Weather Station

From Covington Village Wiki
Jump to: navigation, search

The Covington Weather Station is a cheap and cheerful La Crosse WX2350 powered by a Linksys NSLU2 pico-computer.

ALSO: Check out the Garden Cam.

Service Log

19:44, 15 September 2016 (UTC)
Camera is back. I borked the server by deleting /var/lib - slip of the keyboard due to having a dog on my lap and someone coming to the door that I hadn't noticed. All back, finally.

19:15, 14 September 2016 (UTC)
20160908 14:45 22.7 22.7 22.7 61 14.8 10 23 228 54.4 1002.7 -100000 -100000.000 -1.0
Yes, don't believe the 54.4mm of rain in one 5 minute period. It was dry all day, but I'd unplugged the rain gauge to try to unstick it and this is the result.

16:18, 17 June 2016 (UTC)
Recent lightning has taken the Garden Webcam off-air. Hope to get it working again soon.

09:49, 1 May 2016 (UTC)
Well. No idea what took the system off-air. I guess it must have been a time/date glitch. All working again, now.

15:59, 28 February 2016 (UTC)
Camera was inadvertently unplugged on the 23rd and we only noticed just now! All working again, now. You haven't missed much. ;-)

22:22, 26 December 2015 (UTC)
Weather station decided to have a holiday over Christmas. I suspect the UPS is unwell, but it all seems to be running at the moment. Only the weather station does not come back on after a power fail. Taken UPS off-line as it's rather warm,. Another APC UPS that cooked its battery. I get this a lot. No warning, just everything gets hot and it stops working.

18:33, 28 November 2015 (UTC)
Playing with the old anemometer as the new one hasn't turned up yet. Ignore all the whacky wind speeds. The speed multiplier in the control software isn't playing ball.

19:14, 22 November 2015 (UTC)
Wind speeds still not right. Suspect the anemometer is giving up the ghost. New one on order.

13:36, 12 May 2015 (UTC)
Rebooted station as noted some very strange wind speeds for today. I suspect mechanical problems with the anemometer. Will keep an eye on it.

12:34, 25 February 2015 (UTC)
Noted station has not been logging since 06:25 on the 22nd February. The garden Cam is also looking a bit strange. It's come off it's mount! At least the weather station continued to log data, even if it wasn't uploading it. A reboot and it's re-populated the online data. Phew.

18:44, 2 January 2015 (UTC)
Repairs to the WS3200 Rain Gauge completed using a reed switch from Maplin:

Part    Description            Qty.
CL36P   MicroMin Reed Sw N/O      5

Yes, I got five. Making sure I have spares! They're slightly smaller than the original, but that doesn't matter too much - the magnet switches them perfectly reliably. The Rain Gauge was put back on the system at 17:50 UTC today. Ignore the big rain spike just before that - I was testing that the system was working properly.

16:36, 28 December 2014 (UTC)
Well, we have yet another broken reed switch in the rain gauge. The gauge is a tipping-bucket type, which has a small magnet on it's bucket. As the bucket tips, the magnet passes a reed switch and momentarily opens it. This tells the controller that a measured amount of rain has fallen. Unfortunately, the reed switch has stopped responding to the magnet. I've ordered some more reeds, but this being the holiday season, I don't expect to see them for a few days. Luckily, it's not raining!

09:00, 27 December 2014 (UTC)
Rain gauge seems to be either stuck (it's a tipping bucket type), or the controller has forgotten it's there. Loads of rain last night, but nil on the database. I have checked the connections and wobbled the tipping bucket to make sure it's moving freely. That's the "1mm" of rain for 8:30UT. Let's see if that fixes it.

14:09, 12 December 2014 (UTC)
OK, so the 66mm of rain at 3am didn't happen. What happened was a lot of wind and the mast offset the rain gauge is connected to had a damn good shake. I'm going to move the rain gauge to a location where it won't get shaken or have birds sit/poop on/in it.

14:09, 21 September 2014 (UTC)
Webcam (finally) replaced. There now follows a period of testing and alignment.

14:19, 2 August 2014 (UTC)
The big peak this morning was due to VIBRATION from moving the Met Mast and NOT rain. Everything AFTER that peak is RAIN. I was desperately trying to beat an incoming front and had to rush the re-rigging of the mast after fitting a UHF Land Mobile Repeater antenna on it. This sort-of proves that it's not a good idea to install a tipping bucket rain gauge where it can get shaken!

10:27, 20 July 2014 (UTC)
No rain this morning. That was a pair of woodpigeons fighting on my rain gauge!

18:20, 9 July 2014 (UTC)
Our ISP had a fibre outage between 00:30 and 18:30 today. Whilst the weather station was off-net during this period, it continued to function and now we're all back up again the machine has updated the web page so nothing was lost.

18:20, 24 June 2014 (UTC)
System stopped updating graphs at about 16:00, although the database kept on being updated. System rebooted at 19:15.

11:54, 13 June 2014 (UTC)
Extra sun shield fitted to temperature sensor. By fitted, I mean, a wooden fence post is now leaning aginst the wall and holding a plastic seed tray that is positioned to keep the morning sun from reaching the sensor. If it works, I'll make a nicer version.

10:42, 15 February 2014 (UTC)
Yesterday's storm twisted the anemometer around again, probably at 14:30utc. It has now been reoriented and will be properly tightened once its safe to go up to it.
The mounting bracket was twisted out of the wall, which is amazing, as it's a solid commercial K-T bracket with proper expanding bolt fixings.

13:39, 24 January 2014 (UTC)
Gauge replaced. Let's see how we go, shall we?

22:10, 22 January 2014 (UTC)
OK, SO THE RAIN GAUGE IS OVER-EGGING THE AMOUNT OF RAIN FALLING. It's broken, the sensor board is a mess. A new one is on order. Estimate end of January, 2014.

08:58, 19 January 2014 (UTC)
Weather webcam broke yesterday. I only have one spare, and it's got a slight focus issue on the left-hand side of the image, but it's that or nothing until I get a replacement in.

15:57, 4 January 2014 (UTC)
Having some fun with the internet connection today. Some twit is DDOSing the ISP's authentication servers, which means we're up and down like a yoyo. The weather station may seem to be off, but it's working fine and will, next time it connects, update everything.

11:22, 27 December 2013 (UTC)
Back where it should be...

10:16, 27 December 2013 (UTC)
Anemometer has been blown around the mast in the recent storms. It's now showing 90 degrees more than it should, i.e. South is now East, so a South-Westerly displays as a North Westerly. I'm adding a calibration setting to the system to get it about right until the storms abate enough for me to climb the mast and tie it back down.

00:31, 27 December 2013 (UTC)
We are considering a change of control software. Over the next few days there may be breaks in service as the configuration is tweaked and systems set up for the change. The change in control software should not change the storage of historical data as this will all be migrated across, nor should it result in much change of the display of data on the main weather station page.

17:11, 25 December 2013 (UTC)
Well. Don't know what happened there. The weather station simply stopped logging. Controller fine, displays all OK, but simply not logging any data. Ho hum. Rebooted. Couple of hours lost. Nothing serious as the weather has been fine and stable all day. Sunny and cold.

20:00 13 December 2013 (UTC)
The weather station has been isolated from the internet for a couple of days whilst maintenance took place on the internet connection and the internal network the station sites on. During this time it was still collecting data and it has now updated all the information on the web site so there should be no gaps in the data shown here.

14:58, 19 November 2013 (UTC)
Weather station is back. The whole thing has moved about 10m to the NE of its previous location. The anemometer is in a better position, the temperature gauge may not get so hot as it won't catch the sun on summer mornings and is now 2m off ground level instead of 5m, which should improve accuracy. I'm still looking for a permanent home for the rain gauge. It's currently at ground level, which makes it vulnerable to damage by insects (ants) and larger animals (people).

14:44, 17 November 2013 (UTC)
Weather station closed down for move to new location. Will be off-air for 24 hours.

17:49, 7 September 2013 (BST)
Weather station cabling is being re-routed to allow for a move to a new equipment room. This meant the station was off-air for most of the day.

23:13, 24 August 2013 (BST)
Something went wrong in the morning on the 19th August, whilst the station was unattended, and we didn't get to restart it until today.

19:12, 23 July 2013 (BST)
We had thunder last night. It took out the ELCB, which meant the WX station ran on UPS until the UPS ran out just after 2am. After that, we had no weather data until the system was recovered at about 8am.

11:43, 6 May 2013 (BST)
SLUG ran out of available memory. Not sure why this should be, but it seems to have filled it's logging space up with errors over the space of a few hours. No service between 2 May and 6 May.

11:00, 23 March 2013 (UTC)
Heavy snow. This usually falses the rain gauge. Yes, we had the equivalent of a good couple of inches of rain, but not quite as much as is recorded.

22:25, 5 January 2013 (UTC)
Merde. Aberration. Weather station stopped responding at 17:20 until 22:10 when it was rebooted.

23:38, 27 December 2012 (UTC)
FINALLY. It's working again!

16:45, 27 December 2012 (UTC)
Power now back on again, but weather station will take a while to come back to life. It's currently rebuilding its database. The SLUG takes a long time to rebuild it's database. Usually takes at least two hours to come back to life after an outage.

14:29, 27 December 2012 (UTC)'
Power Cut at about 13:00. Power has been off for just over an hour and the UPSs are now almost completely depleted. Weather station has just gone off.

16:39, 15 December 2012 (UTC)
At some time recently, a large bird landed on the Anemometer and broke it off it's mounting. It was re-attached at 16:00 utc today.

10:36, 9 December 2012 (UTC)
Replaced the weather webcam as almost half of the daylight captured images were dark. We need to get a reasonable reliability from this webcam as I want to get a year's worth of pictures to make a time-lapse movie.

21:47, 25 August 2012 (BST)
My. That Pi is toast. It has completely munged it's hard drive. I've reverted to the SLUG, but lost the database, which was not backing up properly. We still have the daily archive data, but are missing 18 July to 10 August in the database and archive data files from 10 August to 25 August, which includes some pretty interesting weather. Grrr.

20:53, 18 August 2012 (BST)
Weather station is off-air. We're working hard to get it going again. Hopefully, we'll be up by next weekend!

17:32, 5 August 2012 (BST)
10mm in 15 minutes. Yup. It even started coming through the roof into the house...

08:19, 4 August 2012 (BST)
AirPort Extreme doesn't like XBox360. Every time the XBox is brought too close to the AirPort the AirPort stops responding. Of course, the weather Station relies on that AirPort for external connectivity... :-(

08:06, 30 July 2012 (BST)
We had a power cut yesterday due to an overhead thunder storm. I took the opportunity to move the RP onto a UPS but it was a few hours before I could do that and we lost a bit of data before the machine was up and running again. The Garden Cam is also off-the-air due to the same storm. It was on a UPS, but has dropped off-net. I am unable to get to it until at least 6pm on 30July to find out what's happened.

13:09, 26 July 2012 (BST)
Someone unplugged the weather station! It was off-air between midnight and 12pm on the 26th July, 1012.

08:30, 21 July 2012 (BST)
Notice RP has not been updating the web site since 23:30 the night before. Investigations show that the Apple Airport it was plugged into had locked up. Once it had been rebooted the PI FTPd all the missing data up to the web site and everything was rosy.

23:37, 18 July 2012 (BST)
Hardware change. Linksys NSLU2 retired. Replaced with Raspberry Pi Model B. Database migrated. Archives re-created.

10:33, 30 June 2012 (BST)
Software updated to WView 5.20.2-1. Daemons restarted.

16:05, 28 June 2012 (BST)
The station managed to log 23.8mm of rain in 5 minutes due to the NSLU2 running out of disk space. Will try to remove this measurement, along with a couple of other aberrations, from the database when the hardware platform is replaced in July.:

20120628 16:00	30.0	30.0	33.0	60	21.5	17	23	279	0.0	994.0	-100000	-100000.000	-100000.0
20120628 16:05	29.9	29.9	32.6	60	21.2	17	14	264	23.8	994.0	-100000	-100000.000	-100000.0
20120628 16:10	29.7	29.7	32.1	59	20.8	22	21	268	0.0	994.1	-100000	-100000.000	-100000.0

10:35, 30 June 2012 (BST)
Noticed the graphics were not updating and had not done so since the NSLU2 ran out of disk space the day before. Before restarting, took the opportunity to update the wview packages.