Hi,
I monitor the lithium battery voltage as part of my program. It's been reporting 3.44 volts for the past 3 months or so. Recently, the lithium battery voltage reported 0 volts for about 21 seconds or so and then it went back to 3.44 volts and has been reporting that ever since. This has happened on three different dataloggers already, in one of them it went to 0 volts for one second and then back to normal. My scan rate is 1 second and ~13 volts are powering the datalogger at all times. Any thoughts?
Thanks.
The lithium battery voltage is only measured by the datalogger once per day and the value held in the status table when it is running normally. The voltage will normally be extremely stable if the logger is powered from another source and its temperature stable.
I believe the battery voltage is also read when the logger is reset or its program is recompiled. I cannot think of a reason even in those circumstances why it would read zero volts though for a very short time. It could be that your own program is overwriting the variables holding the voltage or the logger itself have overwritten the status variable holding the measurement possibly as a result of a crash. Does the status table for the logger show a non-zero value for Watchdog errors?
Hi Andrew,
Thank you for your response. There are no watchdog errors reported in any of the dataloggers that have shown this situation and the program has not recompiled since it was loaded in January 2011.
I made sure that I was not overwriting the variable to which I'm saving the voltage. The instances below are the only ones in the program in which I'm using the variable.
LithiumBatt = Status.LithiumBattery
And then I have the following instruction in a one-minute interval data table.
Minimum (1,LithiumBatt,IEEE4,False,False)
The panel temperature is stable and the datalogger has acceptable power during the times in which has happened.
I went a step further and checked the value directly in the status table and it indeed reports the Lithium Battery as 0.00 V during "x" amount of time (in seconds) and then going back to the previous value and remaining there.
Hi Andrew,
Thank you for your response. There are no watchdog errors reported in any of the dataloggers that have shown this situation and the program has not recompiled since it was loaded in January 2011.
I made sure that I was not overwriting the variable to which I'm saving the voltage. The instances below are the only ones in the program in which I'm using the variable.
LithiumBatt = Status.LithiumBattery
And then I have the following instruction in a one-minute interval data table.
Minimum (1,LithiumBatt,IEEE4,False,False)
The panel temperature is stable and the datalogger has acceptable power during the times in which has happened.
I went a step further and checked the value directly in the status table and it indeed reports the Lithium Battery as 0.00 V during "x" amount of time (in seconds) and then going back to the previous value and remaining there.
Can you please tell me which operating system version you are using and the datalogger type so we can check this out.
My dataloggers are CR1000-XT and I have OS 20 loaded in memory.
I understand there is a newer OS but I cannot upgrade it in my field dataloggers yet.
Thanks.
I have run a simple test program in OS20 and 22 and do not see the lithium battery voltage jumping. (I recorded the minimum value over long time periods with the logger running at 1 sec.)
This does not mean there is not a problem but I think someone needs to see your program to investigate this further.
Andrew,
I can send my program to Dana which is the person that I have worked with previously. However, I want to mention that I have 12 of these CR1000 dataloggers all running the same program since January 2011 (without any watchdog errors or crashes). Only 4 of these have reported this error, once each, over this 5-month period. Since it seems to be something random, and not happening on all dataloggers, I'm not confident that you would be able to recreate it right away. I don't understand how the program could affect the status table measurement that is being reported though. What's reported on the Status table is something that the datalogger performs internally regardless of the program running, right? Assuming that power is available and the program does not crash.
Thanks.
What you say is theoretically correct. However, there have been a few circumstances found where either due to the construction of the users program or some issue with the firmware the logger can write into over areas of memory it should not, which can lead to variables being corrupted. Usually this results in an error message (out of bounds) or a reset but not always. I have seen status values corrupted in the past.
If possible I would advise upgrading the operating system as a couple of such bugs have been uncovered since OS20.
Andrew,
Thank you for your help. I need to load the OS 22 on my workbench first to make sure everything works correctly with my current program. After that I'll load it on my field dataloggers and I will let you know if I run into the same problems again.
Regards,
Tatiana