1438   Question: representation of db, dbAng, zeroDb

This tissue has following status: green

Created: 17 Sep 2015

Links: #1437 Deadband, dbAnd, zeroDb units

Page: 64, 77

Clause: 8


Category: Issue for edition 2 of this part

Issue: Is it prohibited:
- to declare db in percents from previously transmitted value?
- to declare dbAng in degrees?
- to declare zeroDb in absolute value?
, describing that in MICS/PIXIT, of course.

Proposal: Just a question, see also my tissue #1437 (part 7-1).

Discussion Created Status
Ballot until Editor
solved with proposal in Ed 2.1 / CDV which has been circulated to NCs and approved 26 Apr 18 green
This needs more discussion in the Working Group. 12 Feb 16 yellow
The type of RangeConfig.min and .max is AnalogValue.
Therefore dbRef should be AnalogValue and NOT FLOAT.
Also, the comparison should be dbRef is exactly equal to zero.
Also, the constraint on dbRef should be "non-negative".
06 Nov 15 final proposal
Today, we know how to interpret deadband, if we have min and max attributes present.
We have two issues not solved
(1) In some cases, it is reasonable to have a deadband value that depends on the actual value
(2) In some cases, we still want to have a deadband value that is constant (independent of the actual value), but we can not express a range of min / max, because semantically, there is no min or max.

Final proposal:
Add a new attribute dbRef (or zeroDbRef, dbAngRef) with FC=CF, trgOp=dchg, INT32U that represents the deadband reference. If the deadband reference is present and >0, db would be redefined to: "The value shall represent the percentage of deadband reference (dbRef) in units of 0,001 %".

To solve case (1), it will be defined that, if the value of dbRef is 0, the value db shall represent the percentage of the last transmitted values in units of 0,001 %. Note that for zeroDbRef, a value of 0 shall not be selected.

The presence condition of dbRef shall be mandatory if db is present. (same for dbAng / dbAngRef and zeroDb / zeroDbRef)

22 Oct 15 final proposal 21. Nov 2015
I'm in favour of allowing dead bandcalculation related to percents from previously transmitted value, at least when the attribute max and min are not provided, instead of related to percents of (max-min).
A percentage from previously transmitted value allow to
1) avoid the parameterization of each max and min attribute
2) a dead band sensitivity constant over all the range instead of a range dependant sensitivity.

19 Oct 15 red


Privacy | Contact | Disclaimer