The Unhealthy Noise....

Sorry for the lack of updates... the "unhealthy noise" reported by Jonathan has been the main focus of the observatory for the last few weeks...

Shortly after Jonathan's last post, I headed out there with him to listen to the noise (which I hadn't heard until then) and see if I could work anything out.  It didn't take too long before it became pretty obvious that something wasn't right.. .

We managed to narrow down the "noise" - which was really more like a "clunk" sound - to the declination axis - it only seemed to happen when the scope was slewing in dec. The clink noise didn't really seem to happen at the same point in the slew (which I theorised might indicate an issue with the ring gear), nor did it seem periodic enough to happen on every turn of the worm.

What was slightly worrying though is that the OTA was being visibly deflected every time there was a clunk, and you could certainly "feel" the roughness in the motion by putting a light touch on the OTA when it was moving.  As you might imagine, we didn't so this for long. My thoughts were firmly that this was a more "mechanical" issue, than any variation in the sound of the slewing caused by software/pointing model adjustments.

I really wanted to isolate there being a gear meshing issue, so I disengaged the dec worm assembly from the gear (the same process used when balancing the OTA) and commanded the mount to slew in declination again.  I recorded a video of this (not for the visuals so much as to record the sound of the issue to demonstrate to the AAS/Stardome team - so please forgive the iphone camera work!)

This was both a good and a bad result.... the good news being that it couldn't be any problem with the worm or gear as the noise still happened when these were disengaged - the bad news being that it was now even more obvious that we have a mechanical problem to solve.

The old "mechanics stethoscope' trick (using a screwdriver handle held to the ear and the "blade" touched to various parts of the mechanism) seemed to narrow down the source of the clunking sound to be the region of the dec drive assembly close to the pulley on the end of the worm shaft. This led to a working theory of a possible bearing issue.

We didn't proceed with any more troubleshooting that night but did seek help and advice from both Grant Christie and Marc Bos (Stardome Curator of Instruments), and also to assess what options we might start looking into to find parts and repair if it does end up being a failed bearing.

Another visit to Kumeu with Marc seemed to lead towards a general agreement that possibly a bearing may be the issue - but confirmation would require removal of the assembly for closer examination to narrow down exactly what the cause might be, also eliminating other possibilities such as the servo  motor.  Permission was to be sought from Stardome (who own the mount on loan to Kumeu) for this investigative work to proceed.

After agreement that the dec assembly really should be removed and examined with the aim of repair (either locally, or by sending back to the manufacturer) - Jonathan and I again headed up to Kumeu to remove the OTA from the mount again and remove the offending assembly as previously discussed with Marc.

The paramount worm shaft -there is a bearing at each end
at least one of which seems to be causing the problems

It didn't take too long to carefully remove the dec assembly from the mount. Once on the desk, loosening and removing the belts (well one of them - the other had already had to be removed with the motor to remove screws bolting the assembly to the mount) - it became obvious that whilst the motor and pulley shafts were both very smooth to move - the worm shaft was anything but...  The rotation of the worm shaft was very rough, even "binding" occasionally - certainly not the feel that precision rotating parts should have.


Next steps will be to have Marc, and our other engineering experts take a look and see if we think this is something that we can have repaired and set up locally, or whether we package it up and send it off to the US.

In other news - you might recall a few blog posts back that I had been having trouble getting a response from the dome rotation "azimuth" sensor.  Well, whilst the scope has been out of action, we took the opportunity to get electronics wizard, and dome rotation expert Tony Burns out to the site to have a look and give us his valued opinion.  We ended up removing the sensor and the board and he took it away to check.

Turns out there were a few problems with the system - so it was a good  thing we did.  Firstly there was actually a problem with the USB board itself - at some point in history a short circuit or similar had caused some damage on the board (lifting some of the copper "track").  Also - and probably unrelated - there was an issue with at least one of the input channels that needed investigation.

Tony replaced the control  ICs on the board and repaired the broken track - effectively returning the interface board to "good as new" condition.

The Azimuth sensor however was not quite a simple.  The postion of the sensor (as I've mentioned before) is very close to the dome slot in the open position.  This means that there's a very good chance that any moisture may find its way into that area (as the dome is opened, or closed, usually in that position).  Add to this Auckland's very humid climate - and there is a very good chance that corrosion might at some point give some trouble.

Looks like the connections to the photosensor and IR LEDs has corroded away - meaning there was no chance of this working!.  Tony has fabricated a new set of optoelectronics for the encoder wheel using much more robust connections that will hopefully last many years to come.

All that remains is for this all to be reinstalled and tested - and hopefully the "break" we've just been forced to have will give us the opportunity to do this over the next few weeks.

Phew... that was a big post - and that'll teach me for waiting so long between updates.  I will post more when we know what is happening with our dec assembly and hopefully when we will be back up and running....

Steve H

Automation Elation & Slewing Problem Again

I went out to Kumeu last night determined to figure out the T - Point automated calibration run, I was setting up when I was pleasantly surprised to see Allister had turned up to set up his giant binocular Dobsonian for some observational fun, there was a few clouds about but there were also plenty of big gaps so after getting the C14 focused, I started an automated calibration run, this time I just paused the run after each image was taken of the first six sample points that were all in the start region, I then right clicked the images and astrometrically solved them via the image link feature of the Sky X, this seemed to do the trick, as every sample point after these six would automatically solve and add the sample points, needless to say I was elated to have finally got the automated process working, I then proceeded to capture 61 sample points covering one half of the sky, Grant Christie had advised me to do half the sky at a time to avoid mirror flop issues, so once this was done I then proceeded to capture sample points on the eastern side of the sky, unfortunately the first row of points was below the level of the dome so I just skipped these points, I had captured about 17 points on the eastern side when clouds put a stop to any further efforts, so I took a break and went outside to hang out with Allister who by this point was packing up his equipment, by the time Allister left the sky had cleared up again so I continued the automated calibration run but after just a few sample points the slewing of the telescope started making the same unhealthy noise that I had reported after doing my manual pointing model, knowing that the solution to this problem was clearing the pointing model and being happy that I could now easily do a new automated pointing model, I deleted the pointing model that I had spent all night capturing, I then re-homed the telescope and turned off the Paramount thinking that this would solve the problem but when I restarted the mount and gave it a test slew it was still making the same unhealthy noise, so again I re-homed the telescope turned off the mount and this time turned the Sky X off, I then took a good half an hour break before restarting the Sky X and the Paramount but again the mount was making the same unhealthy slewing noise, I'm starting to think the problem may not have had anything to do with the pointing model after all because this was happening with no pointing model at all, not wanting to cause any damage to the Paramount I just re-homed the mount and turned it off, the next clear night I will head out and try the automated calibration run again but if the mount is still making the unhealthy slewing noise I will have to just wait until someone can come out and take a look at it as I have no idea why it would still be making this unhealthy noise when there is no pointing model, heading home I was elated that I had finally figured out how to do the T -Point automated calibration run but this feeling was tempered by the concern I have for the unhealthy slewing noise the mount was making, hopefully it's nothing too serious.

^ T - Point sample plot screen shot.

^ Allister's giant binocular Dobsonian ( built by David Moorhouse )

Posted by Jonathan Green

Automation Frustration

We had a clear night out at Kumeu last night, so I went out there with the plan of doing the automated T- point calibration, after taking a few images manually with the Sky X and astrometrically solving the images via the image link feature, I started an automated calibration run, I quickly became frustrated though as none of the automated images would solve, I found this to be inexplicable as I could pause the calibration run, right click on an image that was captured via the automated calibration run and open it up in the image link feature and successfully solve the same image that failed to solve during the calibration run, it wasn't like it was taking a long time to solve the image either, so after trying this more than a few times, I then decided I would just slew to each mapping point of the calibration run, take an image and manually solve each image, this would work up to at most 11 sample points but then after that I couldn't solve anymore images, it was like the image link would just start to fail to recognise the stars anymore even though to my eye the images looked the same from start to finish, I remember Grant Christie telling me that you could run into problems if the temperature compensating focuser (TCF) wasn't correctly tracking the temperature, I was controlling the TCF using Maxim DL as the Sky X still can't connect to the TCF since we had the power cut that I reported a few weeks back, so this might be the issue, although I'm really unsure about it, I did have the temperature tracking turned on in Maxim DL and I was watching the TCF and I could see it was making adjustments over the night so I really am quite perplexed, after attempting different exposure lengths and different binning modes without any luck, I really felt at a loss to why the images captured during the automated calibration run wouldn't astrometrically solve, maybe I need to adjust the detection threshold in the image link or something? I kept at it until 4 am in the morning becoming even more frustrated as the hours went by, clouds rolled in around 4 am putting an end to my efforts so I closed down the dome and called it a night heading home in frustration, I think I'm going to need some help with this as I re-read the T-Point manual today and could find nothing that would help me resolve the problem, hopefully once we get the TCF working with the Sky X again this will help although I'm still at a loss to explain why none of the automated calibration images would astrometrically solve.

Posted by Jonathan Green

Slewing Problem Resolved

I rang Grant Christie today to let him know about the unhealthy noise the mount has been making when slewing, he let me know that the problem was probably due to me having to home the scope during the T - Point calibration run, apparently doing so changes two of the parameters of the pointing model and can cause problems, so I went out to Kumeu tonight to clear the pointing model to see if this would fix the slewing problem and I'm very happy to report that the slewing of the mount is now back to sounding healthy again, I also finally figured out how to connect the CCD to the Sky X, the reason I was getting error messages was that you can't have the camera connected to Maxim DL and the Sky X at the same time, up until now the first thing I used to do when I arrived at the observatory was to connect the camera to Maxim DL so I can turn on the camera's coolers as it takes a while for the camera to cool down, I know now this was the reason why I was getting no response from the camera when attempting to connect to it through the Sky X, I'm feeling a little bit sheepish that it took me so long to figure this out but I guess it's all part of the learning process to have a bit of trial and error.

Unfortunately the sky was pretty cloudy out at Kumeu tonight, I did take a few images using the Sky X to control the CCD but I failed to astrometrically solve the images, this was probably due to the images being affected by clouds, hopefully the next time it's clear I will be able to do a new pointing model using an automated calibration run, I plan to take two to three hundred samples of the sky so we will end up with a very robust pointing model, I stayed out at Kumeu until 10:30 but the sky seemed to be getting worse and with the internet not working ( another problem that needs to be addressed soon ) I had no way to know if it was worth waiting around to see if it would clear up, so heading home a bit frustrated that I couldn't get more done tonight I had to console myself that at least when the next opportunity presents itself the telescope and mount are ready for action again.

Posted by Jonathan Green

New T - Point and Slewing Problem with the Paramount

On Friday night I spent the entire night working on making a new pointing model, I scrapped the last one after reading the T - Point user manual and finding out that it was not advisable to take samples down near the horizon due to atmospheric refraction that causes offsets between the actual position of an object and it's observed position, I started collecting samples around 9:30 and only stopped collecting samples early in the morning due to the mount making an inexplicably unhealthy noise when it was slewing, I couldn't see any obvious reason why the mount was suddenly making the noise ( sounds like a rattle with clicking or cracking type noises ) so I just put the mount back in the home position and called it a night, by the time this had happened I had collected 82 sample points, the new model showed that we are only off by 2 arc-minutes in azimuth while the altitude was still excellent and we don't need any more adjustments in that area.

After attending the Auckland Astronomical Society council meeting at Andrew's house on Saturday, I thought I'd head out to Kumeu and see if the Slewing was still making that same noise and to also test the new pointing model, The first ten or so slews sounded healthy but then again inexplicably the unhealthy noise started up again, so again I just put the mount back into the home position and called it a night, I've talked to Steve about the problem and he is at a loss to explain it as well so we might need Grant Christie or Tim Natusch to come out and take a look at it, hopefully it's nothing too serious, before this happened I was quite happy with the accuracy of the new pointing model, objects were always just a few arc-seconds from the center of the frame although I didn't get a chance to slew all over the sky so some areas may be worse than others, I still had the same problem with the mount in the area of sky to the South East that I had the first time I attempted a pointing model so I couldn't take samples around Centaurus, Crux or Carina, below is an example image of the Galaxy NGC 1532, the image was a 100 sec long exposure and I was quite happy with how close the galaxy was to the center of the frame, the image exhibits some pretty bad vignetting but I think this was due to the Moon being almost full and very bright.

Posted by Jonathan Green

T - Point

Last night the waxing gibbous Moon was up and more than a few clouds were about as well but seeing that there were plenty of large gaps in the clouds I thought it would be a perfect night to get the new pointing model done using T - Point, when I arrived at the observatory I noticed that the computer was off and the temperature compensating focuser was flashing, I'm guessing that there was a power cut during the storm we had earlier on in the week, I booted up the computer and connected to both the telescope and camera without issue but when I tried to connect to the focuser it would not connect and came back with an error message from both Maxim DL and the Sky X, after ringing steve for help I switched the focuser back to manual mode and manually set the focuser back to the last known good focus, this worked well enough for me to be able to do the pointing model, the next issue was aligning the finder scope again which must have been knocked fairly far out of alignment at some point, this took me a while to get right but once I was happy that the finder scope and the field of view of the CCD were aligned again I got started on the pointing model, at first I had wanted to do an automatic calibration run but unfortunately the Sky X would not connect to the camera and I got error messages, I also noticed that it will be quite hard to keep up with an automated calibration due to the dome being fairly slow, not wanting to waste a good opportunity to get a pointing model done, I set about doing a manual pointing model using stars that I knew and that I could confirm through the finder scope, by the time the clouds really rolled in at 2:30 am, I had captured 57 samples, one problem I noticed when I was capturing the samples was that I couldn't take any samples from around Carina and Crux any attempt to jog the telescope in this area of the sky caused the mount to become confused and try to point the telescope at the floor, each time this happened I had to re-home the telescope before it would accurately point and track again.

The new pointing model confirmed that the altitude of the mount is correct but that the azimuth still needed to be adjusted ( by about 3 arc minutes ), I then enabled the Pro Track so that the new pointing model can make minor adjustments to improve the sidereal tracking rate of the Paramount GT - 1100S, at this stage the clouds had become so thick that trying to test the new pointing model's accuracy was pointless, so at 3:30 am I closed up the dome and called it a night, I'm looking forward to the next clear night when we can test out the adjusted tracking rate and also the accuracy of the new pointing model.

Posted by Jonathan Green

Polar Alignment Refined & Observations of Arp 123 & Arp 279

Last week we had a one good clear night and few nights with sporadic clouds, Thuesday night was the one clear night and on that night I took images of two pairs of galaxies that are in the catalog of interacting galaxies of Dr. Halton C Arp, one pair was NGC 1253 & NGC 1253A, collectively known as ARP 279 they are located in the constellation Eridanus and are thought to be lie around 75 million light years away, the other pair is NGC 1888 & NGC 1889 which are collectively known as Arp 123, they can be found in the constellation Lepus, after doing some research I've found that NGC 1888 is thought to lie 33.5 megaparsecs away which is over 109 million light years! for now this is the most distant object I have ever imaged, I'm looking forward to breaking that record soon hopefully, the image below is of Arp 123, image data was captured by Jonathan Green and then processed by Amit Kamble.

On the Thursday I also collected more image data on Arp 279 but not long after the Moon set clouds started to roll in, there were still gaps in the clouds so I attempted to image a couple of comets, I imaged both 81P/Wild and 203P/Linear but finding them took some effort, I realised after finding the 1st comet that our pointing model was off by about 13 arc minutes, this was due to our refinement of the polar alignment on a previous night.

On the Friday night the sky looked nice and clear and the satellite image data also looked promising so I rang Steve to see if he was free to head out to Kumeu to work on the Polar alignment and to create a new pointing model with T-Point, we also had a couple of Auckland Astronomical Society members show up with their Dobsonian telescopes looking to take advantage of the clear skies, unfortunately the clear skies did not last long so the nights observing was a bit of a bust, fortunately we could still detect stars through the cloud so we ended up spending the night until 2 am in the morning refining the polar alignment again, now all we need is some clear skies to do a new pointing model.

Posted by Jonathan Green

Rain, rain and more rain....

After some wonderful days and nights between Christmas and New Year, (when, of course I was away from Auckland!), the rain has been falling pretty much non-stop - not to great for observing, but a perfect opportunity to get a few things done that have been on the list for a while. First order of business was getting the USB interface board for the dome rotation system. This has up to now been screwed to the wall inside the dome, but with the issue being unprotcted and that its location was right below the dome home position.  Whilst we've been working, a number of times we've opened the dome and has mosture drip perilously close to the board.  Installing the plastic enclosure took a little longer than expected, but the board should now be protected from and water or moisture from above and be a little more immune to spiders and dust. 

 With that complete, next step was to install the new red LED strip lighting around the dome.  The old dome lighting was too dim with the red bulbs in when working in the dome, and inconvenient to adjust brightness or turn on/off (as the switch is downstairs).  We switched out the bulbs for white bulbs making it easier when we are working (not imaging!) but also bought some red LED lighting with an RF remote to adjust brightness and/or turn on and off. The new lighting is much more controllable and convenient, and hopefully makes life much easier whilst imaging in the dome.  

The final job for the day was re-wiring the dew heater to the new power supply.  At the same time, i added in a PWM controller to control the temperature, ironically one actually designed to dim LED lighting strips.    A good productive day, even though the rain continued relentlessly. The only challenge at this point seems to be that Azimuth sensor for the dome rotation doesn't seem to be functional  (before and after the board was remounted, so not the loose wire I was hoping for) this could be a configuration issue rather than a hardware fault (the dome home sensor is working fine proving the interface inputs are working) but this will need more work

 

 

A Dark Night

Last night was cloudy so I took the opportunity to spend a night taking dark frames for image calibration, when I arrived at the observatory at 9 pm I noticed that the computer had crashed since the last time me and Steve were there, we had left the computer updating the Sky X with a star catalogue that Grant Christie had advised us to use, the computer fan seemed to be really labouring so I restarted the computer which seemed to resolve this problem. I then connected the camera and set the coolers on to reach -20 but after waiting until 10 pm the coolers still couldn't cool down the camera past -15, seeing as it was a pretty hot night I figured this might just be as cool as the camera would get, so I then set about taking the dark frames, I collected 5 hours worth of 600 second exposures which was about as exciting as that sounds, after that was finished I took 30 bias frames to go with my 30 Dark frames, it was well past 3 am by this point, so I decided to call it a night, I closed up the observatory but when I came to clamp the observatory dome I noticed the big clamp was stuck it couldn't be tightened or loosened so unable to secure that clamp I had to just leave it with only one clamp holding down the dome, hopefully the clamp just needs some CRC lubricant or something, In the meantime I will look to buy a replacement clamp as we can't really afford to have the dome unsecured. Posted By Jonathan Green

Galaxy Hunt

On the 7th of November I went out to Kumeu Observatory, I had loads of fun taking images of galaxies in Sculptor, Fornax and Aquarius, although unfortunately I couldn't get lucky enough with clear skies to grab enough data to make an RGB image of any one galaxy, by midnight the clouds really started to roll in, so while I waited around to see if it would clear I went outside and took a few nightscape images with my 6d, the clouds didn't end up clearing and by 1am I decided to call it a night.   Posted by: Jonathan Green

Impromptu Star Party!

After such a long period of there being few visitors up to the Kumeu site, a clear, dark night brought quite a few visitors out to the site tonight - with the end result being a great night of viewing both in the dome, and outside.

In the dome, Jonathan tested out auto-saved exposures through MaximDL, capturing a whole series of exposures and (and changing the filter wheel - though Jonathan did this manually on Friday) using an automated routine.

We found that the residual polar alignment error (we still were not guiding at this point) whilst causing no noticeable trailing on the 80s or so subframes, certainly did show up in a gradual shift over time between the images - this shows up in a slight framing mismatch on the stacked images that I deliberately haven't "fixed" in the image (blue and green band at the bottom)

Target for tonight was the Helix Nebula NGC7293, a planetery nebula located in Aquarius a little under 700ly away.

Outside the dome Alastair had his 16" Binoscope up and running - and the views through it were simply magnificent.  The detail and immersive nature of seeing familiar objects such as 74Tuc, Tarantuala and the Orion Nebula in "stereo" was really something else and was frankly mindblowing!

Another member had an 18" dob out hunting for faint galaxies and a pair of keen new members came along with their 8" Astronz dobsonian enjoying the dark skies and good company.

Go the All Blacks!

Even though an early morning had been had (watching the All Blacks win the 2015 Rugby World Cup at 5am), the night too was looking "all black" so it was time for Jonathan and I to head out to Kumeu and continue the process.

The original intention of the next visit was to get the dome rotation "slaved" to the mount using Pierre de Ponthiere's excellent "LesveDome" software.  We were upgrading from the old version and so using a trial license until we were up and running.  A little planning snafu though was that I found the trial had expired the day before - so that was not really an option for tonight.   (Many thanks to Pierre though for now extending our trial whilst we get up and running  - if you are interested in dome automation, do check out his website http://www.dppobservatory.net/)

Instead we decided to do a short pointing model run, which would help our ability to "goto" objects accurately, but would also give us a sense of our residual polar alignment error (after only drift testing to there being no noticeable drift after 5 minutes or so).

After opening up the some and starting everything up though it appeared we had a strange problem. Even though the last visit the scope had been "synced" and would "goto" reasonably accurately, and the scope had been correctly "homed", the home point appeared to be a long way off.  We re-synced it using Canopus and this proved to be a mistake - the next slew commend sent the scope into pointing in completely the wrong direction (and manual joysticking to the target resulted in TheSky telling us we were pointing below the horizon).

Once we synced again though on a star much further away from the SCP everything went back to normal.  We eventually traced the home problem to not having set up the "home position offset" - and once we'd done this the mount worked fine even after a power cycle.

We set about doing a short pointing model run, and opted to do this "manually" using the CCD and the jog controls to centre stars across the sky  in the centre of the CCD chip.  The results quickly converged reasonably well, with us ending up with the stars being much less than 1 arcminute off dead centre in every case.

The pointing model also gave us an indication of what error we might have in our polar alignment - and indicated about 4.7 arcminutes out in altitude, and 6.6 on azimuth.  In terms of the adjustment to correct this, this is just over a third of a turn of the azimuth adjustment screw West, and just under a third of a turn on the altitude screw high.

We pointed the scope at both the Tarantula Nebula (NGC2070) and the Sculptor Galaxy (NGC253) and ran a few images, partly to test the filter wheel and also to produce a pretty picture for the evening!  Here's the result - was only 10 frames (2 each RGB and 3 L) unguided getween 40 and 80 sec per exposure.  Stacked in DSS and processed in Photoshop.