P
1
0
2
3
8
-0
0
-0
G
0
-0
3
·
0
1
.2
0
0
4
·
S
al
es
a
nd
d
el
iv
er
y
in
a
cc
or
da
nc
e
w
it
h
ou
r
te
rm
s
of
s
al
e
an
d
de
liv
er
y.
·
S
ub
je
ct
t
o
te
ch
ni
ca
l
al
te
ra
ti
on
s.
Fire detection MSR 3000
The biocybernetic sensor technology
The product range
F
IR
E
D
E
T
E
C
T
IO
N
T
E
C
H
N
O
L
O
G
Y
Approval
The exchange module for the
optical smoke detector.
A technical pioneering achieve-
ment
With this technically developed concept,
great importance was attached to simple
installation technique and possible com-
ponent exchange. For the first time, the
installer is capable of exchanging “worn-
out parts” himself, for example exchan-
ging the measurement chamber quite
simply: Disassemble detector, open
detector, remove soiled measurement
chamber and replace it with a new one.
In this way the annoying problem of
changing the detectors is eliminated.
The disposal of the optical measurement
chamber can be carried out by the
installer himself. The disposal of the
ionization measurement chamber is off
course carried out as usual by Novar.
You know yourself; Service costs time
and time is money. With the SDN- and
MSR optical smoke detectors the length
of time for maintenance and for the
exchange is reduced to a minimum.
Novar GmbH
Johannes-Mauthe-Str. 14
72458 Albstadt• Germany
Phone +49 (0) 07431/801-0
Fax+49 (0) 07431/801-1220
Dieselstr. 2
41469 Neuss• Germany
Phone +49 (0) 02137/17-1
Fax+49 (0) 02137/17-286
Internet
www.novar.de
E-mail
info@novar.de

»»Ò³
The new fire detection series
A convincing system technology
The MSR detector is a new detector
generation with the performance cha-
racteristics:
• differential signal processing
• detector-determined behaviour
pattern
• permanent characteristic curve
correction
• system-determined parameterization
• decentralized intelligence
Stan
dard
SDN
MSR
S 3000– The pyramid of efficiency
MSR - optical smoke detector
multiple criteria evaluation with integrated heat detector
SDN - optical smoke detector
with adaptive signal evaluation and tracking
SDN - ionization smoke detector
with adaptive signal evaluation and tracking
M = Multiple criteria evaluation
S= Signal analysis
R= Raster cognition
2
3
So in order to raise the intelligence of
the detector the variety of information
stored in the detector head is drawn
upon when making a decision. With the
information- and microsystem technology
new technologies are available, in order
to process even more data and informa-
tion. A software-controlled algorithm
and plausibility checks integrated in the
detector head together form a new
generation of fire detectors - the biocy-
bernetic detector for automatic early
fire detection.
The MSR detector
There is not necessarily a fire where
there is smoke. Only an intelligent fire
detector can comprehend and evaluate
this subtle but important distinction.
In order to achieve this, we have develo-
ped the MSR detector. A multisensor
system, consisting of several closely
arranged sensors.
The product range S 3000
The reliability and availability of a fire
detection system is determined decisi-
vely by the detection reliability of the
connected detectors. Automatic detec-
tors must preferably be sensitive and
able to react quickly, but on the other
hand must not to respond to deceptive
influences.
It was the aim of Novar to develop a
fire detector, which combines highest
detection sensitivity with the smallest
possible rate of false alarms.
This is not due to magic but due to the
experience in the fire detection over
many decades. A great number of smoke-
and fire tests have been carried out.
These developments have been recor-
ded and analysed in our research labo-
ratories.
Incidently we have taken nature as a
model.A bionically inspired and compu-
ter-supported model has led to a new
evaluation of the algorithm.
Each sensor delivers partial information
with a specifically usable information
content. In connection with a sensor-
determined signal processing procedure
the stored data is boosted in order to
reach the necessary standard required
for further signal processing. The raster
cognition supports the intelligent data
analysis and vouches for a competent
diagnosis of the stored data and the
initiation of effective corrective measures.
These preventive measures ensure that
false alarms are practically eliminated
even with complete detection sensitivity
and reliability.
062260
062650
062550

»»Ò³
4
5
In connection with the RS bus technology
the analogue digital instantaneous
values of the measuring signal are
transmitted to the fire detection control
panel. The data from adjoining detectors
is compared and drawn upon together
for decision making. If for example only
one detector detects some smoke, and
neighbouring detectors have indicated
no change, then there is no alarm signal.
However an alarm can be signalled in
good time if several individual detectors
determine only a little smoke.
Interactive data communication means:
A reliable monitoring and alarm
detection even in critical sec-
tors.
The new fire detection series
A convincing system technology
Continuous compatibility - to the
delight of both the planner and
the installer
There have been Novar fire detection
systems for a long time.
Off course the new detector generation
is technically and functionally compa-
tible with existing systems.
Off course new detectors can also be
operated together with the previous
Novar generation and integrated into
existing systems. A gradual adaption
and upgrading and the consequent
generation change is therefore possible
without any problems.
A maxim which Novar carries on,
guarantees the necessary detector
availability for decades to come.
Decentralized intelligence and
interactive data communication
with the control unit
SDN- and MSR detectors have a de-
centralized intelligent signal processing,
and are therefore on the spot with
“eyes and ears”.
The detectors are not only very reliable,
but are also just as fast no matter how
many detectors are connected. In this
way microcomputer-controlled fire
detection systems are implemented
with decentralized intelligence, in order
to increase the reliability and constant
availability of the total system.
Due to their stand alone intelligence
the detectors can also be used as intel-
ligent individualdetectors and with the
same detection sensitivity on conventio-
nal d.c. lines, for example in smoke pro-
tection switches and test chamber
detectors.
See and be seen
By using the fire detection computer
BMC 1024-F and the ring bus technique
it is possible to display the measuring
signal and its progress in time graphi-
cally on a large screen.
If the progress of the fire reaches the
defined pre-alarm threshold then a cyclic
transmission of the fire phenomena
automatically takes place from the
detector to the control panel.
The instantaneous values of the trans-
mitted digitized data are now stored in
the control panel and are then available
for documentation purposes.
In this way the data of up to 32 detec-
tors can be stored and when needed,
represented graphically on the display.
The operator himself can decide whether
or not the data should be displayed as a
number format or as a time-depending
fire process. Off course you can readout,
document and if required graphically
display the measuring signal in the dia-
gnostic mode or when carrying out the
one-man revision of the detector.
standard
SDN - function
MSR - function
direct current technology
MEI - bus technology
RS - bus technology
detector base
detector head
Stand alone intelligence – for continuous compatibility

»»Ò³
7
6
Reliable and economical
A convincing system technology
The system structure
SDN- and MSR detectors were develo-
ped according to an up-to-date compu-
ter-supported model and contains a
selective and quantized algorithm for
meaningful information evaluation.
The MSR detector always includes two
high-quality sensor systems, which are
accomodated in one housing and opera-
te as independent units and react to
different fire phenomena (smoke and
heat).
Each spot-type sensor system has its
own response sensitivity adapted to
specific EN-54 requirements. Signal
analysis, signal evaluation and detection
algorithm are independent and run para-
llel without influencing each other.
The identification and differentiating
evaluation of the measuring signals
remain the most important factors of
the system.
With help of the raster cognition “unwan-
ted measuring signals” are evaluated,
unwanted signals filtered out and if
required the necessary drift characteri-
stic curve is corrected.
In practice this leads to a well-balanced
detection- and response sensitivity with
all types of actual fire developments
with a broad spectrum of different fire
phenomena.
When using conventional d.c. detector
bases a logical OR operation of both
measuring signals is carried out directly
in the detector head.
Therefore it is possible to con-
nect intelligent SDN- and MSR
detectors to conventional d.c.
lines.
In connection with the RS bus technology,
the analogous instantaneous values of
the measuring signal are digitally trans-
mitted to the fire detection control
panel.
Therefore an interactive working
fire detection system is imple-
mented, through which different
types of detectors are recorded
in the control panel, linked each
other and deciding together
whether or not a true alarm
exists.
?
t
t
smoke
heat
fire phenomena
typical fire
phenomena features
t
t
t
A
intensity
amplitude
?
t
time function
signal function
t
t
t
A
intensity
amplitude
?
t
time function
signal function
t
E
climate
environment
pollution
electromagnetic
interference pulse
physical
deception values
smoke alarm
heat alarm
?
- interference
analogue value
Full custom IC (ASIC)
Signal processor in detector head
detector head
detector base
model formation
signal analysis
signal evaluation
raster cognition
unwanted measuring signals
?
- alarm
alarm line
remote setting and
parameterization
?
- interference
pollution
function
RS bus
technology functions
detector addressing
alarm activation
interference indication
activation
setting and changing of
the correlation factors
switching on or switching
off of the different sensor
systems
day/night operation
analogue value - measuring
signal transmission
G-016
t
t
S

»»Ò³
UB
optical
smoke detector
heat detector
?
digitization
and quantizing
of the measuring
signals
signal analysis and
raster cognition
detector
interface
direct current
MEI - bus
RS - bus
heat
12:45
time
smoke
?
- alarm
?
- interference
pollution
analogue value of the
measured signals
remote parameter
setting
inactivating
of smoke channel
setting and changing
of the response sensitivity
detector base
RS-bus technology
alarm line
memory
algorithm
alarm line
detector head
9
8
Signal acquisition –
digital and intelligent
Biocybernetics – for an
automatic early fire detection
Signal acquisition and signal
analysis in the detector head
Every actual fire development is diffe-
rent and is not reproducible in its time
behaviour and therefore there is no
mathematical conformity. Therefore
special signal theoretical models have
to be developed in order to attain an
evaluation. So we were able to recognize
that a more qualified judgement of a
fire phenomenon can be made, if the
selective methods can be adjusted to
the physical measured value.
The incoming sensor signals are im-
mediately digitized and fed to our own
developed high performance signal pro-
cessor.
With the aid of a digital filter the mea-
suring signal is freed from interference
influences and pulse spikes and is now
debugged and available for further signal
evaluation.
Inspired by nature
Sensitivity in nature means a neveren-
ding adaption to the presently inhabited
and animated environment. In order to
be able to exist in this world, a highly-
developed perceptivity is required. Eyes
not only enable sight and ears not only
hearing, but both sense organs comple-
ment each other and are also able to
correct errors. The human being is a
complex system of biocybernetical adap-
tion to perceptivity. The sense organs
are always active, however different
levels of sensitivity depend on the envi-
ronmental conditions. The more dange-
rous the situation, the more sensitively
we react.
This biocybernetic algorithm is the
essential characteristic of the new
generation of SDN- and MSR smoke
detectors.
In “stand by” operation generally insen-
sitive, the sensitivity of the detector
adapts itself individually, and automati-
cally and therefore optimal to its speci-
fic surroundings and to a dangerous
situation. A detector which learns to be
sensitive, when a dangerous situation
arises.
1
2
3
4
5
6
7
t ( duration)
% - measured value
output value of
the correlation
quick
development of fire
(open fire)
slow
development of fire
(smouldering fire)
start of
adaption
20
40
60
80
100
alarm
alarm
120
140
detection limiting value
adaptable to the actual
fire development
Adaptable adjustment of the detection sensitivity to the actual fire development

»»Ò³
11
10
Raster cognition for filtering-out
of disturbance variables
MSR 3000 –
a convincing system technology
Signal analysis and adaptive
tracking of the response thres-
hold
Traditional smoke detectors have simple
evaluation electronics. The acquired
sensor signal is intensified and triggers
an alarm when a fixed response thres-
hold is exceeded. Should a fire develop-
ment be detected in the early stage
using this evaluation electronics, then a
high response sensitivity is needed and
therefore a lower static alarm thres-
hold.
This rigid structure of the decision pro-
cess would only guarantee an optimum
response sensitivity in a few special
cases, but in the case of fire phenomena
of similar criteria, wrong decisions would
be made and undercertain circumstances
trigger false alarms.
As every fire development is different in
its intensity and its speed, it is obvious
that the response sensitivity has to be
optimally adapted to the different fire
phenomena. The diagram of the adapta-
ble adjustment of the detection sensiti-
vity shows the correlation between the
fire development and response thres-
hold.
The raster cognition for the
filtering-out of disturbances
The cyclic, digitized instantaneous expo-
sures are fed to an arithmetic logic unit
(ALU) and by means of a pattern com-
parison are compared according to fixed
stored tabulated values. An important
role in the development of this sensor
system is the optimization of these cha-
racteristic curves, which were empiri-
cally determined by means of compre-
hensive test series of a model character.
With the aid of raster cognition the
relationship to the typical fire deve-
lopment can be acquired and evaluated.
If the instantaneous exposure does not
fall into the predetermined raster, then
the signal is filtered out for further
alarm evaluations and checked for other
criteria such as: soiling, aging, tem-
perature drift or electromagnetic inter-
ference pulses. Depending on the raster
cognition the appropriate measures are
carried out.
Future-oriented successive
tracking of the output value
The SDN- and MSR smoke detector can
naturally not prevent environmental pol-
lution, but there is no false alarm and,
what is still more important, it keeps
its response sensitivity programmed ex
works in its broad field of application.
In this way the detection reliability and
the unchanged response sensitivity
remain the same even after long use
just as they did on the first day of ope-
ration.
In comparison to conventional smoke
detectors, the SDN and MSR detector
is very insensitive with a relatively high
correlation standard value. If smoke
aerosols are now measured, the res-
ponse value of the measuring signal
starts at a defined threshold. The alarm
signal is the intersection point between
the dynamic response threshold and the
measuring signal.
With this algorithm one achieves, inde-
pendent of fire development, a high
detection sensitivity and a high immunity
against possible physical levels of
deception. The detector is sensitive
over a broad band and equally suited
to all type of fire developments.
permissible range of
identification and
pattern recognition
measured
signal
unwanted signals:
environmental in-
fluences, climate,
pollution, drift
t - duration
20
40
60
100
unwanted interference signals
physical size of deception
80
1
2
3
4
5
6
7
120
electric disturbing pulse
1000
0111
0110
0101
0100
0011
0110
0001
1
2
3
4
5
t
measured value (Bit)
Raster cognition - for filtering-out disturbance variables
measured value
standard setting
output value of
the correlation
t (h) duration
digitized
measured value
measured
value
?
?
digitized
measured value
correlation
standard value
drift tracking of the
correlation standard value
= constant sensitivity
limit A -
limit B -
limit B +
limit A +
correlation
standard value
measured
value
UK
UK
interference signal
diagnosis
diagnosis
interference signal
UK = const.
Long-time response - Tracking of the correlation standard value
SDN- and MSR detectors constantly
collect the stored physical data and
the environmental conditions which
influence it. All data parameters are
fed to a signal processor. In the case
of a drift away from the set value, the
signal processor calculates an optimum
setting and corrects its output value.
In its normal operating condition the
detector immediately signals whether
or not there is an interference or soiling.
In the diagnostic mode the detector
already recognizes future developments.
From the stored measuring signal the
detector determines the rate of change
and reports to the fire detection control
panel, whether or not its soiling and
ageing is so far advanced that it would
be advisable as a precaution to exchange
or to clean the smoke detector.

»»Ò³