Incorrect weather extremes displayed since yesterday morning.

[Bristolian]

Hello,

Since turning on my PC yesterday morning, Cumulus MX is displaying wildly inaccurate extremes for both temperature and rainfall when loading the archive data from starting MX.

The current weather conditions it displays are correct, but not the extremes.

I am running the latest version of Cumulus MX, on Windows 11. My station is the Davis Vantage Vue.

If anyone can help, I'd be very grateful.

Very Best Wishes,
Bob.

[HansR]

Could you switch debug/data logging on; switch CMX off and wait at least one interval (to get some archive data to load) and start CMX again. When normal running for at least one interval, please post the logfile.

The latest version of CMX... does that mean b3231?

[Bristolian]

Could you switch debug/data logging on; switch CMX off and wait at least one interval (to get some archive data to load) and start CMX again. When normal running for at least one interval, please post the logfile.

The latest version of CMX... does that mean b3231?

Hello Hans,

Yes, I'm running build 3231,
Could you please explain how to turn debugging on, and what one interval would be?.

Thank you .

Very Best Wishes,
Bob.

[HansR]

Could you please explain how to turn debugging on, and what one interval would be?.

Switch on logging: Program Settings=>Logging Options : tick the ones you need (debug and data)

The interval you configured for the internet : Internet Settings=>Interval Configuration=>Normal Interval Settings

If you restart immediately there is no data in the logger and you need one interval at least to see what is send to the internet (if you have that) If you don't use the internet at least wait some 10 to 15 minutes to see the cycles CMX is using.

[Bristolian]

Could you please explain how to turn debugging on, and what one interval would be?.

Switch on logging: Program Settings=>Logging Options : tick the ones you need (debug and data)

The interval you configured for the internet : Internet Settings=>Interval Configuration=>Normal Interval Settings

If you restart immediately there is no data in the logger and you need one interval at least to see what is send to the internet (if you have that) If you don't use the internet at least wait some 10 to 15 minutes to see the cycles CMX is using.

Thank you Hans,

I've just done as you suggest.

Very Best Wishes,
Bob.

[Bristolian]

Evening everyone,

Which logfile should I post?.

Very Best Wishes,
Bob.

[water01]

The latest .txt file in the MXDiags folder, all the logging is put in there. You may have to zip it up if it is large.

[Bristolian]

Thank you David,

I've attached the file as a zip file.
This morning, the minimum temp and maximum temp (both at 0827) are only .4C apart, vastly different from the previous few days.
The rainfall totals are hugely wrong too...

Very Best Wishes,
Bob.

[HansR]

Hi Bob,

I see the following for the rain:

Code: Select all

2023-03-10 08:23:28.813 Midnight rain found in the following entry:
2023-03-10 08:23:28.813 09/03/23,07:18,0.7,100,0.7,0,3,0,0.0,0.0,995.0,44.0,17.9,49,0,0.7,0.7,0.0,0,0.00,0.00,-1.2,27,0.0,0,0.0,0.0,0.7,0.7
2023-03-10 08:23:28.815 Midnight rain found, setting midnight rain count = 44
2023-03-10 08:23:28.816 Checking rain counter = 73
2023-03-10 08:23:28.816 Rain counter set to = 73
2023-03-10 08:23:28.817 Getting rain totals, rain season start = 1
2023-03-10 08:23:28.817 Today = 10/03/23
2023-03-10 08:23:28.818 Rainthismonth from dayfile: 2911.8
2023-03-10 08:23:28.818 Rainthisyear from dayfile: 2949.6
2023-03-10 08:23:28.821 Station type = Davis
2023-03-10 08:23:28.821 LOOP2 enabled
2023-03-10 08:23:28.821 Serial device = COM3
2023-03-10 08:23:28.821 Serial speed = 19200

So it is obvious that there is an error at some point because 2911.8 mm in this month alone is somewhat overdone. However the rain for today (44 since midnight) seems not unreasonable. So my guess is that an error has been at some date (that should be easy to find) and from there it propagates through your system. That is correctable but I suggest somebody else takes that to guide you (e.g. @water01 or @PaulMy)

I also find in the file:

Code: Select all

2023-03-10 08:23:29.910 Console clock: 10/03/2023 08:24:23
2023-03-10 08:23:29.910 Console clock: Console is -53 seconds adrift but automatic setting is disabled - you should set the clock manually.

Setting the automatic correction is probably a wise thing. I myself don't have Davis so you have to look in the settings for that.

Then you have a repetitive reading of the archive:

Code: Select all

2023-03-10 08:24:47.337 GetArchiveData: Response data - F4-FF-3F-3F-03-FF-01-FF-01-FF-01-FF-00-00-00-FF-FF-FF-FF-FF-00-E3-03-3F-7F-07-FF-07-FF-FF-FF-FF-FF-FF-FF-FF-00-07-0E-03-00-03-00-00-FF-FF-FF-FF-FF-FF-FF-01-FF-7F-3F-3F-FF-FF-01-FF-01-FF-01-FF-00-00-00-DF-FF-FF-FF-FF-FF-F3-FF-3F-7F-07-FF-07-07-FF-FF-FF-FF-FF-FF-FF-00-07-0E-03-00-03-00-00-FF-FF-FF-FF-FF-FF-FF-01-FF-7F-3F-3F-FF-FF-01-FF-01-FF-01-FF-00-00-00-DE-7F-FF-FF-FF-FF-E3-FF-3F-7F-07-07-07-FF-FF-FF-FF-FF-FF-FF-FF-00-07-0E-03-00-03-00-00-FF-FF-FF-FF-FF-FF-FF-01-FF-7F-3F-3F-03-FF-01-FF-01-FF-01-FF-00-00-00-DF-7F-FF-FF-FF-00-C3-03-3B-77-03-FF-FF-FF-FF-FF-FF-FF-FF-FF-FF-00-07-0E-03-00-03-00-00-FF-FF-FF-FF-FF-FF-FF-01-FF-7F-3F-3F-FF-FF-01-FF-01-FF-01-FF-00-00-00-DF-7F-FF-FF-FF-FF-E3-03-3F-77-03-FF-FF-07-FF-FF-FF-FF-FF-FF-FF-00-07-0C-03-00-03-00-00-FF-FF-FF-FF-FF-FF-FF-01-FF-FF-FF-FF-FF-60-24
2023-03-10 08:24:47.337 GetArchiveData: Loaded archive record for Page=244 Record=0 Timestamp=01/01/1900 00:00:00
2023-03-10 08:24:47.337 GetArchiveData: Ignoring old archive data
2023-03-10 08:24:47.337 1221 archive entries processed
2023-03-10 08:24:47.337 GetArchiveData: Loaded archive record for Page=244 Record=1 Timestamp=01/01/1900 00:00:00
2023-03-10 08:24:47.337 GetArchiveData: Ignoring old archive data
2023-03-10 08:24:47.337 1222 archive entries processed
2023-03-10 08:24:47.337 GetArchiveData: Loaded archive record for Page=244 Record=2 Timestamp=01/01/1900 00:00:00
2023-03-10 08:24:47.337 GetArchiveData: Ignoring old archive data
2023-03-10 08:24:47.337 1223 archive entries processed
2023-03-10 08:24:47.337 GetArchiveData: Loaded archive record for Page=244 Record=3 Timestamp=01/01/1900 00:00:00
2023-03-10 08:24:47.337 GetArchiveData: Ignoring old archive data
2023-03-10 08:24:47.337 1224 archive entries processed
2023-03-10 08:24:47.337 GetArchiveData: Loaded archive record for Page=244 Record=4 Timestamp=01/01/1900 00:00:00
2023-03-10 08:24:47.337 GetArchiveData: Ignoring old archive data

and I conclude that it does not read correctly the archive. That may be the cause for the erroneous data. It maybe that your logger is incorrect.

Then it does the day reset:

Code: Select all

2023-03-10 08:24:47.345 === Day reset, today = 10
2023-03-10 08:24:47.345 === Day reset for 09/03/2023 00:00:00
2023-03-10 08:24:47.354 Dayfile.txt entry:
2023-03-10 08:24:47.354 09/03/23,12,285,22:31,-25.2,08:45,3.7,23:08,0.0,08:45,995.1,09:10,4.8,20:43,28.8,2.5,6.8,3,08:45,0,08:45,100,08:45,0.00,0.0,3.7,23:08,1.7,23:11,-30.1,08:45,969.6,09:19,-30.0,08:45,3.7,23:08,0.0,08:45,316,2.4,0.0,0,08:45,0.0,08:45,3.7,23:11,-30.0,08:45,3.7,23:08,384185.8,1916.0,23:07
2023-03-10 08:24:47.354 Dayfile.txt opened for writing
2023-03-10 08:24:47.354 Writing entry to dayfile.txt
2023-03-10 08:24:47.355 Dayfile log entry for 09/03/23 written
2023-03-10 08:24:47.355 Raincounter = 73 Raindaystart = 44.2
2023-03-10 08:24:47.355 Rainyesterday (calibrated) set to 28.8
2023-03-10 08:24:47.355 RainDayThreshold = -1
2023-03-10 08:24:47.356 rdt1000=200 ry1000=28800
2023-03-10 08:24:47.356 Yesterday was a rain day
2023-03-10 08:24:47.356 Consecutive rain days = 3
2023-03-10 08:24:47.356 Raindaystart set to 73
2023-03-10 08:24:47.356 StartofdayET set to 0
2023-03-10 08:24:47.358 Writing today.ini, LastUpdateTime = 09/03/2023 23:15:00 raindaystart = 73 rain counter = 73
2023-03-10 08:24:47.362 Writing yesterday.ini
2023-03-10 08:24:47.362 Written yesterday.ini
2023-03-10 08:24:47.395 If required the daily graph data files will be uploaded at next web update
2023-03-10 08:24:47.395 === Day reset complete

That seems pretty good and I assume you can confirm yesterdays rain?

Then it starts its loop to fetch data and apparently it is raining (at time of fixing this log):

Code: Select all

2023-03-10 08:24:59.185 LOOP: 7 - Data packet is good
2023-03-10 08:24:59.444 Realtime[5]: Start cycle
2023-03-10 08:24:59.444 Realtime[5]: Creating realtime.txt
2023-03-10 08:24:59.444 Realtime[5]: Creating realtime file - realtimegauges.txt
2023-03-10 08:24:59.447 Realtime[5]: End cycle
2023-03-10 08:25:00.029 DoLogFile: Writing log entry for 10/03/2023 08:25:00
2023-03-10 08:25:00.029 DoLogFile: max gust: 12
2023-03-10 08:25:00.030 DoLogFile: log entry for 10/03/2023 08:25:00 written
2023-03-10 08:25:00.030 Writing today.ini, LastUpdateTime = 10/03/2023 08:25:00 raindaystart = 73 rain counter = 81.6
2023-03-10 08:25:00.033 Creating standard web files
2023-03-10 08:25:00.076 Done creating standard Data file
2023-03-10 08:25:00.076 Creating graph data files
2023-03-10 08:25:00.121 Done creating graph data files

I would say the realtime loop is working.

So from what I see in the log I think the following may be the cause of the problems:

1. The logger does not function (it may be broken?)
2. The time is incorrectly set
3. You require correction of the rain in CMX

I think it is best to look for a second opinion (@Mark) on how to proceed with the logger.
For the time being I would not shutdown CMX but let it run overnight until decided on how to proceed. This avoid the dependency on the logger.

[Bristolian]

Thank you Hans, and David for your assistance.

I can't confirm yesterday's rainfall total, as I was away at work from just after 10am, until about 8.30pm.
I can confirm though that it certainly rained well, once the snow had started turning to sleet!.

Very Best Wishes,
Bob.

[mcrossley]

Indeed, your station dumped the entire logger contents back to Cumulus when asked for some recent data. It does this when it cannot find the date/time requested in the logger.

But the entire contents of the logger was invalid - with a zero'd date/time.

The first thing I would do would be to do a full power cycle of the console - batteries and power out for a few minutes.
Then change the console logging interval to 5 minutes, then back to your desired 1 minute. Doing that makes the console reinitialise the logger contents.

After that you can check if it is working for future Cumulus restarts.

[Bristolian]

Indeed, your station dumped the entire logger contents back to Cumulus when asked for some recent data. It does this when it cannot find the date/time requested in the logger.

But the entire contents of the logger was invalid - with a zero'd date/time.

The first thing I would do would be to do a full power cycle of the console - batteries and power out for a few minutes.
Then change the console logging interval to 5 minutes, then back to your desired 1 minute. Doing that makes the console reinitialise the logger contents.

After that you can check if it is working for future Cumulus restarts.

Morning Mark, thank you - I'll give that a whirl.

Very Best Wishes,
Bob.

[Bristolian]

I've disconnected the console, removed and replaced the batteries,
CMX still shows the previously reported excessive rainfall, and the temperatures as of the initial daily start of CMX earlier this morning.

image_2023-03-10_103508308.png

I'll leave the station on overnight and see what happens.


Very Best Wishes,
Bob.

[mcrossley]

Fixing the logger and fixing the data already in Cumulus are two different things.

Unfortunately fixing the data is going to be a bit painful.

[Bristolian]

Fixing the logger and fixing the data already in Cumulus are two different things.

Unfortunately fixing the data is going to be a bit painful.

Thanks Mark.