search

NOVACProject / NOVACProgram

This program works together with the NOVAC Scanning DOAS system (Scanning Dualbeam miniature Differential Optical Absorption Spectroscopy system).
7 stars 6 forks source link

Incorrect battery voltage written to FluxLog files #154

Open christophkern opened 3 years ago

christophkern commented 3 years ago

When using the 4th Generation NOVAC scanners (Axiomtek, TaskIt etc.), the battery voltage written to the FluxLog files appears to be incorrect. I'm attaching an example file in which all battery voltage values are > 190. Interestingly, the voltages are displayed correctly in the 'Instrument' tab.

FluxLog_2009175M1_2021.09.20.txt

christophkern commented 3 years ago

Sorry, it appears that the battery voltages may not be displayed correctly in the 'Instrument' tab.

mattias42 commented 3 years ago

Not sure what to do with this information, I'm not sure where the value goes wrong and don't know what the correct value should be for the measurements in the FluxLog. Do we have the pak-files that generated this log? If the battery voltage written into the spectrum file differs from what is written to the FluxLog then it could be possible to step through the code and see where the value goes 'bad'. Also, is there a difference between the battery voltage which is printed by kongo in the log files and the battery voltage which the Novac Program reads out from the pak-files?

christophkern commented 3 years ago

I'm not completely sure if the problem is in kongo or in the NOVAC Program. We have an instrument running at CVO, and I connected to it this morning via SSH. I'm attaching a screenshot of the kongo logfile (obtained using 'showlog'). You can see the Battery voltage is listed as 13.17 V for the scan beginning at 15:58 UTC. I'm attaching the pak and txt files for that same scan. In the txt file, you can see the battery voltage is listed as 216.75. So somewhere between the kongo log file and the Evaluation log, the battery voltage is getting jumbled. This only appears to happen for the 4th generation instruments (Axiomtek + TaskIt+ Avantes), so perhaps it's a kongo problem?

kongo_showlog

2009175M1_210924_1559_0.txt

2009175M1_210924_1559_0.pak.zip

mattias42 commented 2 years ago

It seems as if the battery voltage is written incorrectly into the .pak file itself. Opening this attached .pak file. shows that the value ADC[0] written is 21675, which NovacProgram then will convert to a voltage by dividing this integer value by 100 (in order to get a voltage with decimals), hence the battery voltage written to the file is 216.75V.

BatteryVoltage s