High Apparent Temperature glitch on restart

[gerardwood]

There appears to be a slight problem when restarting Cumulus - the High Apparent Temperature for the day can get incorrectly set in the log/ini files.

On restart, the monthly log has a line added which has an incorrect high value, which can then get transferred to today.ini if higher than the 'correct' value. There doesn't seem to be a way of correcting this, as stopping Cumulus, manually editing the log/ini files and restarting will then generate the same error.

I'm using a LaCrosse WS-2305 station, build 1026, and the monthly log is updated every minute.

Thanks,
Gerard.

[steve]

I suspect the problem is that the order of processing the readings is different for the WS2300, and it is trying to calculate the app temp when it hasn't yet read all of the required 'components'. Does this happen when you start up when there is logger data to be read, or when there is no logger data to be read (i.e. a quick restart), or both?

[gerardwood]

Hi Steve,

The problem happens in both cases, i.e. quick restart, and 'delayed' restart.

What is slightly interesting is that the monthly log file will have a duplicate entry for the minute at which Cumulus was stopped - presumably one for the time at which it was stopped, and a new one created on restart. The time taken between stopping and starting makes no difference, there's always two entries. It is the second entry that has the erroneous data. I've just done another stop-start, correcting today.ini in between. On restart, here's what happens to Jan12log.txt (last few lines):

05/01/12,13:27,10.0,71,5.0,14.0,25.9,326,0.00,0.05,1010.7,14.58,18.1,59,9.6,10.0,10.0,0.0,0,0.000,0.000,4.5,157,0.0,315
05/01/12,13:28,9.9,71,4.9,13.5,25.9,319,0.00,0.05,1010.9,14.58,18.1,59,11.2,7.1,9.9,0.0,0,0.000,0.000,4.6,156,0.0,292
05/01/12,13:28,9.9,71,4.9,12.0,17.4,320,0.00,0.05,1010.9,14.58,18.1,59,11.2,7.3,9.9,0.0,0,0.000,0.000,8.8,0,0.0,292
05/01/12,13:29,9.9,71,4.9,14.2,25.3,331,0.00,0.05,1010.9,14.58,18.1,59,25.3,7.0,9.9,0.0,0,0.000,0.000,5.0,0,0.0,360
05/01/12,13:30,9.9,71,4.9,16.6,25.3,334,0.00,0.05,1010.9,14.58,18.1,59,19.5,6.7,9.9,0.0,0,0.000,0.000,3.7,155,0.0,337
05/01/12,13:31,9.9,71,4.9,14.8,25.3,321,0.00,0.05,1011.0,14.58,18.1,59,6.9,6.9,9.9,0.0,0,0.000,0.000,4.0,154,0.0,315
05/01/12,13:32,10.0,71,5.0,14.1,25.3,319,0.00,0.05,1011.1,14.58,18.0,59,11.6,10.0,10.0,0.0,0,0.000,0.000,4.4,153,0.0,315
05/01/12,13:33,10.0,71,5.0,13.8,25.3,316,0.00,0.05,1011.0,14.58,18.1,59,10.7,10.0,10.0,0.0,0,0.000,0.000,4.6,152,0.0,292

I stopped at 13:28, restarted at 13:29, and have included a couple more minutes - the apparent temperature does seem to have a bit of a wander before sorting itself out...

Thanks,
Gerard.

[steve]

I'll see if I can change the order to match that of other stations.

App temp uses temp, humidity, and wind speed. If you have Cumulus calculating an average wind speed, the wind speed, and hence the app temp, will tend to wander a little after a restart due the data available for the calculation (even allowing for this bug).

[steve]

Could you try this updated cumulus.exe, please? http://nybbles.co.uk/downloads/Cumulus.zip

[gerardwood]

Hi Steve,

That looks a lotbetter... the last few lines of the monthly log, 20:18 being the restart time:

05/01/12,20:17,8.2,75,4.0,7.2,15.0,326,0.00,0.05,1019.0,14.58,20.4,58,7.8,6.2,8.2,0.0,0,0.000,0.000,4.6,0,0.0,315
05/01/12,20:18,8.2,75,4.0,6.8,15.0,326,0.00,0.05,1019.0,14.58,20.4,57,5.1,6.3,8.2,0.0,0,0.000,0.000,4.7,0,0.0,315
05/01/12,20:18,8.2,75,4.2,5.0,10.7,321,0.00,0.05,1019.0,14.58,20.4,57,1.1,6.9,8.2,0.0,0,0.000,0.000,5.3,0,0.0,315
05/01/12,20:19,8.2,75,4.2,4.5,10.7,321,0.00,0.05,1019.0,14.58,20.4,57,1.3,7.1,8.2,0.0,0,0.000,0.000,5.3,0,0.0,315
05/01/12,20:20,8.2,75,4.0,3.9,10.3,313,0.00,0.05,1019.0,14.58,20.4,55,10.3,7.3,8.2,0.0,0,0.000,0.000,5.7,0,0.0,315
05/01/12,20:21,8.2,75,4.0,5.4,12.1,325,0.00,0.05,1019.0,14.58,20.4,56,5.6,6.8,8.2,0.0,0,0.000,0.000,5.2,0,0.0,337
05/01/12,20:22,8.1,75,3.9,6.4,14.3,330,0.00,0.05,1019.0,14.58,20.4,56,14.3,6.3,8.1,0.0,0,0.000,0.000,4.8,0,0.0,337

Thanks very much for the quick turnaround!

Gerard.