Skip navigation

Category Archives: Equipment

It’s been around three years and five months since I started this blog. Over that time its been clear that to make the most of the limited clear skies we have in the UK you really need to be into automation. You need to build a system that will maximise the time you have available. I’ve built up an imaging setup that can run automated without much need for me to go outside besides opening up and checking it. Sorted, right?

Even with this type of setup in place imaging in the UK is very difficult. First of all you have to battle the elements: the skies in the UK are renowned for driving Astronomers to despair, the estimated percentage for clear skies is around 15%. Next up is everyone’s pet peeve: light pollution. I live in an suburban location just outside Edinburgh, and enjoy some of the finest light pollution my country has to offer. I’ve done battle with my local council, which eventually got me a shade on a street light that overlooks my Observatory. This has made a big difference, but the worst of it is being stuck between a lit motorway to the south and this monster to the north:

Even being a good mile away from the bridge and facing the opposite way, on a cloudy night you can see a beam of LP projected right across my field of view. Not the best. Next problem is this field of view. I have two options where I live, in front and behind my house. Being N-S orientated, the back gives a north view, the front gives a south view. The front became the only option as most of the targets are in the southern part of the sky. This southern view is further restricted by trees and the house itself, with the meridian splitting the view, if you have ever mixed LP & sky gradients with mid-session pier flips you will know what a nightmare this can be.

Added to all this, the biggest trap in Astroimaging is gear lust. To improve your images you have to be constantly working on your setup. Buying, selling, building, having various development projects on the go and using precious clear spells to test these out. This got to the point with me that I had my head in the technicalities so much that I really lost touch with what I was trying to do. So many times I would get a clear night and then realise I had spent no time planning and researching what targets to go for, or learning about the night sky.

So, what is the solution to this? I’ve toyed with the idea of my own fully remote observatory in the UK for a few years, it solves the LP and FOV issues, but unless you can afford to have it hosted at a pro site, you are locked into an incredibly difficult technical challenge, and still have to deal with the UK weather.

The nice folks over at have provided a great solution to this. Imagine being able to use kit remotely not in your garden but in the best locations around the world. Add to this being able to choose from a huge variety of scopes with different fields of view. They currently have 3 sites in operation: Australia, New Mexico and Spain. At each of these sites is a variety of research grade scopes that you can book time on or use when free. It opens up fantastic opportunities to see objects you can’t from the UK, and have access to them all year round.


This is all based around a monthly subscription plan, where you buy points to use on the scopes. Each of the scopes have their own points per hour rating depending on how powerful they are. This can be expensive, but considering how much you need to spend to get into Astroimaging and the quality of the kit you can use, works out as being reasonably good value.

It’s the type of thing that’s not for everyone, as you don’t get to use any of your own equipment. For me, it’s the ultimate experience. I’m in the process of selling all my equipment to concentrate fully on this. The final point that nailed it for me was being able to spend the afternoon imaging in the southern hemisphere, then go out to the pub that night. No more tacking the elements, technology and giving up precious spare time! I’ve had more fun in the past month than I have since I started getting into imaging. You can the results in the past few posts. That says it all.


I’ve never got along well with flat frames, I’ve always had trouble with getting the levels just right. It’s tempting just to fire it all through PixInsight and let the background extraction tools take care of it, but it doesn’t get better than having good calibration already done, espacially when you have light pollution to battle on top of that.

So, I thought it was time to get to grips with this part of imaging, so I bought a Gerd Neumann Aurora Flat Panel.

The first problem when it comes to flat frames and the Atik 383L+ is just what is the exposure length I need to get above to overcome any smearing from the mechanical shutter? I’ve heard 0.5 – 1 secs as a rough guide, but it should be easy to work out and have a measurable effect on the data.

Well, here is a range of exposures through a Red filter using the Aurora. Analysed using CCD Inspector. The resting position of the shutter is obviously on the bottom right looking at the first exposure. With increasing times this gradually decreases the effect on this corner and starts to show the real vignetting in the system on everything above 2 secs, it may be possible to get away with cutting it fine somewhere between 1 and 2 secs, but for the sake of headroom, 2 secs will be my minimum exposure length.

Scope: Borg 101EDII (@ F5.3)
Filter: Baader 1.25″ Red CCD

When I built my automation system, the one thing that became apparent is how poor the user interface is. Running multiple apps, each with multiple windows of their own, is a bit of a mess and quite hard to keep track on what is happening at any given point. ACP presents a nice user front end in a website, but is priced at the higher end of the market where people can have the open of renting out their Observatory. There seems to be no open source middle ground for hobbyists to explore.

I’ve started to develop an open source solution that can interface with automation applications in different ways. The idea is that if you can have a client that brings all your applications into one standard communication protocol, then this can be used to interface with anything; a web server, database, or external controller. This could simply be displaying information on a website as ACP does, but with touchpanels now being ubiquitous in everyday life, why can’t your remote Observatory have an interface on an iPhone or iPad?

The Mobai Windows Client will run as a console application on your Observatory PC and connect into each of your applications (supported so far are MaximDL, CCD Commander, ASCOM mounts, the Aurora Cloud Sensor and MJPEG IP webcams). It will then send all the stats for these applications over the network to another device. What I plan to develop with the client is that you can pick and choose your configuration. This could mean support for different automation applications, and the choice of how you want your user interface to work, this could be a simple webpage, or an iOS / Android application. So far this information is read only, but it also has the ability to send commands into your applications from the interface and there is a lot of room for development on this front.

The way I am creating a user interface is to send all the client data to an AMX Netlinx controller. This sits inside the house on the same network as the Observatory PC and can also control lots of devices via RS232, relays and digital sensors. These are high-end AV & Industrial controllers, but can be found occasionally on eBay for a few hundred pounds. AMX controllers can also be connected to their own touchpanels, which again are quite expensive, but some cheeky Russians have brought out their own AMX compatible iPhone, iPad and Windows application called iRidium that allows the use of AMX touchpanel files normally meant for proprietary devices. The license for this is a few hundred pounds, but in effect this means quite a high-end automation system for not a lot of money.

For more info and to try the source files visit the Yahoo Tech Group:

One thing I noticed with the 77ED was that at least one of the colours in an RGB run would be slightly out of focus, I noticed this on a few images and checked it all with a bahtinov mask. Going between RGB and having to change focus all night is not something that can really be done manually. There is also the problem of compensating for temperature changes over the course of a night. Focusing is the last part of my workflow that is not automated, so it was time to build a full auto focus system.

Accurate auto focusing is not really possible with the helical focuser that original came with the 77ED, so the first step was the upgrade this to the M57 Feathertouch. I’ve always had difficulty with heavy imaging trains and 2″ eyepiece holders, I’ve never found them to be very reliable and can introduce tilt quite easily. The M57 version of the Feathertouch allows the Borg adaptors to screw directly in making it all rock solid. I added in a camera rotator as well, it won’t really work without this as it locks your CCD angle in wherever the threads tighten up.

I then noticed a great thread on Stargazers Lounge by Neil Chase, who had also started his own automation group (SGL Observatory Automation). The idea was to use a combination of Arduino and Motor Driver to build an ASCOM compliant Motor Focuser. The programming side was really straightforward, it just took a wee bit of tweaking some of the variables to match what I was doing. The Feathertouch doesn’t have any fixing points so it took a bit of trial and error and a few bits and pieces from eBay to properly mount a stepper motor on it’s own frame bolted onto the main section of the scope. The result works quite well on a artificial star, now to just get it working with AquireStar to automatically find focus stars during imaging runs.

Astronomical darkness returns to 55° North on Tuesday so it’s time to get my finger out and finish a few summer upgrades. The first and the most major is moving from the 77ED to the 101EDII. It is compatible with the same components as the 77ED, so all you have to do to upgrade it is to change the front objective.

It has a focal length of 530mm compared to the 77ED’s 430mm so gives a slightly tighter FOV, and is also slightly faster with the 0.85x reducer giving it F5.3 compared to F5.6. I’ve still to run some flat frame tests but still being under F5 I’m hoping 1.25″ filters will still just about do the job.

I’ve wanted one of these ever since seeing the excellent results people were getting with them. Being an Astrograph it is perfectly suited for imaging, giving a nice flat field over up to 35mm full frame sensors. I’m 0.8mm off the 55mm spacing requirement but even with that the curvature is extremely low, if my guiding is up to scratch this should result in much sharper images.

The major advantage of the Borg system is it is completely modular. Each component can be broken down and re-arranged. It can even be upgraded to the 101mm version just by changing the front lens. I’m running it with a 0.85x reducer at the moment, so will at native F6.5 this brings it down to F5.5. It has a super reducer that can bring it down to F4. I will eventually buy this and then switch from 1.25″ to 2″ filters.

I’ve just been giving it first light and been playing with the helical focuser, which being more like a camera lens focuser is completely different to the usual crayford types on most scopes. So far it’s turned out to be excellent, it can take the weight of the imaging train easily and is nice and smooth and can be locked off once set.

Living of the edge of a big city with a big airport to the south means loads of light pollution in my area. Quite a lot of my clear nights also coincide with the moon being in the sky, so with my QHY8 I found myself doing more Hα imaging than anything else. The obvious move was to sell it and get a mono CCD. I was lucky to get an newly released Atik 383L+ in the second batch, they have now just gone out of production due to component shortages.

It uses the Kodak KAF-8300 sensor, which has been available for some time on the QHY9. This sensor has been a challenge for CCD manufacturers as being a full-frame sensor it requires a shutter to block out light before and after taking images. It is also significantly noisier that a lot of other sensors, so relies on as much cooling as possible and the use of dark frames.

The 383 has set point peltier cooling, and it can get 40 degrees below ambient temperature in a few minutes. The software was really easy to install and has worked flawlessly for me so far with the plugin for MaximDL. The download times are a lot faster than the QHY8, as it operates at full USB 2.0. It’s nice and light as well at only 500 grams, putting less pressure on the imaging train, which is really impressive for a cooled CCD.

There has been a lot of interest in the CCD due to the fact that the sensor is quite close to the front mounting. The majority of mega-pixel CCD’s are too big to use with nice and cheap 1.25″ filters. The KAF-8300 sensor is 22.5mm diagonal, which is just small enough to be used if you can get the filters close enough to the CCD. It also depends on what f-ratio of the optics are, as the steeper the angle of incoming light, the more vignetting there will be on the chip. I’ve have been able to get a good result on my first few pictures using a special adapter from Scopestuff. I will be running some further tests to compare this usage with using a filter wheel at f5.

Around 6 months ago, I ran a trial on Global Rent-a-Scope (GRAS). This system runs on quite high-end imaging automation software called ACP Observatory Control. It allows a user to control an imaging telescope from anywhere in the world. The obvious advantage of this is that from a cloudy UK during the day you can be controlling a scope in crisp clear skies in Mexico or Australia.

Seeing this in action gave me a lot of inspiration. Building this sort of system for imaging in the UK gives you a huge advantage, you can use every single bit of available darkness and clear skies, without losing sleep. ACP involves quite expensive software and hardware, but given automation is my profession, I thought, it can’t be that hard…right?

I’ve been building and testing for following system for the past few months. The two main programs it uses are CCD Commander, which is similar to ACP ( and Homeseer, which is a Home Automation program (

Software Configuration: CCD Commander

CCD Commander provides the automation of the all the mount control, auto-guiding, and imaging CCD functions. It can also automate auto-focus by using Focusmax, but I haven’t fitted stepper motors to my focusers yet. Once you have made an imaging plan for the night, it runs each of the actions.

CCDC can operate along with MaximDL for all the key imaging related functions, but for full observatory automation it needs some way of interfacing with external equipment for roof control and weather sensing. The recommended way to do this in CCDC is to use a Weather Sensor such as the Boltwood from Diffraction Limited ( and an ASCOM compliant dome controller such as LesveDome (

The Boltwood was out of the question as it was too expensive, and LesveDome wouldn’t give me the flexibility I would need for a roll of roof design with added functions, so I decided to use Homeseer.

Software Configuration: Homeseer

Homeseer is primarity for Home Automation purposes, but being an open-ended interface with a huge number of plugins, it works well in a variety of uses. In this case I use it along with a Velleman K8055 USB Digital I/O board ( to provide interfacing with the observatory.

This does a few critical functions in the workflow. While CCDC is running an imaging session, Homeseer will monitor both a rain sensor and an security beam. If it detects rain it will notify CCDC by editing a Boltwood formatted text file that CCDC is monitoring. CCDC will then stop the session, park the mount on it’s side and notify Homeseer that it can now close the roof. If it detects an intruder via a security beam it will also follow the same procedure.

As my Observatory is a small motorised roll off roof, the scope has to be parked on it’s side to allow the roof to close. This raises the obvious problem. To allow the roof to close with any degree of certainty, I really need to know the scope position. To combat this, I have fitted an IR Beam Kit to both the scope and the Observatory wall. This gives a contact closure to the input the K8055 when the scope is in its parked position. This allows Homeseer to then close the roof, if it doesn’t, it fires a alarm event.

There is also a sensor that monitors the position of the roof. If the roof doesn’t close after it should, this also fires an alarm event.

Homeseer can also make use of another Home Automation protocol called Xap. This allows networked communications between Homeseer and either other versions of Homeseer, or different compatible applications. In this case Homeseer on the Observatory PC can communicate with a version of Homeseer running on my Laptop in the house. This means that if any of the error functions happen during the session (Rain, Intruder, Roof Fault) an alarm can be ran on the laptop, waking me up. I see this as being really crucial to the whole process, as a number of things could go wrong.

If the session is complete without any errors, Homeseer can then switch of power to the mount and CCDs. As my CCD has a built in shutter, there is also the option here for allowing CCDC to automatically takes bias frames, dark frames, and flat frames by mounting a Electro-luminescent Panel on the Observatory wall. Given the scripting power within Homeseer, it’s even possible then to stack the images in Deep Sky Stacker

The one thing that bothered me about my permanent set-up is that because it was on the north side of my house I was losing everything below 40 degrees to the south. This meant I was missing a large amount of objects. The answer was to build something on the south side of the house. I don’t have much space so it would have to be as small as possible, the smaller the better. I also want something that can be at least semi-automated, so I can set it running and go to bed.

I cut down my pier to it’s smallest practical size, and by parking the scope on it’s side, I managed to get the space it needs to operate down to 5″ x 5″ square by 4″ high. The main problem in doing this is that I don’t have the space for a traditional roll off roof, but I need the same functionallity. I opted to go for a self-supporting roof design, something I hadn’t seen done before but it seemed like a good idea

I made some plans in 3D using Sketchup, then started building it using wood and cladding from B&Q. I used heavy duty rollers that are usually used for steel gates, they were quite expensive but do the job really well

The great thing about this design is that when its closed it only takes up a really small space, it’s so short that you can’t see it over my fence which is good for security. The hardest thing was making the legs stable as it rolled in and out. Because the whole thing is on a slope, the back leg is offset to clear my front step, and the roof is quite heavy, I had to reinforce the two legs by joining them together. I hadn’t imagined this would be an issue when I designed it

The next stage is to go for full software automation: from powering up the scope, doing an auto-alignment process, auto-focusing, auto-guiding, taking sets of exposures and then parking the scope on it’s side and closing the roof. It will also need to have built in security and some sort of weather sensing to protect against possible rain

Nano Observatory (roof on)

Nano Observatory (roof off)



Now that I have moved into a house with a garden, it was time to get myself a permanent set-up. I started using my kit from the back of the house, where I don’t have space for a full sized roll-off roof shed. I decided to modify a wheelie bin storage unit from B&Q. At least then I could leave everything outside, the mount would stay polar aligned and everything would be ready to go



The design has interested a few people and Bill Leslie from Forres has built his own version here: