Dome rotation (1)

Having tackled the mechanisation of the two shutters I thought that although manual rotation of the dome was  simple,  fast  and easy it would be cool if I could motorize that as well. Certainly, were I to decide on a fly-by-wire remote control of the observatory in the future, hands-free operation of the dome window orientation would be necessary.

My first attempts involved rubber wheels attached to low voltage motors and attempts at rotation of the dome by friction from the wheels applied to the undersurface of the doughnut beam at the base of the dome. Sadly this proved unreliable and beyond the capabilities of the motors I had at hand. I must confess also that I had little faith in the arrangement from the start and although I have seen similar systems work on some sites I visited on the net, it struck me as a rather crude way of providing mechanical coupling,  liable to slippage and jams.  

A “rack-and-pinion” sort of system or one that provided drive via a chain and sprocket coupling would provide a more secure and controllable linkage between motor and dome base. The problem of course was that a chain like the one I had used on the shutter would not be easy to mount on its side and therefore some sort of belt had to be used that provided holes all along its length for the sprocket wheel to sink its teeth in and drag. 

Commercially these systems are available for purchase and for mounting on certain ready-made domes but they are quite expensive and I was not even sure I could get one to fit my precise needs. Looking for some ready made alternative punched steel flat bar I had no joy and so once more I had to make my own.

Like any DIY metal worker can tell you, a jig is a great time-saver!  So if you are planning to make equally-spaced holes, half-an-inch apart all the way around a hoop of  15x 4 mm thick flat bar over 7 meters in length  you had better take your jig-making seriously.

This is what I came up with…

HoledbeltJig2HoledbeltJig3

Although perhaps it may look all a little daunting at first (and the backdrop of jumbled workshop ware does not help!) it is actually quite a simple contraption. It consists of a channel 15 mm wide made by welding two short sections of angle iron with the gap in between that allows lengths of 15mm flat bar to be run through beneath the vertical drill. At the exit end of the channel a sprocket wheel (salvaged and modified from the bicycle gear set) was mounted on a lug welded to one of the angle bars and held there by means of a bolt and a lock-nut that act as a pivot allowing the sprocket wheel to turn freely. The hole for the bolt is made in the form of a slot to allow adjustment of the sprocket wheel enabling it to just about reach the bottom of the channel as it turns. This allows the sprocket to dig into the flat bar at each hole, thus checking that  they are properly spaced. 

To further ensure that the drill-bit is at the right place and held there while it turns, two adjacent spring-loaded bolts were installed between the sprocket wheel and the drill and spaced exactly 3 hole lengths away from each other and six hole lengths away from the drill and from the sprocket at its lowest.

The spring loaded studs were simply made from 6 mm round bar that was threaded 2/3rds of its length and fitted inside a “U” made from a piece of flat bar. Compression springs were inserted in the section of stud within the ”U” and their tension adjusted by a nut and washer threaded below them. Another nut was threaded at the top of the stud around the bit that extended from the top of the “U”. The picture above shows the arrangement. When both  bolts slammed down into the holes beneath it meant that the distance between the holes was perfect and the next hole can be drilled. Movement of the sprocket wheel with advancement of the holed belt underneath further confirmed proper hole alignment.

HoledbeltJig1

The arrangement was very sensitive to minor deviations in alignment and altogether this jig was to a large extent the main contributor to the project’s success. Some sections were deemed unsatisfactory and had to be discarded. A total of over 500 holes were drilled using this gadget and when one got the hang of it, it became a breeze! 

HoledbeltJig4

The belt was built up from several sections of punched flat bar screwed to the inside perimeter of the doughnut beam at the base of the dome, making sure the joints themselves were as low profile as possible and performed in the gaps between holes but keeping the spacing in sync. The holes themselves were slightly countersunk to allow easier release of the sprockets and prevent them ‘catching’ inside. Small 3 mm countersunk holes were further made at intervals of each six holes in the spaces between adjacent holes to allow 15 x 3mm screws to be inserted to fix the belt sections to the wood underneath.

20160219_131419

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Advertisements

Front hatch control

IMG-20171126-WA0009
Front hatch servo can be seen disconnected from the front hatch frame

Having tackled the problem of raising and lowering the main shutter, the lower hatch at the front of the dome was tackled using a 12v DC servo motor with a worm driven rod linked to the right side of the inner frame of the  lower shutter. When operated this would push open or close the lower hatch depending on the polarity applied to the motor. The motor was obtained from the orient via ebay and a small bracket cut out of aluminum sheet used to mount it to the inside of the main arched beam on the right close to the base. The best  mounting site for the motor to provide the desired range was found (as usual) by trial and error and a few tentative screw holes remain as evidence. 

The  opening of the front hatch was purposely kept  to about 10 degrees. This limits  viewing at very low angles but this was not considered necessary in view of the buildings nearby and to some extent shielded the OTA from street and house lighting. 

Hatch control (3) …schematic and build.

Control of the hatches needed to be  simple, battery operated and adaptable to any possible future upgrades and Ascom mediated synchronization. The schematic below seemed to offer this and required a minimum of circuitry to control both motors, namely the Main Hatch control motor and the Front Hatch servo that both were chosen to operate off 12 Volts.

RavenAndOwl_Hatch Control  

A lead acid gel accumulator of the type used in UPSes and burglar alarms is cheap and gives a sterling service for many years provided it is kept charged at an ideal dwell voltage. So it is very important to choose your charger well and make sure it hovers on or around 13.7 V under trickle conditions.

The workings of the schematic are simple but at the same time cunning :),  making use of diode switches to stop the travel of the hatch at both ends. Normally OFF  (NO) magnetic reed switches sited at the upper and lower extremities of the hatch window opening on the right are turned on at the arrival of the magnet that is bolted to the side of the hatch in front on the right close to the anchoring point of the chain.

When this happens, relays are actuated that open up a normally closed (NC) contact that was shorting out a diode rectifier wired to oppose the current flow to the motor. This causes the motor to stop.  A  red LED lights up to alert you that the fully open  position has been reached and a green LED when fully closed. These also indicate that the relay is still activated. The battery can handle several hours with the relays’ current flowing but the switches were purposely chosen with a centre-zero state to allow manual shut down of current to the relays once the open /closed state is reached thus conserving the juice.

I added  bipolar LEDs to indicate the direction of travel of the hatches. These light up red or green depending on the polarity applied to them. This is obviously an optional but I thought they looked “cool” .

SW3 allows operation of the hatch from an external supply and the wiring also permits the battery to be charged from a separate supply while the dome is in the parked position. This charging is best left on 24/7 when the dome is not in use.

The three double pole switches have to be capable of handling at least 10A. Most car type switches easily manage this and they sell cheaply (usually in quantities of 5 or ten!) on ebay or such.  A digital voltmeter was added at the front end to ensure proper monitoring of the all important charging state of the accumulator. 

The external ‘cigarette lighter’ output socket is wired in parallel with the external power socket. This allows  hatch control by an external 12v source in case of battery failure. I have a small 12V portable camping type supply that I keep for such eventualities and it also serves to start my car if that battery goes dead!  The 7amp 13.7v bench type power supply I use to run my AVX would also manage the hatch if the need arises. Conversely 12v can be tapped off from the external supply to feed the AVX via the extra socket.

2015-12-05 17.01.23
Wiring up the circuit for the hatch controller
2015-12-05 17.01.08
Front panel design stage

The panel was made from scrap aluminum sheet and screwed to two short lengths of timber and made to fit neatly between the two arched risers to the right of the front window. The small accumulator cell also fitted well in the space behind the panel. A cardboard template of the space at the top of the panel allowed me to make a roof for it out of plywood. This was just push fitted over the top to allow for easy access to the battery should the need arise. The charger was placed underneath at the spot where it is usually parked but its output cable gave it substantial leeway.

20160115_190659
Testing out the Hatch panel

 

Hatch control (2) – A successful chain-reaction.

Lifting and lowering the main hatch by chain drive meant that the chain itself would need to be firmly anchored to the hatch at some point.  Owing to the fact that it would need to be opened fully, the anchoring point had to be at the foremost edge of the hatch on the right which was the side chosen for the motor housing. The anchoring  bracket itself was designed to join up with the links of the chain by a rivet and to the front corner of the hatch frame on the right by two short stainless steel screws. 

The distribution and layout of  the drive chain and the array of associated roller bearings, take up spools and spring tensioners were almost entirely arrived at by that age-honoured method of trial-and-error. This naturally involved several modifications to the initial design and ultimately took the shape of the diagram below that shows the two sprocket wheels, a large take-up one at the top and a driven one at the bottom of the loop. The intermediate wheels are six home-made “chain guides” made from stainless steel washers sandwiched over a roller bearing mounted over 8mm roofing bolts driven into the arched beam at the spots chosen. This allowed the chain to track easily and diminished the risk of derailment had I opted for sprocket wheels instead. It also allowed the anchoring bracket to glide over them without jumping over cogs and wearing them down over time.

To reduce chain sag  with reversal of  direction and especially at the ends of its travel  the spring loaded guides seen in the diagram were added.  The double one in the middle helped to keep the chain aligned and the one below, just above the driven sprocket was essential to keep the chain well apposed to the drive, taking up slack and compensating for  the varying loads on the motor. 

Main_Hatch_Chain_drive diagram

The anchor point in the diagram shows its location with the hatch fully closed. 

This worked fine, lifting and lowering the hatch on reversal of polarity to the DC drive motor below BUT as previously mentioned the load on the motor varied with the position of the hatch and its direction of travel. Past the meridian in both directions, gravity would tend to speed up its descent and some form of braking was required.  This needed some thought. Indeed I could not find anywhere any mention of this problem and any useful reference, so I had to devise my own solution.  

Since a gradual brake was necessary on the descent in both directions past the meridian but was obviously not useful when the motor was lifting the load, some sort of  toggle system was required that automatically and continuously adjusted the load on the motor drive. 

… a sporting solution

I converted three of my home made “chain guides” to double ganged ones by adding another washer / roller-bearing gang on the inside of the ones carrying the chain, holding the two resulting channels tightly together by a locknut at the back. I added another of these guides to the back of the roofing bolt “axle” of the larger sprocket wheel at the top. I then drilled a hole in the horizontal support of the steel frame of the hatch on the right side about a third of the way from its lowermost edge and directly above and in line with the newly added roller guides. 

A length of non-stretch polyester marine-type cord was fixed to the hole in the hatch frame and guided over the bearings as shown in the speeded up animation below.

finalanim_improved

The free end of the cord was then attached to one of the springs “borrowed” from a  chest expander. This spring was channelled through a length of black garden hose pipe of the sort used for drip irrigation and the free end of the spring anchored to the rising right main arch beam  at the back via a hooked tension screw. The garden hose was kept neatly tucked away in the corner on the top edge of the arched beam.

The length of the cord was cut to size to allow the spring to be fully relaxed when the hatch is at the meridian. The tension screw  at the anchor point allowed fine adjustments to the spring tension to be made. When properly ”tuned” the spring would just peep over the top of the black tube at both ends of travel of the hatch. Silicone grease was liberally applied to all roller guide channels and silicone spray supplemented from time to time to reduce friction and wear on the chain.

The completed system worked well and the stored energy in the spring at the extreme ends of  travel of the hatch in fact served also as a boost to the motor when overcoming the inertia at the start of traction.

IPCAM - 2017-08-12 09.38.03
Completed system of  Main Hatch control drive.

 

20160115_190308
Detail showing drive and spring brake anchoring points and control unit.
20171207_123011
Detail showing “chest-expander” spring housing and anchor point.
IMG-20171126-WA0014
Low angle view of the main hatch chain drive showing motor location,
20171207_122844
Complete Main Hatch chain-drive system (low angle view).

Watch some of the stages in the making of this drive on youtube

 

 

 

 

 

 

 

 

 

 

Hatch control (1)- Alas the Windlass

As mentioned earlier lifting and lowering the main hatch manually proved difficult from the start because of the height of the roof and the poor mechanical advantage obtained when trying to drag the lid against gravity from its fully open position. Furthermore, controlling its free  descent to its closed position when past the meridian required a substantial braking force to counteract gravity once again.

My first attempt at mechanizing the main shutter employed a windlass and some heavy steel cable. In theory the cable would be attached to the shutter and conducted to a windup barrel in a continuous loop much like the old transistor radio sliding  frequency pointer used to have.  For this purpose I constructed a steel cylinder reel with two compartments and mounted horizontally on a frame inside roller bearings. . Turning this would wind the steel cable on one side of the reel while unwinding it from the other side thus enabling a controlled movement of the steel wire  anchoring point on the hatch. 

A temporary winding arm was added that would later be replaced by an electric motor .  Well so much for the theory…

In practice the device proved a frustratingly infuriating absolute waste of time!!!! Contrary to what you may think from reading these chapters I am not generally known for my patience and this infernal creation of mine nearly drove me to the brink! The steel wire showed no sign of cooperating and insisted on curling, knotting, twisting, crossing over itself, jumping lanes… you name it …it tried it! Another consideration was the hazard it would pose to the user (me!) should the steel cable or some anchoring point snap under stress!

Back to the drawing board, and this time I changed tack and decided I’d go for a chain-driven system that would allow forward and back movements controlled by motor driven sprocket wheels.  A visit to a friendly neighbourhood cycle shop rewarded me with two used-but-in-good-condition ½ inch 7-cog gear cassettes.

I dismantled one of the cassettes using WD40, heat and lots of  ‘axle grease’.  Naturally the individual cogs having been designed to fit over each other and around a common hub, needed to have their centres filled in before they could be used individually as cogwheels over a shaft. 

Having measured their internal diameter with a caliper I proceeded to cut circular discs  out of  a scrap piece of thick  steel sheeting to fit the inside of the cogs.  This was done using a metal jigsaw making sure the initial rough cut was somewhat larger  to allow for a more accurate sizing and shaping later.

An 8 mm bolt was driven through the centre of the discs and secured by means of a locknut and this was then mounted in  a slow speed vertical drill and  the final rounding off and size reduction performed with the drill turning and the edges rubbing against a hand held grinding wheel. (I must emphasise that this proced ure is fraught with danger and  it is very easy to injure oneself unless using full hand and eye protection as well as a suitable leather apron.  The spinning edges of the metal disc are as sharp as razor blades and fragments of grinder wheel can fly off and end up anywhere!! It provides a poor man’s alternative to a lathe but with proper care and precautions and frequent caliper checks of the diameter, one ends up with very acceptable steel discs complete with central bore. Indeed the discs were so well fitting that I had to actually hammer most of them in place! I repeated this for several of the cogs scavenged from the cassette and then proceeded to braze them in place using my trusty propane/oxygen torch and brass rods.

One of the smaller discs can be seen fitted inside one of the cogs of the set below.

2015-09-10 19.05.292015-09-10 19.04.49

This way, at budget price  I procured seven very usable cogwheels that provided me with an assortment of ratios to try out when it came to decide which ones to use to drive the canopy control.

I decided to buy the bicycle chain new and I resorted to a UK very comprehensive cycle store and got three full lengths of this sent 

http://www.chainreactioncycles.com/mt/en/kmc-z410-1-8-chain/rp-prod120670

I received these by post very rapidly and proceeded to test for proof of concept. I built a mount for a 12V DC low gear motor, one of several cheap chinese imports I bought on ebay and rigged this up. The other sprocket wheel in the picture was used as a free wheeling take up spool mounted on an 8mm shaft. A length of chain was strung up around the two gears and the system tested for traction.

Flying colours! 

2015-09-14 15.49.35

 

Camera(s) and a key.

Having looked around the forums and read what others had remarked about DSLRs suited for astrophotography I decided to settle for a used Canon 40D unmodded. It was in excellent shape and had been lovingly cared for by its previous owner. According to him it had travelled round the world and had served him well. I got it for a bargain price and with two lenses thrown in.

Of course I only wanted it for its body and having obtained the proper fittings and adaptors I had it rigged up to the Celestron in no time. For software I decided on that excellent package with the name of Backyard EOS and bought the full version. It was worth it because it made hooking up to the gear very simple and provided a lot of useful tools  for the Canon that it detected automatically.

All was fine and for a few months I was happy snapping at this galaxy and that until one fateful night disaster struck and the usb lead to the 40D got snagged against the AVX rear azimuth adjustment knob and to my horror pulled off a bite of the camera’s printed circuit board with it!!!  My Canon was destroyed. A new Canon 1200 was bought as a replacement but somehow I felt the 40D had given me superior pictures.

To enable long exposures an autoguider was found indispensable and again following the recommendations of others from the forums I decided on the Orion Starshoot.

It is good value for money and has so far given me a stirling service. The PHD2 software provides a very suitable and versatile interface for the autoguider.

20160115_190214
Canon 1200 and Celestron F6.3 reducer plus Orion Starshoot autoguider.

Another invaluable help in successful astrophotography is an f-reducer. This allows a wider field of vision and admits more light through the ODA providing faster resolution of faint and distant objects. I opted for Celestron’s own f6.3 reducer. It does introduce a bit of vignetting but I think its benefits outweigh this and anyway vignetting is fairly easily dealt with in post processing.

For planetary work I was using the Philips spc900nc webcam. I believe this model is no longer produced but its sensitivity to low light made it particularly suitable for planets and the moon. Video sequences are taken and then stacked using Registax 6 or some similar program.

At present I have replaced the Canon with a CCD monochrome cooled Atik 383L

 

I can’t say I am completely satisfied with it and I feel it does have its shortcomings. Many of these can be masked during post processing but even so I was expecting better results. I use an Atik EFW2 9-filter motorized wheel loaded with L,R,G,B,Ha, Hb,Oiii,Sii, filters. This is controlled using either the Atik provided Artemis software or via APT through the Ascom driver.

20170220_113902s
Atik 383L and EFW2 filter wheel attached

The added weight of the camera gear has significant effect on the AVX balance and for this reason an extra counterweight ring is added. This is probably overkill but with this in place balance is reestablished with the counterweights about half way along the shaft.

I have digressed a little and have jaunted through time to the present but I wanted to complete the chapter about the cameras so that in the next one I can return to the business of mechanizing the openings of the dome… a need that was felt very soon after first light!

But before I wind this up, as an anecdote, I would like to introduce you to the makings of my observatory door key. The lock I had chosen for the door was scrounged from some old door and lacked a key. Admittedly since the observatory was on my own roof it was highly unlikely to fall victim to  theft or sabotage but even so I thought it really ought to have a key… and so… I reverse engineered one taking measurements from the lock itself and  transferring these to the tip of another key body I had lying around. Having made a template I then forged a casting in aluminum..The propane forge itself was also home made and has served me well for several other projects.

 

 

Temperature and Humidity control

Now that first light had been and gone it was time to start looking at ways to make the observatory more functional so as to exploit its advantages to the full.

Three things demanded priority :  

  1.         Since astrophotography and Deep Sky Objects were my main targets, an f-reducer upgrade to the optical lineup and some sort of autoguiding was essential. A decent DSLR was necessary and it would have to be one with a proven track record in astrophotography and most of all affordable. Some optical filters e.g. a CLS for streetlight pollution and a Neutral density for lunar work might come useful. A low-light video camera for planetary and lunar work would also be desirable. Some complementing free or reasonably priced software would be required to allow control of the shooting from the PC corner.  
  2.         It became rapidly obvious that,  handle or no handle, for a 5ft 7in tall and mostly sedentary member of the species, dragging at the main shutter to open and close it was making unrealistic demands on my shoulder muscles and was possibly responsible for an extended course of physiotherapy.  The greatest difficulty was in pulling the shutter up from its fully open position and dragging it across the zenith and then slowing down its momentum as it fell to its final closed position. This was easy to do with the dome at ground level but lifting it 1.2m above ground it became a nightmare. The angle of attack was lost. A chair provided some help but at some risk. 
  3.          Thirdly and this in fact was probably the most urgent was provision of a constant and reliable control of the internal environment for periods when the observatory was to be on standby. This naturally would be a huge part of the time. Temperature and more especially Humidity regulation was always high on the agenda. 

The running expenses would prohibit installation of a full A/C so I had opted for the four extractors to provide a down draft and heat release from inside by convection.. and  the white paint on the outside to favor reflection. This is not entirely satisfactory of course and falls well short of good control but it does provide an acceptable range of temperatures the year round.

Humidity control was more crucial because in Malta this averages higher than optics feel happy with. The solution was a 230W small mobile dehumidifier. I routed its condensation water by means of a small plastic pipe that drained to the outside from beneath the back wall of the observatory and from there to the roof. This water would track down the roof drain and end up in my well to water my lemons… ahh the cycle of life ! 

20160115_190351
Dehumidifier and IR heater

One problem with dehumidifiers that rely on condensation is however that when the Temperature falls much below 15°C their efficiency falls too and at lower temperatures they stop working altogether.

To avoid this I added a home-made infrared heater. This was  based on three 50W IR ceramic bulbs that I mounted on a short section of rectangular aluminum.  This I fixed vertically above a round wooden base and added four small castors.

I surrounded this with a cylindrical tube of 13 mm square “chicken wire” that was anchored by small brackets to the bottom and capped on top with more chicken wire. This shape was adopted to limit its overall footprint and allow me to get close to the dehumidifier.

I position these close to the central pier when the observatory is “off duty”. When actually inside and observing there is little point in leaving these on with the hatch open. So I override the control and turn them off. The dehumidifier then retires under the desk and the heater in the wedge between the desk and the wall leaving the arena clear.

The design allowed me to keep the centre of gravity of the IR unit low and although tall it is very stable and will not topple over easily. It tends to skate instead!

20180202_124112.jpg

By using this approach and making my own heater I could experiment on how much heat was just sufficient and therefore optimise on my energy requirements. Infrared lamps are available of different wattages and one can mix and match.

In this way the total drain from the dehumidifier/heater was kept at around 400W and this only when they are actually ON because they are under control by relays operated from the sensors on the Control Panel described in a previous chapter.

Another useful fact is that the internal motorized webcam that I installed beside the Control centre can see the invisible infrared glow emanating from the bulbs and confirm that all three are OK.

Some IOT charts to indicate effect of environmental control

 

20180204_samp1_IOT
Recent chart showing triggering of Dehumidifier/Heater at  66%RH and approx 4% hysteresis
20170622_Temp1_Temp2_ comp_IOT
Comparative  display of diurnal fluctuations in temperature between the inside (Left) and the outside (Right) for one week in June 2017

20170805_20180329_dry_wet_compare_IOT

Comparison  between a hot dry Summer day and a recent humid one

All charts are produced via data uploaded from the observatory to the IOT via appropriate microprocessor controlled homemade sensor units and software that I will hopefully find time to describe at a later stage.