Trend History Files vs variable tag engineering range change

Hi,

I have a variable tag defined with an engineering range of 0-30 DegC, and that variable tag drives a trend tag which is set up to log over 12 trend history (log) files. If I then need to change the engineering range of the variable tag to 0-60 DegC, am I right in saying that the already created and logged to trend history files are not affected by this change and do NOT need to be deleted/re-created? Would they be capable of handling a value of e.g. 55 which the original engineering range would not allow?

As far as I know, deleting the trend history files would only be necessary if one of the timing (No. Files, Period, Time, Sample Period) or Storage Method fields changed on the trend tag definition.

Please confirm.

Thanks

Ash

  • It depends on whether you are using 2-byte or 8-byte storage method for this trend tag--check the trend tag definition.

    For 2-byte trends (AKA scaled trends), if you change the scale of the variable, then existing data for the trend will be read back incorrectly because Citect assumes it was all recorded with the current scale. To avoid this, you need to delete the data files for that trend tag after making the tag change. If you need to keep the data, try converting the trend to an 8-byte trend with Trend Converter.

    For 8-byte trends, the actual engineering value is stored with no additional scaling so it doesn't matter if you change the raw or engineering range.

    See TN3428

  • Hi Erik,

    Thanks for your quick reply. Happily, all my trend tags are defined with Storage Method=Floating Point (8-byte samples), so it seems I can change the source variable tag engineering range and the trend history files will be unaffected.

    That was the answer I was hoping for!

    Thanks,
    Ash