Jump to content





  • Posts

  • Joined

  • Last visited

Everything posted by Mognet

  1. I'm curious as to what you will make next. Your build threads are fascinating
  2. That's better And the finished rocket looks fantastic. Well done to all involved
  3. I got the alert on Friday night too. First one I've seen in the four years I've had the app. Went outside to check and it was completely clouded out. As is typical the sky started to clear again once the alert level was back down to yellow
  4. Seems to be an issue with the link. I get the same message too
  5. The Register interviews Brent Waller, one of the designers. Apparently it's 99.8% accurate on relative timing https://www.theregister.com/2021/09/15/lego_solar_system/
  6. Looks good. Glad you're getting on with it and you got your MQTT input sorted I know what you mean about the clear skies and full moon. Went out to check the skies last night and Jupiter was sparking, but the full moon washed everything else out. Tonight had been promising until about ten minutes ago when clouds rolled in again I didn't bother updating the about page with the last change as v1.2.5 only affected the install scripts. They'll be realigned with the next change
  7. Does it work if you set it to manual exposure? My D3100 and D3400 refuse to take photos when attached to the scope if I haven't set that
  8. Last night was the first decent clear night here in ages so I took some time to experiment with the wedge. And it does work. Took some thinking about and a little swearing The SynScan alt-az software isn't designed to be used like this so initial attempts at using star alignment had to be ditched. Either it wouldn't align, or it would give alignment stars I couldn't see, or it would cause trailing. It seems the way to do it is to set the location to 90 degrees north (or south if you're in that hemisphere) and then skip the star alignment. Target alignment can then be done by guesswork and trial photos. The camera stayed pretty much with the chosen star field for the twelve minute period I was shooting in. These are 200x200 pixel crops from the top left corner of the first and last images. For a first test and for my purposes it's acceptable This is the result from last night once I'd figured most things out. It's shot on a Nikon D3100 with a Nikkor 35mm 1.8G lens, 22 subs of 30 seconds each, F4, ISO1600, no darks or flats, processed roughly in Sequator and GIMP I would have taken more subs, but the camera battery ran out of power after those 22. Which led me on to a problem I was expecting. My print sizing isn't quite correct so the holes in the wedge don't engage with the teeth in the mount and it seems that friction alone cannot be relied on to keep the wedge in place so when I went to change the battery the whole thing rotated. I definitely need to modify that part of the design, and I think I'll just make up an adaptor for my existing one as that piece was a ten hour print
  9. Thanks. My designs vary between crazy and good, and I think this one could be one of the good ones I've not seen one of those before, but it is quite similar I noticed that when servicing the mount a few years ago. I was never quite sure I put the right tension on the azimuth nut when I put it back together. It's possible that I may not have set it tight enough
  10. As I'm being a bit tightfisted economical with my money at the moment, I thought I'd experiment with adapting my old style SkyWatcher SynScan alt-az mount by designing and printing a wedge to enable me to experiment a bit more with astrophotography. I'm not expecting it to be hugely accurate, just vaguely. But if I can get it to follow the same patch of sky to get 30 second exposures with my longer camera lens and without trails then I'll be happy. I seriously doubt it's strong enough to take the weight of a scope though. I haven't tried it out yet as I'm waiting on a clear sky, but I have mounted a camera and rotated it using the controls. There is some slippage in the mount in a couple of places which isn't unexpected as the balance is off, and there is some flex in it as it's only printed in PLA with 30% cubic infill I'm not publishing the plans yet as it's still a prototype. I know already there are some small changes that need to be made. The teeth on the original mount don't mesh with the holes in the print so might cause further slippage, and I need to tidy up a few other dimensions Hoping that this is actually a practical idea instead of one of my usual bonkers ones!
  11. I've found a list of four clubs in Suffolk http://www.astronomyclubs.co.uk/Clubs/Default.aspx?CountyId=67 Orwell Astronomical Society will be the closest to Hadleigh
  12. That's brilliant. The space clearance/empty directory problem should be fixed in this version along with a little problem in creating new periods that one of mine was having issues with. It looked like it was just a timing thing between the capture and new movie scripts. Space clearance now happens three times a day; twice for the new periods, and a third at noon when the timings are regenerated. And that third one also reports to update.log in ~/Mognet-All-Sky-Camera/back-end/ This latest version should be stable. I've had it running for about three weeks and haven't managed to break it yet I'll update I've updated the instructions so that removal of the downloaded files and installer directory are included, and added version messages to the autodeploy and update scripts too along with a version history in the README.md too And thanks. Job hunting is tedious. Twenty years experience in IT; 3 as a dev and 17 as an automated software tester, and I've only managed to get one interview out of 80+ applications so far.
  13. The error trapping on the main page isn't great, so the message of "couldn't open info file" means it's looking in a history subdirectory and not finding that file. I wouldn't be surprised if the directory is empty too, but the directory itself hasn't been removed. It was a problem I was seeing on one of mine before some of the recent fixes (v1.2.2 I think fixed it). The directory in question looks like it's /var/www/html/history/1623186661-night/ There might be others too, and they will probably be the oldest ones The find command is new to me. I'll have a play with that tomorrow in between job hunting and learning Python Just checked the update script on GItHub (https://github.com/MarkGrimwood/Mognet-All-Sky-Camera-install/blob/main/update) and it should copy the entire contents of the front-end directory to /var/www/html/ while preserving the history. (Just seen your update) It should handle Y or y too Not sure about the "No crontab for root" message. I think I've seen it on some of the clean installs I realise that 'works for me' doesn't mean it works for everyone else. Might need to see what messages are actually coming out of the update script. Might also be a good idea to add version number reporting to the update and autodeploy scripts too in the next update
  14. The Datyson image looks good There does seem to be a glow with the Pi cameras. This is from a V2.1 NoIR showing the same glow across the top. It's the second one I've had as the first one, while fine during the day, had an horrendous glow across the image at night that made it unusable
  15. Not quite a rocket launch, but the Virgin Galactic takeoff is currently scheduled for 1530 BST https://www.virgingalactic.com/
  16. Hi Jason. The latest version is v1.2.4, and that should be visible on the about page. There was a fix a little while ago that would stop an occaisional problem with space clearance. Looks like you might have run into that if your card ran out of space
  17. Hi Jason, Yes there is. There's a bit about updating here https://github.com/MarkGrimwood/Mognet-All-Sky-Camera-install/wiki#updating First bit is just the plain update, second bit is just to update GPS coordinates And by coincidence, I made some bug fixes today too. Just minor stuff this time
  18. With some help from Bob (jmrp) on GitHub who discovered the right settings, the code has been changed so that it can capture images every minute at night too. Only problem is that not all cameras can do it. It works with my Pi V2.1 NoIR camera modules and should do with the Pi HQ camera too, but it doesn't and does work with the WaveShare camera module. The solution lay in using --exposure off instead of --exposure night in the nighttime capture. There is a note in the raspistill documentation to say that whether it works or not depends on camera tuning
  19. A few little defects fixed and tested, and I've added a version history too. I've called it version 1.2 as I hadn't really recorded the updates before Fix history creation where an unused period containing only the initial files was being archived in the history section Fix minor defect in autodeploy and update that would display an error if enter was pressed with a blank entry instead of a response Change autodeploy to only copy the necessary files Change autodeploy so that the gps file is created in the execution directory Change update to copy LICENSE and README.md too Add a link back to the GitHub page and a this version history Tidy up on sun times page And I'll put Hardware Watchdog on all mine soon as a router reboot ten minutes ago has just caused one of them to lose connection
  20. Focus looks pretty good there. I know with mine I ended up settling for a close enough focus set in daytime The Pi Zero in my conservatory had a CPU temp of 25 at the time your image was taken, and the spare room one recorded 29 degrees. Looks like the Zeroes run cooler than the other models, so that might account for the change in condensation
  21. Thanks. It's looking good there, and it makes me happy to see someone getting good use out of it I did mean to link back to the GitHub page from the About page, so I'll do that when I do that little fix tomorrow If you want to capture MQTT data to add onto the image then look in capture.sh and newmovie.sh in back-end/pics/ as that's where the images are captured
  22. That's a known issue and does get mentioned in the wiki section. The cause comes from creating both day and night directories at the same time. When the changeover code in newmovie.sh is run it just assumes there was something in the directory to archive rather than checking. Along with checking the directory existance I should also check the count of files, and then only create the archive if there are more than four (the newly created directory contains image, thumbnail, mpeg and info files). I'll fix and test that tomorrow I've had similar happen with a router update and I've had to power cycle the Pi to reconnect, but I've also seen it reconnect fine too. Something to look into so I can add fix notes to the wiki. One of mine I set to a fixed IP address on the router, and I think I set the other to a fixed IP address on the Pi. The former is probably the better solution I'll have a look at that hardware watchdog. It could be interesting. I was planning to keep mine within easy reach but I know other people will have different plans
  23. That's good news. Odd about the connection problem, but these things happen. Let me know if you find anything wrong other than the little bits I've documented already All we need to do now is fix the weather!
  24. I suspect it's March winds and April showers making up for being late this year
  25. Quite possibly. Tonight it's 22mph, tomorrow gusting 50 according to the BBC forecast
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.