Our full technical support staff does not monitor this forum. If you need assistance from a member of our staff, please submit your question from the Ask a Question page.


Log in or register to post/reply in the forum.

Calibrate Command


Terrapin Aug 12, 2015 10:52 PM

Hi Folks,

I have a large CR1000 program that is throwing a lot of SkippedSystemScans. The program has a 20-second scan interval, reads numerous SDI-12, differential voltage, pulse channel, and a serial output sensor. It also has a slow sequence scan of 60-seconds to read SDI-12 output from a CH200 unit.

I've been reading up on the Calibrate command and would like to play with this as a way of reducing, if not eliminating, the SkippedSystemScans. My question is: Regardless of whether the parameters are specified in this command or not, are the calibration coefficients always written to the appropriate Status Table fields? I would assume so, but it's not clear from the manual if so.

Also, what is the recommended slowest execution interval to use for this command?

Thanks.


aps Aug 13, 2015 10:25 AM

The calibration process runs in background and the calibration process is split up into stages to minimise the disruption to the main program. If you get the occasional skippedsystemscan it does not matter too much as this just means that one step of the rolling calibration is held up, temporarily. The only down side to this is it means the logger might not respond to any change of calibration caused by a rapid change of temperature so quickly.

If you put the calibrate instruction in the main scan the logger will then do all parts of the calibration process in one go which will hold up the measurements whilst this happens. The calibrate instruction does all the necessary calibrations are updated in the logger, whether you choose to store date in variables or not. If you put the calibrate instruction in its own slow sequence the disruption to the main scan is not so bad but this will be little different to letting the logger run the calibrate instruction in its own System Scan (which is another slow sequence).

One other thing I would try is putting some of your other SDI-12 measurements in their own slow sequence as serial I/o can run in parallel with normal measurement tasks.


jra Aug 13, 2015 12:27 PM

Here's some general information on skipped scans, system and main:
https://www.campbellsci.com/news-tip-skip

Log in or register to post/reply in the forum.