Experimental Physics and
Industrial Control System

Maren Purves via Tech-talk <[email protected]> · Mon, 18 Nov 2019 10:21:53 -1000

We put in an extra record to calculate the logarithm and archive that

(same application: vacuum).

Maren Purves
East Asian Observatory/JCMT

On 2019-11-18 07:08, Jemian, Pete R. via Tech-talk wrote:

To archive a signal that ranges over many decades, separate the mantissa and exponent into two separate values, each to be logged.

----------------------------------------------------------
   Pete R. Jemian, Ph.D.                <[email protected]>
   Beam line Controls and Data Acquisition
   Advanced Photon Source,   Argonne National Laboratory
   Argonne, IL  60439                   630 - 252 - 3189
-----------------------------------------------------------
      Education is the one thing for which people
         are willing to pay yet not receive.
-----------------------------------------------------------

________________________________________
From: Tech-talk <[email protected]> on behalf of Kasemir, Kay via Tech-talk <[email protected]>
Sent: Monday, November 18, 2019 8:51 AM
To: EPICS tech-talk; Bo Jakobsen
Subject: Re: [EXTERNAL] ai records with data covering many decades

Hi:

..PV for a pressure gauge which by nature covers many decades of values.
I wanted to set up MEDM and ADEL values before starting to log the values to our archiver.
Output is in exponential format, and i would like to monitor on e.g. 1 decimal independent of the exponent, it this possible.

I'm not aware of a good solution for efficiently displaying or archiving log-scale-type-PVs via MDEL respectively ADEL.
As you say, end users would like to see updates based on for example changes of "0.1", no matter if that's "0.1E-3" or "0.1E-5" on a log scale plot of vacuum pressure readings.
Sometimes the MDEL/ADEL is set to say "0.1E-5" because the reading happened to be in the E-5 range at the time, and then the vacuum improves to the E-6 range and MDEL/ADEL suppresses all updates.
So the only practical solution for archiving is to set ADEL=0, and store the value every few seconds to limit the amount of data. That's OK for the longer-term histor, but with vacuum readings you tend to care about some detail in vacuum excursions which start as a sharp spike, followed by exponential decay. With just a value every few seconds, you miss the initial spike detail.

Thanks,
Kay

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System