The weather at VE3KSR

VE3KSR Weather Station News

May 2, 2000 Its unfortunate, but shortly after the station is to be featured on the CKCO-TV weather with Dave MacDonald, I'm going to remove the temperature/humidity sensor and send it back to the manufacturer for repair or recalibration - you may have noticed that the readings drop off too steeply below about 50%. The manufacturer tells me that I should be able to plug in a spare temperature-only sensor in the meantime, so we shouldn't lose any data that is correct now. However we all know about 'Famous last words.'

April 7, 2000 Well, I really screwed up this time... when I went to adjust the time for daylight savings, I did a seemlingly innocuous thing - I pushed the power adapter more firmly into the power bar. Unfortunately, this actually caused the main power to be lost - something I did not notice because the backup battery immediately took over. However, the backup battery died later that night but not before the station sent some corrupted data (like wind speeds >1600 Km/h and temperatures over 56C). While logged in remotely to prevent the bad data from being transmitted, some other innocuous-looking changes prevented the data flow from restarting after the power was restored. I finally fixed everything by April 15 and we're now back to a stable situation.

January 8, 2000 The last readily visible problem seems to have been eradicated... the barometer reading now changes smoothly. The value used to jump around within a 1kPa range, making it hard to see the correct value - this was most easily seen in the graph. The cause was apparantly more radio frequency interference caused by the system's processor being directly on top of the transceiver. When the processor was moved away from the radio (in a holder made of sheet metal, which further shields it) the values are now accurate, stable and change smoothly. I also reset the clock, which had drifted 10 minutes fast, and I hope the barometer fix will also cure the clock drift.

January 1, 2000 Had a program crash (no, not a Y2K bug - a New Year's bug) that took until 00:40 to figure out the work around. Lost the first 35 minutes of data. Sorry - won't happen again in 2001! Happy New Year!

December 15, 1999 Yesterday, shortly after 1:00 pm, the wind speed dropped to 0 (during what was a pretty blustery day) and did not start to register again until about 7:30 am this morning. There is a very simple explanation... the anemometer is not heated (the manufacturer advised against purchasing that option because it is virtually useless in our climate) and the plummetting reading coincided with the start of a blast of heavy, sticky snow. Basically, the rotor gummed up with slush that did not melt out until this morning (at which time the rain guage suddenly registered over .5cm because the meltwater from the slush accumulated in it started to drip through).