Our v2 allsky camera was an SBIG ST237A camera and was operational from 2006Dec-2018July. Ever since then we have been toying around with building another system out of existing spare parts and instead of a desktop, laptop or netbook data logging computer, a Raspberry Pi instead.
This pathfinder project was essentially an R&D project to get an operational camera system up and running, write it up and then see if others would like to go ahead and build some as well.
Luckily someone else thought of this first and posted the imeptus for us to follow at: https://www.instructables.com/id/Wireless-All-Sky-Camera/
This project uses a ZWO ASI120MC or equivalent camera, with included wide angle lens, housing and raspberry pi code.
The SCGO system has been mounted and operational since 2019 Dec 03 and work continues on commissioning it. This includes developing a dome heating system as it often condenses and or frosts up and accumulated snow does not normally melt off; programming to create archives, aiming of the housing to include more northern horizon for aurora capture; and more.
Total cost is expected to be under $300Can with most of that being the camera.
To date the system has been very robust with the raspberry pi having an uptime of 12 days since last restart.
This image shows sample site Nightingale, OSIRIS-REx’s primary sample collection site on asteroid Bennu. The image is overlaid with a graphic of the OSIRIS-REx spacecraft to illustrate the scale of the site. Credit: NASA/Goddard/University of Arizona.
Over the course of 91 days this year, we mapped this rock collection of rocks, for the end purpose of selecting a safe landing site for the spacecraft.
In the end, we also mapped (along with 14? others) the primary landing site “Nightingale”.
From https://www.psi.edu/news/bennusite
When NASA’s OSIRIS-REx mission spacecraft arrived at the asteroid Bennu, it discovered more rocks and boulders than envisioned. Mapping all these potential hazards was necessary to select a location to collect a sample of the surface for return to Earth. This effort was the work of multiple teams around the globe. One of those teams consisted of more than 3,500 citizen scientists who used CosmoQuest’s Bennu Mappers project to mark rocks, measure boulders, and map craters. Together, they made more than 14 million annotations of features on a global map of Bennu. CosmoQuest is a project that is based at the Planetary Science Institute in Tucson, Ariz. (CosmoQuest.org).
These volunteers had no way of knowing in advance if the sites they studied and mapped would be the one selected as the final sampling site for the OSIRIS-REx spacecraft. Folks who saw images with fewer rocks could hope “This is it!” But first, they had to first mark all the photos with hundreds of rocks and dozens of boulders so mission scientists could choose the images critical to site selection.
“It is amazing that more than 3,500 citizen scientists participated in CosmoQuest’s project to map Bennu and help mission scientists find the best place for OSIRIS-REx to collect a sample,” said Pamela L. Gay, Senior Scientist and Senior Education and Communication Specialist at PSI. “This kind of a volunteer effort makes it easier to find safe places to sample and scientifically interesting places to explore.”
Season 4 of The Expanse, a television series based on the works of James S. A. Corey (the pen name used by collaborators Daniel Abraham and Ty Franck) released all 10 episodes last week and we made the time to watch it over two evenings.
Although we still have discussions about binge-watching vs weekly released, it was still great to watch the 10 episodes without forgetting plotlines, characters and who was in danger with what when.
Highly recommended. If you have read the book first, it helps set the tv season and keep the characters familiar.
acurite weather station
It’s been over a week now since the new weather station (Acurite Pro model 06004) went operational in the back gardens. There have been signal dropouts and there has been cold temperatures (-20C), so we assumed it was the cold weather affecting the 4AA alkaline batteries powering the transmitter.
They were replaced 3 days ago with 4AA lithium batteries, rated for -40C.
And we still had signal dropouts!
So last night the receiver/display was moved from the front of the house to the rear of the house, closer to the transmitter, with fewer walls to penetrate, and with fewer sources of RFI between the transmitter and receiver.
The signal immediately jumped to 100% and has been solid for 12 hours now.
The issue is, we had 100% signal a *lot* when the station first went in.. and it got progressively worse and more intermittent over time.
Hopefully not a weather station failure in progress..
Only time will tell.
Updated 2019Dec13 – two days later the signal strength has remained strong and mostly consistent… just a few blips here and there of signal drop.
Pi Projects update
The weatherpi (3B model) data logger for the Acurite Pro Weather Station has been operational now for over 5 days straight, with the weather sensor out in the back vegetable garden. Everything is working fine and nicely correlates with the remote thermometer at the front of the house.
The oldest pi (model 1b) has failed at the simplest of tasks we wanted, running a 55″ display with web browser contents.. it just could not keep up. So it has been permanently retired.
The 2nd oldest (model 2b) is running the display and can even play music at the same time.
The next project in development is the wireless allsky camera, as seen here:
https://www.instructables.com/id/Wireless-All-Sky-Camera/
Installing raspbian “buster” with desktop went in fairly well on an 8GB microSD card, but because we would be storing imagery onboard for some unknown amount of time (we ran it for one night and found 2600+ images at 300K each totalling 770MB for the one night), we upgraded to a 16GB microSD card.
We hope to move images off the raspberry pi microSD card on a regular, automated process, but sometimes life happens and it does not work out as planned 🙂
This is a pathfinder project for the local astronomy club. Once operational, a detailed project plan will be written up with parts, supplies, sources, software and instructions… and then we will go from there.
The software install and initial configuration is complete.. now we turn our attention to the camera housing.
pi pi everywhere a pi
raspberry pi
We recently did an inventory of all of the Raspberry Pi computers we had kicking about:
1x pi1b 2014ish 1 core 700 MHz, 512MB, 2xUSB2, ethernet, hdmi
1x pi2b 2016ish 4 core 900 MHz, 1GB, 4xusb2, ethernet, hdmi
3x pi3b 2017ish 4 core 1200 MHz, 1GB, 4xusb2, ethernet, hdmi, wifi 802.11n
The oldest pi1b we had used as a command line interface datalogging from a davis vantage vue weatherstation, since 2014. Then the weather station died. I wanted to try and see if this would drive a 55″ display running in a lobby, putting out a dynamic website with the chromium web browser *and* using xrdp, use windows remote desktop protocol to command and control it remotely.
One step at a time. Downloaded and installed the latest Raspian OS, “Buster” and used “Balena Ethcher” to flash the biggest graphical version with software onto an 8GB microSD card with SD adapter. That took a couple of hours but did in fact worked.
The Pi came up, repartitioned itself, came up again into a nice graphical desktop. Opening up a terminal we did a “sudo apt-get update” then a “sudo apt-get upgrade” then a “sudo apt-get install xrdp”.
A little bit of research showed how to start up the built in chromium web browser automatically, and in kiosk mode:
sudo nano /etc/xdg/lxsession/LXDE-pi/autostart”
Add the following line:
/usr/bin/chromium –kiosk –disable-restore-session-state
also, start up the web browser, go to the website we wanted, and added that into the startup settings as the initial home page to open up.
Rebooted and it all worked!
Now mind you, displaying on a 1080p display was a little slow, but the functionality was there.
Attempting to remote desktop from a windows machine did also in fact work… but very slowly. On the order of a minute and the background wallpaper appeared. another minute and the web browser appeared.
It was unmanageable.
So the moral of this story … old tech is in fact a little too slow for some things. So we will shelve this model pi and try the next one up… the Pi2B with 4 CPU cores instead of 1 core, 900 MHz clock speed instead of 700MHz, and more RAM.
We should be able to remove the microSD card and transfer it across to the new Pi… will find out tomorrow!
Early Christmas
After many months of being without a weather station, we finally found one that should work for us.
This is an Acurite Pro (model number yet to be determined). It has arrived, been assembled, and is waiting to go outdoors and then to be connected to a data logging computer. We will try and use the Raspberry Pi 2 Model B that we used in the past with Wee-wx software and see what happens.
It will be mounted out back in the vegetable beds, about 2m off the ground. We are more concerned with ground temperatures where we work and play, and the veg grow, then we are about “more accurate” wind speed readings by being 10m in the air. In addition, this station does not log wind gusts, only regular wind speed.
Garlic to bed
The 847 garlic cloves of 99 varieties in 142 rows are nicely to bed and sleeping under a coat of straw and snow.
Hopefully all will survive and grow to be Big Big bulbs next year!
They had a *lot* of composted manure this time around to help out and we will be adding in irrigation in the spring.
Transit of Mercury
You may have heard of the Big astronomy event… the Transit of Mercury across the Sun.
Well… this is the view we had from our location….
No Sun. No Mercury. Just cloud.
Very disappointing!
The next event will be 2032 Nov 13… but not visible from North America.
After that 2039 Nov 7… 20 years down the road!
The latest linux fedora distribution came out a few days ago.. and for once I actually missed the event.
So starting yesterday, I started the twice a year process of upgrading all of the linux fedora servers. It is getting better and better!
The first one went without error and completed within 15 minutes. The next two also went without issue and completed within 30 minutes.
The base process remains the same:
sudo dnf upgrade –refresh
sudo dnf install dnf-plugin-system-upgrade
sudo dnf system-upgrade download –releasever=31
sudo dnf system-upgrade reboot
sudo dnf system-upgrade clean
Since yesterday, 5 servers and 1 workstation have been upgraded and so far only one had boot grub issues. Those issues have been fixed.
Garlic Prep
Two days after work or prepping for garlic planting. about 3 1/2 hours in total to crack open about 150 bulbs to get 847 cloves for planting in 142 rows of six each.
Working inside to do this part really helps in the field when planting. IT used to take us two full days in the garden, cracking and planting. Now we hope to be done outside in only a few hours.
This year we have at least 99 types, some of which are named the same but come from different sources.
acropolis greek
aglio Rosso Creole (2019)
armenian (2015)
artichoke formidable
barcelona red sp
bubbas chesnock
carpathian
chesnock red
chilean silver
chloe
crème de la rasa (2018)
czech broadleaf
dujanski
f3V
Fishlake F1
Fishlake F21
Fishlake F23
Fishlake F3
Fishlake F30
Fishlake F40
Fishlake F7
french
genki
georgia crystal
georgia fire
georgian (2018)
german red
german white
hungarian(2015)
inchelium
italian lorenz
italian purple (2019)
italian sicilian (2018)
japanese
japanese sikura
khabar
koji (2019) Perth
korean purple
kostyns red russian
kostyns red russian (2019)
lacota (2019)
lucian sicilian
magnificent
marino (2016)
mennonite
metechi
mexican Red Silver
music
music (2019)
musical/music
newfoundland (2019)
northern quebec
northern quebec (2019)
northern siberian
northumberland (2018)
omas
persian star (2015)
persian star (2018)
phillipino
polish jenn
port au george (2018)
purple glazer
purple stripe sicilian
railway Creek (2019)
red rezan
roja (2019)
rose de lautrec
rose de lautrec (2019)
rosewood (2015)
russian red
russian red (2019)
saltspring Select(2016)
seversky palisek
shouldice
siberian
sicilian gold (2019)
sicilian gold (2016)
silverskin40
slovak
solent wight
spanish antihelion (2015)
spanish roja
spicy korean
susan Delafield/Music (2016)
sweet candy (2015)
tibetan
transylvanian
ttv5
us polish
us republican
us romanian
vampire (2018)
wettergren (2016)
wettergren (2019)
wild garlic
wyld garlic
yugo (2018)
yugo red (2019)
zemo
True Field of View
True Field of View
In the past, when imaging Jupiter. I did a full field image run, put the image up on the screen, found out the apparent diameter of Jupiter and physically measured the screen and Jupiter to determine TFOV.
This past weekend, Kim & I tried it out on a star. Using firecapture and the reticle turned on, we centered the star, stopped the tracking and timed how log it took to hit an edge. Luckily for us, it went on a diagonal and exited almost exactly at the corner in 12.69 seconds.
* drift test for FOV wth ASI290MC camera, no reducer, no barlow, on the Torus telescope D=400mm F10, FL=4000mm
12.69 seconds from centre to top right corner exit.
camera pixel resolution=1936×1096.
The centre reticle location is 1/2 of that, so 548×978
Using right angle geometry, the diagonal is 1121 pixels, full diagonal would be twice that or 2242 pixels
calculate drift 1/2 FOV is 12.69 seconds
full FOV is 25.38 seconds
360 degrees / 24 hours = 15 degrees/hour
15 degrees/hr / 60 minutes/hr = 0.25 degrees/minute (15 arcmin/minute)
15 arcmin/minute / 60 seconds/minute = 0.25 minutes/second (15 arcseconds/second)
So just count how long it takes in seconds for a star to move across an eyepiece’s FOV and multiply it by 15 to get the FOV in arcseconds.
TFOV(diagonal)=15*Time
TFOC(diagonal)=15*25.38=381 arcseconds = 6 minutes 20 seconds diagonal
ZWOASIM290MC FOV 1936×1096 pixels gives 329 x 186 arc seconds rectangle
Turns out this method is WRONG! (*Unless* you are using a star with a declination of 0)
Use this formula instead:
The True Field Of View (TFOV) is then:
TFOV = 15.04*T*Cos(delta)
where “delta” is the star’s declination, “Cos” is the Cosine function, and “T” is the measured drift time interval. If the time is measured in minutes, the true field will be in minutes of arc, and if the time is in seconds, the true field will be in seconds of arc. For example, if a star has a declination of 25.5 degrees (ie: 25 degrees 30 minutes), and a measured drift time of 2.75 minutes (ie: 2 minutes 45 seconds), the true field of view is then 37.3 arc minutes in diameter. For stars within 3 degrees of the celestial equator, the Cosine function can be approximated to 1, and the formula becomes:
We used a star near saturn (mag 0.52 38 arcseconds with rings)
HIP 93423 mag 6.2 dec -22deg 39min = 22.65 degrees
cosine 22.65 degrees is 0.923
TFOV=15.04*25.38seconds* 0.923= 352 arcseconds (less than original calculation of 381)
so ZWOASI290MC TFOV= diagonal 352 arcseconds, rectangle 303×172 arcseconds
Then someone mentioned you can do the same with just the focal length and the camera sensor chip size, which led to this
formula:
I found: sensor size (mm) * 135.3/focal length (inches) = arcminutes
mixing units seems to be just wrong, but perhaps the fudge factor of 135.3 accounts for that?
the camera sensor size is 5.6×3.2mm
focal length of 4000mm=157 inches
5.6*135.3/157= 4.8 arcminutes
3.2*135.3/157=2.8 arcminutes
drift solution:
ZWOASI290MC TFOV= diagonal 352 arcseconds, rectangle 303×172 arcseconds = 5×3 arcminutes
So the two answers are reasonably the same, within error limits.