|Hall's Harbour Observatory, Hall's Harbour, NS|
Baie Ste. Marie Blog: 2015
|22-Nov-2017, 10:23 AST||22-Nov-2017, 14:23 UTC|
Baie Ste. Marie Observatory is now closed. My wife and I will be moving in a few weeks. It is time to pack up the equipment and prepare for moving the building.
Watch this space for news on the new site and for the grand re-opening, under a new name.
Shutter Controller Complete!
I am smiling!
A couple of days ago, I completed the assembly of my Shutter Control unit, and, yesterday, I completed the installation, including connecting up the limit microswitches. Today, after a minor adjustment, I tested it for several full open/close cycles. IT WORKS!!
This is the first step in replacing the old dome controller. In order to have a home-built, simple solution, it was critical to have a reliable shutter controller that did not require in-dome battery support or radio-frequency communications. This unit requires a simple two-wire interface. Give it positive voltage and the shutter opens; give it negative voltage and the shutter closes. The controller ensures the proper sequencing of the two doors without electronics, simply using the existing limit microswitches.
One disadvantage of this design is that the shutter can only be operated in the home position. I can live with that. Power is brought to the rotating part of the dome through a two-conductor commutator that was used by the old system for battery charging. Now it conducts positive or negative motor power. In the next phase of this project, I will be adding a third conductor so that the computer can sense when the shutter movement has completed.
I used surface-mount telephone jacks as junction boxes for the limit switches, just like the original controller did. However, it only used two wires per switch, and four-wire cables. I use all three connectors on each microswitch, and six-wire cables, so the junction boxes had to be replaced.
The controller box sits in the tray that formerly held the battery for the old controller. The mounting bracket for the old controller is now a cable guide, helping to ensure that all the cables remain out of the way of the home position sensor as the dome rotates.
The operating cycle is much faster with the new controller. The old controller had a five second delay between door movements. In addition, it slowed the upper door down to half-speed as it opened. The new controller runs both doors at full speed with no harmful consequences, and the lower door opens immediately as soon as the upper door hits its limit switch. More importantly, in the event of a rain emergency, closing is equally fast.
Another advantage is the ability to resume opening or closing from an intermediate position. The old controller got confused in such a situation, and I had to close the shutters manually. My new system will complete an open or close cycle from any intermediate position as soon as power is re-applied.
Last week, my dome controller finally croaked. I assume the problem is permanent, because several reboots, firmware re-flashes, and software updates would not fix it.
It started during one of my few winter photo sessions. Monitoring the session remotely from indoors, I noticed that the guiding suddenly went off-scale. When I looked at the guider image, the stars were drifting out of the field. A quick trip out to the dome revealed that the controller had turned off the power to the mount and camera. The guider was still running only because it draws power from the computer via the USB port.
On the next clear night, I attempted to complete the interrupted sequence, with identical results.
After upgrading the firmware and software, I could not communicate with the controller at all. Reverting back to the last "working" versions did not help. So, it is with relief that I declare the system dead.
I have been working, whenever I get the chance, on wiring up my home-built controller. My priority at the moment is to complete the shutter control unit, since operating the shutters manually is tricky. I think one more session with the soldering iron will allow me to finish it to the point of testing.
This week, I had one clear night. With the Moon almost full and a non-functioning dome, I decided to use the opportunity to do a drift alignment of the mount, using PHD2 software. I operated the dome manually for the procedure.
For the first time in over seven weeks, I was able to open up the dome and do some imaging. I was a bit rusty, but, it seems, some of the hardware was even rustier. My @#$% dome controller was really acting up.
With my camera cooled down to operating temperature, I opened the dome to cool it down. Some time later, the dome controller lost contact with the shutter controller and decided to close the shutter. In my effort to get it working again, I had to restart the controller's software, which meant losing power to the camera. By the time I had everything restarted, the camera had warmed up to ambient temperature. @#$%!!
I got the camera re-cooled without too much frost. But the controller agai closed the shutters. This time, not having time before darkness to re-cool the camera, I left everything running. I opened the shutters using the hand controller and then unplugged the motors to prevent them re-closing.
During the session, the dome controller would only listen to the DomeSync application for 10 seconds at a time. In order to keep the opening centered in front of the scope, I had to manually disconnect and reconnect the dome software every three minutes or so.
In a final indignity, the dome controller saw fit to turn off the power to the pier, including the camera and the mount. The guide camera kept running, since it is powered from the USB port. But, of course, without power to the mount, the guiding software lost its lock, and the guide star drifted out of the frame.
To make a long story short, I took fewer frames than I had intended, and I didn't get any blue frames at all. I had to synthesize blue by subtracting red and green from white. It's not as simple as it sounds, but it worked. The result is here.
We had a clear night Saturday night! Of course, there was a big Moon polluting the darkness, but I thought I could at least grab some Ha frames of something. Unfortunately, it was not to be.
I had already cleared what snow and ice I could off the dome. The upper shutter door opened slowly but almost all the way. However, it hung up on ice near the open position, and would not trip the limit switch. As a result, the motor pulled the actuating cable almost right out of its clamp. Fortunately, it stopped when I released the manual switch.
Without the "open" signal from the limit switch, the controller would not start a close sequence. This is a limitation of the controller, one that I have encountered before. It will not close from an "in transit" position.
For dealing with such contingencies, I have made up a pigtail cable with a connector that I have modified so that it can connect to the motor cables in either polarity. By disconnecting the motors from the controller and connecting them directly to the power supply, I was able to close the shutter again.
The excessive slack in the opening cable from the failed attempt to open the shutter caused the cable to jump off the pulley as the shutter closed. I tied a couple of knots in the cable to take up the slack so that the shutter would close.
So now I have some re-rigging to do. There is rain in the forecase later this week that will hopefully melt some of the ice.
The next time I see the sky, I won't recognize it. The last time I used the scope, six weeks ago, Orion was just rising at a decent hour. The next time I see it, it will likely be setting.
It is still winter here. *Sigh* The snow banks keep getting higher, and the sky is nowhere to be seen.
I had a cold-related casualty in the observatory last night.My observatory / weather computer shut down unexpectedly after 1:00 am. I had to bundle up in the -28 wind chill this morning to go out to investigate. The computer was powered off. When I tried to start it, it ran for about 45 seconds and powered off again.
I normally control this computer remotely, so there is no monitor attached to it. I had to go inside, fetch a monitor and connect it up. The result left me no wiser. It got as far as starting Windoze, and then shut down.
Finally, recalling a similar event some time ago, I unplugged some USB devices. Lo and behold, it started up just fine.
I think there are two culprits: a USB power supply in the computer that is not adequate to handle its ten USB ports, and an external hard disk (used for auto backups) that draws all its power from the USB. I suspect that the cold weather caused the disk drive to draw more than its normal amount of power, causing the USB voltage to drop, and the system to shut down.
I'll have to move the external disk, and a couple of other power-consuming devices to an external powered hub.
This weather gives a while new meaning to the term "drift alignment". The photo is from yesterday. This morning, the path is drifted in again.
Yes, another blizzard whine. After three blizzards in a week, I think I'm entitled so a nice glass of whine. We have more snow on the ground than at any time last winter, and last winter was apparently the worst in living memory in these parts.
The guy who plows our driveway came on Sunday to dig us out after the second blizzard. Unfortunately, he got stuck. The snow was more than his truck could handle, and he left with a huge mount of plowed snow in the middle of the driveway and the upper end not done at all. After some scrambling, we got an alternative plow operator with a bigger truck to finish the job.
He came again today after blizzard number three, along with a partner in a second truck. Between them, they got the job done. The snow is strange. Imagine trying to plow styrofoam: once you break it up, the chunks are lightweight, but before it is broken up, the entire mass is one solid object.
This winter got off to a more civilized start (January) than last winter (November). Still, it is making up for it with two blizzards this week, and another forecast for tomorrow. There is no point in digging out more than necessary today, since everything will blow in again tomorrow.
The observatory is holding up well under the circumstances. The dome does admit a bit of blowing snow, though less than last year thanks to my weatherstripping efforts. I have a tarp over the scope to protect it from what does come in. Unfortunately, that means I cannot move the mount, so testing the new beta release of the ASCOM mount driver software will have to wait.
I managed to get a few pictures of comet C/2014 Q2 (Lovejoy) in the last couple of weeks.
The ASCOM driver for the CEM60 still needs a bit of work. It is a new product, and they are still getting the software up to speed. However, iOptron is being very responsive, and are accepting user input about the driver issues.
Right now, I have a beta release (v3.21) of the ASCOM driver, and it has fixed the major SideOfPier issue that was interfering with dome synchronization and with automated meridian flips in Sqeuence Generator Pro. Another user, who has had clearer skies than me, reports that he was able to perform an automated meridian flip for the first time, last night.
Unfortunately, the beta driver has introduced another bug, so software cannot tell when the mount is slewing. With any luck, they will have that fixed in the next day or two.
Meanwhile, I have started building the new dome controller hardware. I have all the electrical parts I need, and I have completed the mechanical assembly of the controller box. Now I have to do the wiring.
I have had to decommission the old power controller, because I need some of its parts, so I no longer have automatic rain detection working on the dome. I will have to rely on accurate weather forecasts until I get the new unit completed.
In my last blog entry (Dec 28th), I mentioned having to control the dome manually. I have decided not to wait for the new mount driver that might fix the problem.
Instead, I thought of a way that I could bypass the problem by writing my own software to do the dome synchronization in a way that would be independent of vendor software errors. This means that it would be a useful product for other observatories experiencing similar compatibility issues, regardless of what hardware or software they have.
I now have the software (named, with astounding originality, "DomeSync") working. After testing it on simulated mounts and domes, I have tried several tests on the real hardware. To my pleasant surprise, it works extremely well, positioning the dome to within 2-degree accuracy, which is the limit of the dome-position-sensing hardware.
I am now polishing the software a bit, and doing some more testing. I want to ensure it will work in the southern hemisphere as well as the northern. And, of course, I want to ensure there are no hidden bugs lurking in it. Once I am satisfied that it is reliable, I will be publishing it.
Copyright © 2015 Kathleen Walker
Last modified: 27-Dec-2015