coupled to FocusMax. The set up is autonomous. I upload plans and out spits data. I don't manually start imaging runs nor determine what target to shoot and when. The Scheduler dispatcher process does this based on plan input such as moon up/down/avoidance, object altitude, etc. I recognise that such a level of automation is not everyone's cup of tea as some like to sit beside their imaging rig, watch guide stars, logs and make subsequent equipment tweaks. For me, its not the journey as I've travelled the road before, its the destination (output result) such as post processing where I prefer to spend my time.

[jase]

I should also add that don't make a swift choice on observatory automation software. Research the options, pro/cons. Much of it depends on what you want to get out the software. I've had the pleasure of using CCDcommander, ACP and CCDAutoPilot. They all come with strengths and weaknesses. I'm going to resist getting involved in a cost discussion, but simply state that people fail to realise the importance of software. They drop huge dollars on hardware (OTA's and big red mounts) yet can't justify spending a few thousand on software. In particular, observatory automation software makes your hardware investment work for you and can improve productivity.

[frolinmod]

What's even worse is that even though I write software and like to get paid, I still have that particular mental disfunction myself!

[Paul Haese]

Thanks all the replies. All very imformative and plenty to think about here. For now I will get the Robofocus here, install that and get my v curves working, with focusmax before looking at this next step of further automation.



Don't worry Jase. I have been looking and the purpose of this thread is to find out which is better. Price point is an important consideration, but I want the software to work more than the price. I wasted money on CCDsoft, so I have learnt my lesson here. :) Total smooth automation is my primary goal here. If it means I pay more for software then so be it.

[DavidTrap]

Paul - if you get a reply from Robofocus, can you let me know. I've been trying to contact them for the past few days (multiple e-mails to multiple addresses) without success. Anyone know if they've gone out of business?

ta
DT

[gregbradley]

Yeah, I sent Robofocus an email also about 4 days ago and no response.

Greg.

[jase]

Good to hear Paul. To be diplomatic, all observatory automation software packages will undoubtedly do what you are after. Completely automated imaging runs with everything in between, not to mention automated starts up, shutdown and bad weather recovery. How they execute the data acquisition process and perhaps the flexibility available will be the distinguishing points.

I liken CCDAP as a blackbox design where as ACP you can actually look under the bonnet to see its inter-workings. Don't like the flow of a control script? - you can make changes to it. For this reason alone, you'll find many research and education institutes utilise ACP. Not to mention it caters for multiple users on a system (commercial license required) and has a web interface which is ideal for operating over slow, high latency links. Of course, you can still use CCDAP in a similar manner via a remote desktop session. The leading rental telescope operations such as GRAS and Lightbuckets all run ACP but have heavily customised it to improve their operations. Bob is actively working on numerous new features listening to the ACP user base. You can view some of them here - http://dc3.ideascale.com/

[DavidTrap]

Not good news Greg... Anyone know of any alternate stepper motors that are compatible with the Robofocus controller I already have?

DT

Additional: I'll try to give them a call Tuesday morning and report back - found a direct number for them on a US Yellow Pages website.

[DavidTrap]

Doesn't look good guys - their website is not responding...

DT

[jase]

You looking to build a SuperRobo David? How good are your soldiering skills?

I'll let you in on a secret (of which I'm about to tell the world)... Having a heavy payload (ApogeeU16M with D9 cooling package along with 7 slot filter wheel) on the back of the FSQ, I killed the original robofocus motor. It died a glorious death with slipping, teeth grinding, crunching and a puff of smoke. Truth be told the original robofocus motors are crap. Their gear reduction is around 50:1 so don't have much torque and their step size is relatively large though but still sufficient for nearly all telescope critical focus zones.

Enter the SuperRobo...
http://www.hurst-motors.com/lsg42geared.html
>>>>>>LSG42012E98P<<<<<<
These are the same form factor as the original robofocus motor, so no need for another mount bracket, 12vdc, small step size providing improved resolution and a massive 300:1 reduction...enough grunt to scare a troll. FocusMax V-curves are now text book perfect with a heavy payload.

Disclaimer: This configuration is not supported by Technical Innovations. Modifications are at your own risk.

[Paul Haese]

David that might be a groan. I did not get a response myself but people do go on holidays too. Maybe that is the case here too.

I made an order from OPT and they did not mention anything about any problems. Otherwise I am gonna have to look at another system.

[cventer]

Jase

Those look like the goods. Did you order the motor from the same place in the link ? Any pictures of how you wired it relative to the original robo focus serial pinout ?

I have been running my robo motors at 50% duty cycle to keep the payloads from slipping.

[DavidTrap]

I'm interested. Do you just have to solder a DB9 connector onto these motors? I can manage that (or if not my father-in-law can - he was an electronics tech in a past life)

DT

[DavidTrap]

Might be a false alarm - their website is working again
DT

[Paul Haese]

I am looking at alternatives though with a kit for the TSA and autofocus from starlight instruments. It will cost a bit more but if I cannot have a robofocus quickly I will get that instead. I might be able to get a bit of a discount from them too. I have bought a few units from them now.:)

[jase]

OK, here's the catch. Hurst don't ship international. You'll need to find a company that stocks them. There are a few distributors in the US. I couldn't find any in AU at the time so I got a an astro imaging friend in the US to ship a few over. They ordered direct from Hurst.

Super easy to wire up guys. You can strip the same DB connector off, but I simply bought another one. You can also heat up the brass collar to get it off the shaft and reuse that. The wiring colours are also also identical to the original robofocus stepper so you can't go wrong. Attached is a shotty phone camera picture to prove its the same profile. I used the same bracket and even the same mount screws to fasten it to the bracket.

PS, Paul apologies for hijacking the focus automation thread. I guess this is still focusing related...:lol: Happy to change the topic back on to automation again. :D

[Paul Haese]

No its ok, I am looking at all options at present. This might present as an issue later down the track if Robofocus does go ahead.

[OzRob]

ACP uses Aquire Star to select an appropriate star and then everything else to do with focusing is handled by Focusmax. I even thing it slews the scope back to the target.

I think that for a remote setup ACP is the way to go. I am just trialing it at the moment but due to various issues it is a lot more robust than CCD Commander that I was originally using. Although CCDC quite happily would run a 6 hour multiple target run with no problems, albeit without guiding.

For my setup in the back yard I think I will continue to use CCDC. I like the way that you can set up every step. It is a bit of a pain to set up but it is more efficient than ACP. The thing I really like about CCDC is that I can change anything in the run (other than what is currently being executed) without stopping the run. You can even pause the run if you like. With ACP these things are not possible and you have to kill the run if you want to change something.

[DavidTrap]

Way cool Jase. I don't have a particularly heavy camera (QSI583) so I haven't had slipping issues thus far. The price of the motor is similar to the Robofocus motor alone, and you get the brackets and mating adapter too. If slipping is an issue, it might be worthwhile arranging a bulk shipment of these upgraded motors though.

DT

[jase]

Yep, you're right Rob. When you do terminate the run however, ACP writes into the plan where it got up to last so that when the plan is relaunched, it will kick off from where it was terminated.

You can modify chained plans however, providing the are not currently active. Any active plans can't be modified on the fly as they are put through a compiler at run time.

[jase]

Hmmm, most of my work colleagues in the US are not working Monday 10th as its a federal public holiday - Columbus Day. May want to keep that in mind.

[bert]

I should have said that ccdap does not use aquire star as it has its own routine.

CCdap5 parameters can be changed on the fly, without having to start over on the imaging run.

As I mentioned to Paul privately, it depends what you are after. As far as my computers skills, I'm no IT expert, thus I find the interface of ccdap5 simple and has the options for what I need. ACP scripting/plans was beyond my skills/patience.

CCdap does have the ability to run scripts, if I ever get the knowledge to use it!

Brett

[DavidTrap]

Thanks for that! Saved me an early wakeup on Tuesday morning. I'll reset the alarm for Wednesday morning.

DT

[DavidTrap]

Happy to report that Robofocus are definitely still in business. Jeremy answered my e-mail overnight (as did Precise Parts, so the public holiday thing might not have been a big issue). He said he'd been away installing a dome.

DT

[Paul Haese]

Yes I got an email too today. I have already made the order through OPT and might get it early next week.

[cventer]

I tried to order one before I saw this and Hurst cancelled my order. Looks like I need to find a US distributor who will ship internationaly. Anyone have a mate in USA they could get to buy us a couple and pop them in the post ?

[gregbradley]

I got an email from Jeremy of Robofocus also the other day. He must've just been away or ill or something.

Greg.