Setting up for an allnighter

[skybadger]

I must confess that while I've been doing this lark for 40years and done visual all-nighters, getting instruments to image all night is proving hard..

For example last night, I opened up the obbo at 7.30, imaged Mars until 9 and then it was dark enough for deep sky. 

45minutes later and I still wasn't imaging, I had to break into the failed focusing routine, reboot to reestablish connection to the DSLR and fix some pointing issues. 

The the guiding needed recalibrating and then I was imaging. I came in at ten while I thought the guiding was working until I checked over rdp to see the dome had had a fit and gone pole-wards walkabout. 

I fixed that and babied the mount to good RMS Guiding on m33 until almost 2am and got called indoors. I really wanted to be able to leave it running all night.

Back to the crux, what prep do people do to get to the point where you can start a run and go to bed, expecting success and outstay in the morning ? How does that work ?

How did you develop your process and work out the kinks until they were reliable ?

What systems are you using for multi-target imaging ? 

Is it all scripted ?

For info I'm using maxim 5 to manage the overall setup with a homebrew dome and EQ8. 

Grateful for your input.

Mike

[hughgilhespie]

Hi Mike,

I can't offer any help but I can very definitely sympathise. I am also involved in trying to get automated overnight imaging going and finding it a real struggle. I have a Pulsar dome with homebrew automation for opening/closing the shutter and rotating the dome. I use the Lesvedome system to control it. I also have an EQ8 and I am running everything using SGP. For various dull reasons, I can't manage to stay up all night with the equipment, so I am limited to trying to get everything working in a couple of hours up until about 9.30 pm and then leave it running on auto. So far I haven't managed that. I am trying to do one thing at a time but as every single thing - focusing, guiding, plate solving, etc. etc. is new to me, my nice, simple plan for the evening rapidly descends into chaos. 

Last night was guiding night. I started with a firm resolve not to include anything that wasn't needed. So, I started by doing everything actually in the obsy - normally I would be running everything from the house via RDP - firing up Sharpcap to check the guide camera was working and to tweak the guidescope focus. The results were poor but I did my best. Then I used CdC to slew the scope to a target then manually rotated the dome until the target was visible through the shutter opening. Using Sharpcap I got a reasonable image from my main scope but sweet nothing from the guide camera. OK - check connections, try a different USB port, pull out hair. Then it dawned on me - the OTA was pointing in the same direction as the obsy computer and the miniguide scope I use doesn't have a dew shield - the objective lens is right bang at the end of the tube and was picking up loads of light from the computer screen. 

OK - now pick a different target, this time one in the East rather than the West. Slew, etc. and bingo! We have an image from the guide camera. Switch to PHD2 and - nothing! Camera connects but screen is completely white. Some time later - yes, I was using the wrong camera (QHY5Lii) driver. At last, some sort of image on the PHD2 screen. More faffing about to try and improve focus and eventually after doing dark calibrations I can actually start guiding. That was actually the high point of the evening. For the first time ever I was producing my own guiding graph and it was working! According to PHD2 I was getting an RMS of about 0.55 arcseconds which sounds pretty good to me.

So, in a spirit of wholly unwarranted optimism I decided to try an actual sequence with SGP. I chose M31 - well, who wouldn't! When I connected the dome driver in SGP odd things started happening - not surprising as I have no idea at all what I am doing. The dome decided that it needed to go the wrong (long) way round to my target. OK - why not. The 'why not' was because I had put the lens cap from my Tak FSQ106 on the little flat shelf sort of thing at the base of the dome, so it was carried around as the dome rotated. At least it was carried around until it hit my Dome Home Position Sensor, £22 of fragile plastic. I tried to leap to my feet when I heard the crunch but it was (a) too late and (b) I hit my head on the EQ8 counterweight bar.

Still - all in all I count the evening a great success. I have a PHD2 profile that actually works and RS have a really good delivery service. Saturday night is looking light there will be some clear skies in North Devon - here comes Plate Solving!!

Regards, Hugh

 

 

[skybadger]

Thanks Hugh. Frustration shared is what it's about

I have just boxed up a 12v relay unit so I can remotely control power to the bit s that need it. Once I have that cabled up, I'm in the position where the only thing I can't remotely do is unlock the shutter and open/close it. Oh, and uncap the scopes. Working on that bit. 

However right now I'd settle for automated capture and guiding working well 

It's a permanent setup so maxim just does a 'connect all', I connect the cameras and it's working.

Slew to target, plate solve, re-slew and solve again. Sync. Start sequence. 

The killer bit (assuming scripting of those bits) is getting good guiding and good guide re-acquisition. Brighter guide stars seem less susceptible to chasing seeing resulting in better guiding. 

I shall practise tonight ....

 

 

[hughgilhespie]

Hi Mike,

Same as you really - I used one of these http://www.robot-electronics.co.uk/files/dS378.pdf to control the obsy power switching. It lets me turn on the 13 volt DC power to everything and turn on or off the charger for the shutter motor battery - I have a bit of electronics that connects the shutter battery to the charger when the charger is running. When the charger is turned off, the battery connects to the shutter motor drive. That system works together with a sort of spring connector that connects the charger on the static wall to the rotating dome when the dome is in the home position. All the mechanical stuff works OK - or OKish anyway. The problems all come with the operator! In theory I can run everything remotely- except taking the cap of the Tak and guide scope. Actually I also put a cover (bin bag) over the scope and its associated computer and the Pegasus hub. Pulsar domes let in a lot of water when the rain comes horizontally as it often does in North Devon. So, one visit to the obsy in the afternoon, take off the cover and lens caps and back to the nice warm house for a comfortable imaging session.....................I wish!

Regards, Hugh

[skybadger]

So are you running the acquisition scripted or kicking off sequences remotely by terminal ?

 

[hughgilhespie]

Hi Mike,

I use SGP that allows me to set up a sequence in advance then automatically runs it. It is very powerful and gives you a lot of control over what events you want to include in the sequence such as when to autofocus, change filters, etc.

Regards, Hugh

 

[skybadger]

Just the shutter, the flip flaps and the filter wheel controls to take care of then....

Does sgp handle meridian flips well ?

Cheers Mike

 

[hughgilhespie]

56 minutes ago, skybadger said:

Just the shutter, the flip flaps and the filter wheel controls to take care of then....

Does sgp handle meridian flips well ?

Cheers Mike

 

I'll let you know when I get that far!!! I believe it does but I am hopelessly stuck trying to get my dome synchronising with my telescope. I know it CAN work - I have sat in the obsy and watched keep track faultlessly for hours but every time I try to set it up from scratch something goes wrong and usually SGP just freezes and I have to shut it down and restart everything.

But - on the bright side we had a clear evening last Friday and I managed to polar align using SharpCap Pro. SharpCap says my PA is 'excellent' and less than about 10 arc-seconds. I am yet to be convinced that it's that good but it is nice that something went to plan. After polar aligning I decided to try running an SGP sequence - and had the normal 30 minute descent into total chaos before I abandoned ship.

More trials tomorrow!

Regards, Hugh

[skybadger]

When you say you try to set iTunes from scratch, what are you setting and why, if it's already tracking the scope ?

 

[hughgilhespie]

2 hours ago, skybadger said:

When you say you try to set iTunes from scratch, what are you setting and why, if it's already tracking the scope ?

 

Hi Mike,

A bit confused re iTunes? 

What I am trying to do is to get to a level of automation that lets me start an imaging session remotely. So, during the day I will do the following:

1. Choose a target or targets

2. Set up a sequence of events for that target using SGPro. This will include an initial opening of the dome shutter and slewing the telescope / dome to point at the target. After the first slew SGPro will focus the scope and then do a plate solve and final, accurate, slew to centre on the target. Next start taking images according to the sequence I have already set up. 

Come the nice, clear, (rare) evening I go to the obsy and take off the scope cover (currently a large bin bag) and the lens caps from the main and guide scopes. Next is fire up the power and the computer I am using to control everything. This is a headless (no display, keyboard or mouse) mini computer mounted on the OTA. I connect to this OTA computer using Microsoft's Remote Desktop Protocol. I have another ordinary computer in the obsy that I use just to connect to the OTA computer so I can control things while I am in the obsy.

Once everything is running nicely, I can sign out of my RDP session running on the obsy computer and go back into the house where I will log onto the OTA computer using my normal home pc and then I can start the imaging session from there.

That's the theory anyway.

In practice, as described in my earlier posts, I hit problems as soon as I try to start my SGP sequence running. Basically the telescope moves - as it should - but my dome and shutter do nothing so the telescope ends up pointing at black painted fibreglass. This is almost certainly a case of mea culpa. I designed/built/wrote the software that operates the dome rotation and shutter opening and closing. Somewhere in my code there is a silly little mistake that stops it working as it should. Not an obvious mistake because when I send commands to the dome or shutter manually, it works just fine. So, I have been playing hunt the bug for what seems to be a long time. The good news is that it can mostly all be done during the day and I don't have to wait for precious clear evenings.

Regards, Hugh

 

 

 

[jjosefsen]

Really cool setups you have, I'm quite envious..

I don't have an obsy so the dome is not something I need to worry about. SGP should be able to control a dome through ascom though.

It sounds like you need to make a checklist for all steps needed to start a session. You can annotate it with problems as they arise, and then review them after. I'm sure you will get there eventually!

 

[skybadger]

ITunes -That's a phone spelling of ' it's a'. 

I have scripted out my sequence for startup and shutdown, the key frustrations are reliability and repeatability.  That's about building confidence in the process and tools and the potential things that can wrong. For example, having built a remote power switch so Ivan turn all the 12v things on using web calls, I find my usb3 still needs a press on the switch to wake it up!

[AstroAndy]

Hi Mike

Here is what I do; taking into consideration that I travel to dark sites, you might have to adapt these steps. Even though Murphy (and his law), like to travel with me occasionally, it has served me well so far.

Before I go out, I will have set the necessary software for coordinates (eg. latitude, elevation). I will also have checked the position of Polaris (in Stellarium) at about the time when I start my 3 star alignment. I will also have set a sequence of captures in my imaging software.

1) With about one hr. to sundown, I arrive at my site.

2)Using a smartphone compass, I set up the mount first, align it as North as I can (almost always finding the North Star in my polar scope on the first try), level the mount, put scope and counterweights on, whereby the scope is sideways (along East West, so's I can see through the polar scope). I may swing the scope to where my target is, to see that the camera, which is not connected yet, is not pointing downwards).

3) Then I connect all the equipment (except cam, guidecam and filterwheel), during daylight it's somewhat easier to see and place all the cablery.

4) Collimation, then I focus and center a bright star, or planet, in the main scope and finder (stuff always gets out of kilter driving the equipment around) Visually at first (much easier without cameras, having to run around looking at a computer screen, and never finding stuff through narrow FOVs). By then the North star will have become visible.

5)Then, adjusting azimuth and latitude, I'll put the Northstar "in it's place"with the polar scope, opposite the Stellarium position (because 2 lens optical systems are reversed with up / down, left/right, then put the scope in "park position".

6)Then, all the Equipment gets switched on (except maybe the dew heaters, which, however, are already connected, because once I start the alignment, I want to touch the scope as little as possible).

7) I run through the Synscan setup routine, choosing 3 star alignment (making sure the stars I chose form a big triangle that is not too steep, and whereby one star is past the meridian from the other 2; this eliminates cone error). On the first star, I manually center, rather than using the hand control. On the third star, centering visually as well as possible, the cameras and filterwheel go in and get connected (guide and main). Then I focus and center both. After that (and only after that), will I click enter on the handset (Alignment successful).

8)Drift alignment using a software called EQAlign (less so PHD, although I have successfully used that before, EQAlign works better for me). I usually do 2 iterations, Az, Alt., Az, Alt. (because a move in one axis contributes a small error in the other, so 2-3 iterations will make that negligible. I do the drift alignment through the main scope w/ the big camera rather than through the finder with the guidecam for more precise results (my image scale is 4 times bigger than my guide scale). I should note here that I'd rather do the drift alignment first, but the Synscan setup routine asks for the star alignment before it kicks in). I'm usually precise enough that the little position change doesn't matter too much. Then the dew heaters get turned on.

9) Then I slew to the object, recenter and refocus both scopes, calibrate and start the imaging run (unless the object is too far North, then I calibrate (with PHD2) as close to it as I can.

10) Shiver through a cold night watching graphs, or sleep in the car, occasionally checking how my guiding is doing.

11) If there is a semblance of life left in me in the morning, I wait up for twilight and shoot flats (as if, mostly, I make synthetic ones in PS).

- Andy -

 

 

[skybadger]

Hi Andy

I have a fixed pier in an obbo so life should  be easier than that. 

What is getting me right now is

1. Maxim isn't reliably Co nesting to the cameras

2, it will crap out in the middle of a sequence if I rdp into it from the warm room

3, if I slew to another target, I expect maxim to detect the slew, stop guiding and then pick up a new guide star once the slew has finished.

 

Still frustrating !