Re: Meeting in Norfolk, Virginia
- BACnet - Lighting Applications Working Group
5 October 2004
Participants - Affiliation
Steve Karg - Lithonia Lighting
Robert Hick - Leviton Lighting
David Fisher - Polarsoft
Pete Baselici - The Wattstopper
Steve Bushby - NIST
The meeting was called to order at 8:40am after introductions and
greetings from our hosts.
We reviewed the minutes of the last meeting. David Fisher pointed out
correction: DMF011-13 should be DMF011-3. The minutes were approved
The meeting agenda was reviewed and approved.
Pete Baselici reported on NEMA activity. DALI protocol development is
proceeding. Work is being done on selecting parameter banks. The new
joint committee DALI protocol allows multiple controllers. NEMA
working with IES protocol committee.
Steve Bushby reported on a NIST-IAR (inter agency report) on lighting
applications written by Steve Treado that is due to be published soon.
LA-WG submitted proposal discussion
Steve Karg proposed that this committee drop DMF-018-3 in favor of
DMF-011-6 and DMF-019-3. Reported consensus at Nashville was that
DMF-018-3 and DMF-019-3 were redundant if DMF-011-6 was adopted and
therefore not necessary. Group discussion ensued and consensus
emerged that 019 contains simple changes to BO BV that may provide
easy way for mfg to provide on/off lighting control. Group discussed
about when to implement BO / BV instead of LO object. Steve Bushby
moved to abandon DMF-018-3 and retain DMF-019-3 and DMF-011-6 (vote 5
for and 0 against). David Fisher agreed to add language to DMF-019-03
to clarify the use case for BV in addition to BO.
Multiplexer Object proposal STK-004 and Tripped Breaker Status
proposal STK-015 are in the queue for SSPC-135 review.
Scheduling Famous Times
David Fisher presented DMF-028-1 for Famous Times. He explained of
use of "Offsets" as time value relative to Famous time. This proposal
may additionally require the rewriting of several clauses in the
schedule object. David Fisher suggested an alternate method of
encoding of BACnetTimeValue which would utilize the illegal values.
The group discussed how to handle a reference time versus an
enumerated famous time and their possible encoding format.
Hours | Minutes | Seconds | Hundredths
Legal | | |
Values 0-59 | 0-59 | 0-59 | 0-99
Famous | Famous |
Time 0xFE | time |16 bit signed offset
| enum. |
Time 0xFD |array |16 bit signed offset
The group discussed the use of the 16 bit signed integer offset, and
decided that it should have a limited value of +/- 720 minutes, rather
than an unlimited value of a signed 16 bit integer (+/- 32768).
The famous times enumeration was discussed, and it seems that
initially only Sunrise and Sunset may be necessary. The group decided
that the enumeration should include 0-127 as reserved BACnet Famous
Times (ASHRAE reserved values) and 128-255 to allow additions.
Consensus of the group was to go ahead and create proposal without
adding the 135.1 testing requirements to the proposal in order to
float the idea to the committee.
David Fisher will produce DMF-028-2 based on these ideas.
Speed of Lighting Control on the Wire.
Steve Karg described problems he had with sluggishness of BACnet
lighting control. One of his solutions was to use asynchronous
WriteProperty services instead of synchronous. Robert Hick described
need for 125ms response time within field of view (i.e. Ballroom).
Steve Karg reviewed several protocols used in lighting control:
1) DMX 512 which sends a variable length frame which contains the
lighting level for up to 512 devices at 250kbps giving up to 49
updates per second to all devices on the wire. The protocol is a
single master protocol. It is the industry standard for theatrical
2) Proprietary protocol 1 sends a variable length frame containing a
fade time and up to 128 lighting levels (frames can be chained for up
to 2048 lighting levels). The protocol operates at 38.4Kbps and gives
about 300 updates in 100ms. The protocol is a multi-master token
3) Proprietary protocol 2 sends a variable length frame containing a
lighting level and a fade time that affects up to 16 devices and
operates at 57.6kbps. The protocol operates as a single master with
4) The DALI protocol runs at 1200 bps, and can have up to 64 nodes on
the wire. The protocol is designed for multiple masters, and the
lighting devices act as slaves which can reply. The lighting devices
can also be members of groups and can also store scene level
Problems with MS/TP and 156kbps ARCNET speed seem to stem from the
size of the packet, the waiting for acknowledgements, and sending
individual packets to each device. What might possibly work is some
sort of unconfirmed broadcast write service. Another suggestion is
using a "group number," which may not address setting light to
different levels. There was a side discussion of how some
manufacturers store presets levels at the control device versus
storing the preset levels at the dimmer/relay device. The group
discussed the need of other BACnet control sectors that may need a
fast unconfirmed broadcast write. It was suggested that we start by
defining way of accessing presets or scenes.
David Fisher agreed to write a proposal for a broadcast group write of
Modeling enable/disable ON/OFF using out-of-service
Steve Karg discussed need for way to model the disabling control
stations or individual buttons. David Fisher suggested that BACnet
"overridden" would be more appropriate than out-of-service.
The next meeting is planned for Orlando, Florida sometime around
Submitted by Robert Hick with edits by David Fisher and Steve Karg.