7.2.1 Register addressing
In a parallel 3-ph UPS system, the various UPS in the system can be addressed by using a particular
register address encoding
yyXyy
, where X represents the UPS ID in the system:
0 - UPS1; 1 - UPS2; 2 - UPS3; 3 - UPS4; 4 - UPS5; 5 - UPS6; 6 - UPS7; 7 - UPS8; 8 - System
The
System
UPS is a virtual UPS that reports alarms, status and measures for the entire UPS system.
For stand-alone 3-ph UPS and for 1-ph UPS only one set is available, that is
0 – UPS 1
.
7.2.2 Data Types
The UPS status and measures are available in different data types. Data types exceeding the 16-bit
register length are implemented by combining subsequent registers.
Bit
Discrete Input registers (UPS alarms and status) are single Bit.
If a specific bit is set the corresponding alarm/status is active.
Byte
8-bit [0..255].
The 8-bit data is always stored in the lower byte of the 16bit Input Register – the higher byte is
Zero padded.
Long
Long Integer, 32 bit [-2
31
.. +(2
31
-1)]
Float
Floating-point, 32 bit: 8-bit exponent [2
[-128 .. +127]
] and 24 bit mantissa [-2
31
.. +(2
31
-1)]
The exponent occupies the uppermost 8 bits (MSB's) and is expressed as 2.complement. The
valid range is [-127;+127]. It is transmitted first.
The mantissa occupies the rightmost 24 bits (LSB's) and is also expressed as 2.complement. The
range is [-2
23
; +(2
23
-1)]. The floating point is implicitly assumed to be placed after the LSB.
Integers value in this range can simply be expressed leaving the exponent=0.
Fractional numbers in the range [-32768..32768] can be expressed multiplying the value by 2
8
and defining the exponent=2
-8
. The resolution is 2
-8
, corresponding to 3.9x10
-3
.
As a special case, a value which is not available returns the value
80'FF'FF'FFh
(exponent = -
128).
MSB
LSB
±
±
exponent mantissa
NOTE
: When floating-point handling is disabled all
Float
type measurement will be handled as
signed long integers [-2
31
.. +(2
31
-1)]. Note that the not available value encoding remains the
same.
anual
SNMP/Web Adapter
The floating-point representation used is non-standard. For clarity, below are some examples on the
actual float values encoding / decoding.
Input Voltage
355.16 V
: exponent –8 (F8h), mantissa 90’290 (01 62 28h), coded value F8 01 63 28
F8 01 63 28
01 63 28 h =
90920mantissa
F8 h =
-8exponent
1
90920 x 2
-8
=
355.16
1
The exponent is expressed as 2-complement
Modifications reserved
Page 52/69
OPM_CNT_SNM_BAS_CRD_1GB_V014.doc
Operating M
