Cannot download logger data from La Crosse station

[Matt.j5b]

I copied over the files in the data folder and my cumulus.ini and I started Cumulus MX and put the URL into IE11. I double checked the settings which were all correct. I restarted Cumulus MX but it did not pick up the readings from the station. I am using the La Crosse 2306 with a USB adapter. The COM port was opened and Heavy Weather and Cumulus 1 couldn't open the port. However MX was stuck on downloading logger data from the La Crosse. I have MX set in COM Port 1 and the adapter is set to 1 also. With Cumulus 1 it is takes a few minutes to get the logger data.

Then after 20 minute it connected but the graphs look very strange with two lines of data for the downloaded data. I closed MX down, winded back using the backup and tried again. After 20 minutes it connected but it looks like it did not download from the logger as the Jan15log shows a 6 hour gaps with the strange graphs. But it does update properly with current data and the pages does look nice.

Please find diags and graph image attached.

[steve]

It's reusing the 'old' data that it downloads from the logger because it's checking the wrong 'last update time'. I'll fix that in the next build. I don't know why it was initially unable to download any logger data, we can investigate that if necessary after I fix the other problem.

[Matt.j5b]

I updated to build 3002 but it still fails to download any data from the logger. It was stuck on "WS2300: Downloading history data" in the diags for 5 minute before I closed down MX. This is what it did with the previous build and was stuck for 20 minutes before processing logger entries but there was no sign the logger entries in the monthly log file for this month. I tried it again (after rewinding using the backup) and it has been stuck for 9 minute with the same message in the diags.

[Matt.j5b]

I gave it another go and it got stuck on the downloading history data for 20 minutes, then it downloaded from the logger.

Looking at the diags I see some issues:
2015-01-06 11:40:58.750 Error at line 8374 of data\Dec14log.txt : Constraint
2015-01-06 11:40:58.751 Please edit the file to correct the error
2015-01-06 11:40:58.784 Error at line 257 of data\Jan15log.txt : Constraint
2015-01-06 11:40:58.784 Please edit the file to correct the error

Writing entry to dayfile.txt:
05/01/15,93.3,157,10:05 PM,22.3,10:45 PM,29.3,9:30 PM,1013.9,2:35 AM,1028.2,9:15 PM,260.5,9:15 PM,38.8,23.8,25.6,46.7,10:05 PM,58,9:30 PM,88,11:10 PM,0.00,0.0,31.2,9:30 PM,29.9,9:20 PM,17.8,10:10 PM,21.7,9:15 PM,22.3,10:45 PM,22.2,5:50 PM,18.4,12:25 AM,172,0.0,2.7,0,12:00 AM,0.0,12:00 AM

I though that dayfile is mean't to be 24 hour time without am/pm.

Error at line 1188 of dayfile.txt: Input string was not in a correct format.
Please edit the file to correct the error

2015-01-06 12:02:51.017 Day reset required
=== Day reset, today = 6
=== Day reset already done on day 6


Then after the downloads lots of:

Object reference not set to an instance of an object.

My logger will only hold 14.5 hours of 5 minute readings that I have it set to. It appears that MX has downloaded the logger data from 26 hours ago ( last night). 26 hours ago I ran Cumulus 1 and I then copied the data files from Cumulus 1 to get rid of the mess described yesterday. Then this morning when I ran MX it did not download from the logger but I only waited 9 minutes. So when I tried today it did work but it should have a big gap in the data. But according the "charts" there is no gap. Looking at the log file there is no data between last night and today.

Looking at the graphs I noticed the pressure graph was a bit strange going up and back down. I have restarted Cumulus MX (because the web interface would not work) and the graphs are clearly showing that the logger data was not downloaded, with a line straight from last night to the today with no readings. So no data has been downloaded from the logger. Please find diags attached.

[steve]

I though that dayfile is mean't to be 24 hour time without am/pm.

I'm using the system 'short time format'. I'll have to change it to explicitly use "HH:mm" instead.

So no data has been downloaded from the logger. Please find diags attached.

I'll need to add more tracing to show the data bytes received from the station.

[Matt.j5b]

I upgrade to 3003 and copied across the data from Cumulus 1 to start clean this morning. I ran Cumulus MX and the exe was deleted by Norton as it has with the previous MX versions even though I exclude it. I started up Cumulus MX to see if I get it to download from the logger and it was stuck on downloading history data for nearly 30 minutes. Then it downloaded from the logger (122 records). Downloading from the logger only took 2.5 minutes with Cumulus 1 when I ran it this morning (not at the same time). The time format with the dayfile is correct now, thanks for that. But looking at the graphs the downloaded data is not quite right. The pressure graph for the period of catch up goes straight up 10 hPa or so above the prevous data, then the logged data continues at that level and then goes back down to the expected level and stays there for the live data. The scale is labelled as mb but I have hPa selected in the settings (they are the same but mb in not used here anymore). Also the wind speed graph for logged data is much higher than it should with the scale labelled mph but I use km/hr. This is not as important but the thoroatical solar max is not plotted from the logger data (like Cumulus 1 does) though I don't have a solar sensor.

I am very happy with this new version. It is looking very good and in time I hope that this issue can be resolved which I do hope it can be because I know how hard the la crosse stations can be. Do stay postive, you have made great progress and there is no other software that is as clean and stable as yours. Take your time and you have very nice piece of software.

Some errors are present in the diags:

2015-01-07 07:51:43.905 Error at line 8799 of data\Dec14log.txt : Constraint
2015-01-07 07:51:43.905 Please edit the file to correct the error
2015-01-07 07:51:43.935 Error at line 257 of data\Jan15log.txt : Constraint
2015-01-07 07:51:43.935 Please edit the file to correct the error

Object reference not set to an instance of an object.
Object reference not set to an instance of an object.
Object reference not set to an instance of an object.
---repeats over and over approximately 95 times

During today after copying the data files from Cumulus 1 (which I ran this morning) I ran MX to see how the logger goes. It got stuck on downloading history data for 11 minutes for 70 records. The charts did the same strange things. The wind speed is being recorded live too high ( for 2 hours so far) with it at the moment at 50 and the base station on 0 and this affected the apparent temperature. See this link to view the graphs. The rest of the live data looks all right.

Please find attached diags for today.

[steve]

I've put some tracing of data from the console in build 3004. Turn it on by setting Logging=1 in the [Station] section of Cumulus.ini before starting Cumulus.

[Matt.j5b]

This morning I copied across the data files then ran Cumulus MX. Norton deleted the updated exe file which I excluded from being checked. The logger entries were downloaded and took 15 minutes with 120 records. This morning for the same period Cumulus 1 took a little over 2 minutes to download the data from the logger. Cumulus 1 downloads a logger entry every second or two whereas MX takes around 6 seconds per entry. But don’t get me wrong I do like MX, it just takes quite a while to get the entries. There are also those issues with the charts and are:

Wind: scale in mph and not plotting live data (with a light breeze it did not register with MX)
RH: not plotting live data with a straight line even though the humidity did rise
Temp: straight line for wind chill with the temperature rising when it ran live this morning
Pressure: logger data goes up across and then down and reads one hPa higher than the base unit when ran live.
Solar max: not plotted using logger data, though I don’t have a solar sensor

I ran MX during today after copying the data files from Cumulus 1 that I ran this morning. It took 9 minutes to download 72 entries.

Looking at the graphs:
Wind: scale in mph and the figures are plotted too high, max gust should be about 27 km/hr. When running live it was stuck on same value. Temp and pressure are being plotted live but not wind speed.
Temp: same wind chill issue with data running live and apparent temp jumping up and down too much with the wind speeds too high.
Pressure: Seems right when ran live with logger data plotted higher than it should and then back down to the normal level when ran live.
RH: not plotting outdoor humidity, staying at 69 for last 45 minutes but is now is 68.
Solar max: same issue

Tonight I will run MX to see if the live data issue continues

Please find attached diags for today
See here for this link for the graphs.
Thanks.

[Matt.j5b]

I ran Cumulus MX tonight and the same issues are present. I rewinded the data files to this morning from the files from Cumulus 1 and it downloaded 160 entries which took 20 minutes.

Temperature updates properly when running live while the wind speed and direction and outdoor humidity does not. The outdoor humidity went up slightly according to the base station but was not reflected in Cumulus MX which went in a straight line. The wind chill also went in a straight line while the temperature fell. Also the wind direction changed on the base station (S) but was stuck on SE whilst running live. The pressure updated properly but not with the logger data which was recorded 10 hpa above the normal level.

Diags from tonight attached.

[steve]

I think I have all I need for now, I just need to find time to go through the trace and see what data it is getting and why it is either being interpreted incorrectly or it is being given the wrong values. The other issue is why each read takes six seconds. Other than that, it's actually looking better than I thought, I was expecting to see all kinds of timeouts and read errors, but that seems not to be the case.

[Matt.j5b]

I have restarted Cumulus MX (because the web interface would not work) and the graphs are clearly showing that the logger data was not downloaded, with a line straight from last night to the today with no readings. So no data has been downloaded from the logger.

An update regarding logger data. MX downloads the logger entries but it does not write them to the monthly log file. After running Cumulus MX tonight, I restarted it and the charts plotted a straight line for the period since I last ran Cumulus MX during today.

[steve]

Thanks - I'll fix that.

[steve]

I've uploaded build 3006 which should fix a number of issues with the data. It also adds extra logging to try to establish why each read takes six seconds.

[Matt.j5b]

I updated to build 1006, copied over the data files from Cumulus 1 and ran Cumulus MX. I resolved Norton deleting the updated exe file by right clicking and selecting trust so that it not checked. It took about 17 minutes to download 135 entries. The data downloaded is looking much better, thanks for that.

Temp: the wind chill still runs in a straight line while the temp rises. The other readings look all right. I expect the apparent temperature might be an issue but it is not evident due to very little wind (its early in the morning)
Pressure: issue of logger data 10 hPa higher than expected issue resolved, chart looks good.
Wind Speed : I cannot confirm as there is calm conditions atm. Same for wind direction. However the scale is in mph and is being plotted much higher than it should be.
Humidity: can confirm outdoor rh is now updating correctly when running live.
Rain: No issues
Solar Max: does not plot theoretical solar max for logger data.

I can also confirm that the monthly log file is being updated with data from the logger.
See previously posted link to see the graphs.
I may run Cumulus MX later to see how if the wind speed and direction is recorded correctly.
Please find diags attached.

Thanks.

[Matt.j5b]

I ran Cumulus MX during today to test the wind speed and direction and as before it gets stuck on the same figure when running live. I did restart MX to see if that would make a difference but it got stuck on another value which was evident on the charts. The wind chill issue is still an issue after restarting. The other figures look all right, except for a rain rate of 0.1mm/hr for only a moment, possibly when I restarted Cumulus MX.

I have attached the diags from today.
Thanks for your efforts.