• Process Control System
• Process I/O Systems
• Networking
• Operator Interface
• Display Designer
• Engineering facilities
• Eng
• History
• Article reprints
• Downloads
• Manual (contents)
• Manual (complete)

ABACUS4 Manual

ABACUS 4 BLOCK LANGUAGE

ABACUS 4 - Linux

Intended Audience

Table of Contents

Introduction

Block Parameters

analogue value

• floating point 0, +/- 1e-10 to 1e+10
• integer -32767 to +32767

logical value a/m, s/n,v/n

Logical values represent logical results of measurements or calculations, or status flags.

block reference

• simple inputs 0 to max_block_number
• simple outputs 0 to max_block_number
• switch destination_block_no. switch_type (s/r/+/-)
• cascade destination_block_no. cascade_type

text string

• various permitted string lengths

Common parameters

• tg tag


One to seven character string used to identify groups of blocks which are associated


• sn scan number


Number from 0 to 15. Specifies execution frequency for executable blocks.


• ra analogue result


Floating point number. This is the analogue result of the block.


• am auto/manual flag


Logical flag which may be a (auto/true/1) or m (manual/false/0). This flag enables or disables the block for normal scanned execution.


• vn violation flag


Logical result of block, may be v (violated/true/1) or n (nonviolated/false/0).


• ss single shot flag


This flag may be s (set/true/1) or r (reset/false/0). If it is s and the block is executed the block's am flag will subsequently be reset (to m) thus ensuring that the block executes only once.


• nn b.name


This string parameter may be used to identify the block, and is printed in parenthese () after the block number when printed in engineers mode. The text string is converted to upper case, and may not include spaces. It must also be unique, attempts to enter an existing nn into a block will be regected by eng. A maximum of 29 characters are accomodated. Within eng blocks may be located using th '_' command imedeately followed by a string. All matches will be listed, and connection will be made to the last match. To clear an nn paramter enter nn-

example:


b6301 _FIR
Block 6301 (FIRST-AO) type analogue out
Block 14001 (FIRST-DI) type analogue out
b14001


example: (must be unique)


b14011 nnFIRST-DI
b.name (name in use - block 14001)


• nm name


This string parameter is used by various parts of the system to identify the block. For example the PI-Bus driver inserts card and point identification here for user information.

example:


b5301 b14037;p
Block 14037 () type dig input

Block 14037 () type dig input
nn.b.name
de.descr
nm.name DI card 03 point 05
tg.tag 0 sn.scan 0(0.0 sec) ra.rslta 0.000
am.au/mn m vn.vltn n ss.sshot r
en.A Entpr 0 dd.D Entpr 0
b14037


The Indata (plc) driver uses the information here to identify 'nin' and 'nout' numbers.


• de descriptor


This string parameter is used as a textual description of the block. It is useful in adding short comments which are displayed on the drawing produced by the engdraw program. Text is converted to uppercase.


b14011 deThis is a bit of text to explain that this block is a digital input block
descr =THIS IS A BIT OF TEXT TO EXPLAIN THAT THIS BLOCK IS A DIGITAL INPUT BLOCK

Process I/O

One analogue input block is connected to each PI-Bus analogue input point. The calibrated output is presented on the block ra parameter.


One analogue output block is connected to each PI-Bus analogue output point.


The Digital input block reflects the status of a discrete signal from the process.


The Digital output block asserts the status of its am flag in the form of a discrete digital output.


The Pulse duration Output block is used for plant control via two discrete digital output, producing a timed raise or lower signal, to effect the required control action.


One K48 analogue input/output block is connected to each analogue input point and corresponding analogue output point on a SIOX K48 module. The calibrated output is presented on the block ra parameter.


This block is used to connect one SIOX N45 I/O module to various other ABACUS I/O blocks.

Analogue input block (PI-Bus)

One analogue input block is connected to each PI-Bus analogue input point. The calibrated output is presented on the block ra parameter. The block has the following parameters in addition to the common parameters tg sn ra am vn and ss.

• aa algorithm a



integer_value

• rn amplifier range



integer_value

This parameter sets the gain of the amplifier as follows

• ds data scratchpad



block_number or indirect

• rw raw measured value



floating_point_value

This is the result of the analogue to digital conversion represented as 0.0 to 100.0 (unipolar) or -100.0 to 100.0 (bipolar).

• sp span



floating_point_constant

• ze zero



floating_point_constant

The block result is calculated using the following equation

Analogue output block (PI-Bus)

One analogue output block is connected to each PI-Bus analogue output point. The block has the following parameters in addition to the common parameters tg sn ra am vn and ss.

sp set point floating_point_value

hl high limit floating_point_value

ll low limit floating_point_value

sp set point

The full output range of the analogue output is represented by a setpoint of 0-100. On the first scan after the am is set the sp set point is made equal to the current output.

hl high limit

ll low limit

The sp is constrained between the values of hl and ll , before being further processed.

Digital Input block

The Digital input block reflects the status of a discrete signal from the process. The vn flag will be v when the discrete signal is active, and n otherwise. Although the block has all the common parameters including sn and am it is scanned independently of the normal block scanning process.

Digital Output block

The Digital output block asserts the status of its am flag in the form of a discrete digital output. Depending on the process I/O used the vn flag may indicate the actual status of the digital output. Although the block has all the common parameters including sn, it is scanned independently of the normal block scanning process.

Pulse Duration Output block

The Pulse duration Output block is used for plant control via two discrete digital output, producing a timed raise or lower signal, to effect the required control action. In addition to the common parameters the following exist

sp set point floating_point_value

hl high limit positive floating_point_value

ll low limit positive floating_point_value

sp set point

The required number, and duration of pulses required are cascaded into the sp of this block. Cascading +1.00 produces a raise pulse of one unit duration, -1.00 a lower pulse of one duration.

On the first scan after being set auto the sp is set to zero.

hl high limit

The maximum duration of pulse permitted in either direction is determined by the hl high limit value. If the absolute value of sp is greater than hl then hl pulses will be sent

ll low limit

The minimum duration of pulse permitted in either direction is determined by the ll low limit value. If the sp is less than ll then no pulse will be sent.

ra result a

The value of ra represents the actual pulse length to be sent. If an sp value greater than a predetermined maximum is sent then only that maximum is sent and the value of sp is reduced by that amount. Thus pulses are carried forward from one scan to the next.

The actual execution of the pulse duration depends on hardware. If a SIOX N45 module is to be used, the PDO block must be referenced in the N45 module block, see N45 section.

Indata I/O block

The Indata communications protocol is based on markers. Each of the units on an Indata network has an address in the range 1-31. Markers have a number in the range 1-511 (coresponding to blocks 18001 to 18511). Only one unit on the network should write a marker, but any or all of the units may read that marker. Markers may be either analog or digital.

In the Indata programming language a marker is set using a NOUT entity, and is read using a NIN entity.

If ABACUS reads a marker (set by another unit) it will detect the type (analog or digital), the source address and the value. The value will be put into the ra of the block for analog markers, or the vn for digital markers.Text is inserted in the nm parameter describing the address and type or marker. For example if a digital marker (1) from unit with addres 1 is received

nin d adr 1 mark 1

will be written.

If ABACUS is to output a marker then the nm must have text of the following format.

nout d

or

nout a

The digital value of the am flag or analog value of the ra will be transmitted onto the network.

System blocks used by the Indata driver

• Block 21 abacus address
• Block 22 first uPLC address
• Block 23 last uPLC address
• Block 24 data cycle time
• Block 25 C/I ratio

To avoid scanning the entire address range when only some addresses are present the first and last uPLD addresses are defined in the ra parameters of block 23 and 23 respectively.

The cycle time is defined by setting a value in the ra of block 24. This value is the polling frequency of the driver in milliseconds.

To reduce trafic, the Indata protocol allows polling of changes of signals in addition to taking all values. The C/I ratio is the ratio of the number of Change messages to Init (i.e. all signals) messages.

Multiple Indata Networks

sudo $ABACUS_HOME/indata /dev/cua1 9600 21 18001 &
sudo $ABACUS_HOME/indata /dev/cua2 9600 31 19001 &
sudo $ABACUS_HOME/indata /dev/cua3 9600 41 20001 &

Hardware considerations

One method is to use a RS232 - RS422 (4-wire) converter with the RS422 TX and RX pairs connected together.

    TX+ ----+-----------------+--------------+-----------+
            |                 |              |           |
    TX- ----|--+--------------|--+-----------|--+--------|--+
            |  |              |  |           |  |        |  |
    RX+ ----+  |              |  |           |  |        |  |
               |
    RX- -------+

K48 Analogue input/output block

One K48 analogue input/output block is connected to each analogue input point and corresponding analogue output point on a SIOX K48 module. The calibrated output is presented on the block ra parameter. The block has the following parameters in addition to the common parameters tg sn ra am vn and ss.

ao analogue output block number

ds data source integer constant

mx max value floating_point_constant

ze zero floating_point_constant

The block result is scaled to the range ze to mx corresponding to 0 to 100 % input. This may be 0 - 20mA, 4 - 20mA, or 0 - 10 V depending on the hardware configuration of the K48 module.

The K48 module should be configured to have 4 addresses, (4A/I and 4A/O). The value of ds defines both the SIOX port used and the point address. ds should be 10000 + 100*port number + SIOX point address. i.e. for port number 2, SIOX point address 5 ds should be 10205.

The ao parameter should be connected to an analogue output block, whose result will be sent to the corresponding point address.

N45 Module control block

The block has the following parameters in addition to the common parameters tg sn ra am vn and ss.

ds	data source	integer constant
oa	output a terminal 1	block_number
ob	output b terminal 2	block_number
oc	output c terminal 3	block_number
od	output d terminal 4	block_number
oe	output e terminal 5	block_number
ia	input a terminal 10	block_number
ib	input b terminal 11	block_number
ic	input c terminal 12	block_number
id	input d terminal 13	block_number
ie	input e terminal 14	block_number
if	input f terminal 15	block_number
ig	input g terminal 16	block_number
ih	input h terminal 17	block_number
ii	input i terminal 18	block_number
ij	input j terminal 19	block_number
ik	input k terminal 20	block_number
il	input l terminal 21	block_number
im	input m terminal 22	block_number
in	input n terminal 23	block_number

The value of ds defines both the SIOX port used and the point address. ds should be 10000 + 100*port number + SIOX point address. i.e. for port number 2, SIOX point address 5 ds should be 10205.

The SIOX N45 I/O module is a general purpose digital I/O module with 7 sourcing output and 14 inputs, all 10V - 35V DC. The outputs can supply 0.5A each and have short circuit protection. Through the built-in CPU and a simple PLC programming language, curtain options are available.

If the PLC programme has been correctly loaded and configured, a number of the outputs may serve as pulse duration outputs. The corresponding output parameter (oa - og ) should be connected to a pdo block. Enter the pdo block as a positive number to connect the raise side, and as a negative number for the lower side. Thus it is possible to connect raise and lower in any combination. The pdo function, if programmed into to N45 module excludes ordinary digital output operation for those specified output points. It is possible to configure any combination of digital outputs and pdo outputs.

Logical Blocks

Checklist block	This block type functions as an AND or OR gate for up to five logical inputs, any or all of which may be inverted.
Switch block	The switch block derives a single logical result (vn), and can switch auto or manual two blocks, depending on the status of a single logical input. Two list processing algorithms allow ranges of blocks to be switched.
Pattern block	The pattern block is used to switch up to five blocks auto or manual.

Checklist block

The checklist block derives a single logical result (vn), and can switch auto or manual one block, depending on the status of five logical inputs. The block has the following parameters in addition to the common parameters tg sn ra am vn and ss.

ia	logical input	block_number or indirect_block_number
aa	algorithm a	0 - 7
ib	logical input	block_number or indirect_block_number
ab	algorithm b	0 - 1
ic	logical input	block_number or indirect_block_number
ac	algorithm c	0 - 1
id	logical input	block_number or indirect_block_number
ad	algorithm d	0 - 1
ie	logical input	block_number or indirect_block_number
ae	algorithm e	0 - 1
af	algorithm f	0 - 1
sv	switch if violated	block_number s/r/+/- or indirect_block_number s/r/+/-

The logical result of the five blocks specified by each of the logical inputs in turn (ia,ib ,ic,id,ie) is processed by its coresponding algorithm ( aa,ab,ac, ad,ae) to generate five intermediate logical results in accordance with table ck1. (Note that only aa can have a value greater than one.)

Unconnected inputs are ignored.

af = 0 AND operation

The logical result of the block will be v (true) only if all of the five intermediate logical results are true (1).

af = 1 OR operation

The logical result of the block will be v (true) if any of the five intermediate logical results is true (1).

Table ck1

algoritm	description	logical previous	logical current	intermediate result
0	off (inverting)	-	0	1
	(aa,ab,ac, ad,ae)	-	1	0
1	on (noninverting)	-	1	1
	(aa,ab,ac, ad,ae)	-	0	0
2	on-going	0	1	1
	aa only	1	1	0
		1	0	0
		0	0	0
3	off-going	1	0	1
	aa only	0	0	0
		0	1	0
		1	1	0
4	any-change	0	1	1
	aa only	1	0	1
		0	0	0
		1	1	0
5	no-change	0	0	1
	aa only	1	1	1
		0	1	0
		1	0	0

The block addressed by the sv parameter will be affected in accordance with table ck2.

Table ck2

intermediate result	switch type s	switch type r	switch type +	switch type -
	target block becomes
1	a	m	a	m
0	-	-	m	a

Switch block

The switch block derives a single logical result (vn), and can switch auto or manual two blocks, depending on the status of a single logical input. The block has the following parameters in addition to the common parameters tg sn ra am vn and ss.

ia	logical input	block_number or indirect_block_number
aa	algorithm a	0 - 7
sa	switch a	block_number s/r/+/- or indirect_block_number s/r/+/-

The status of the logical input specified by the input a ia parameter is processed by algorithm a aa and the target block's am auto/manual flag is changed in accordance with table sw1 .

The logical result of the block is set to the intermediate logical result derived by aa algorithm a.

ab	algorithm b	0 - 7
sb	switch b	block_number s/r/+/- or indirect_block_number s/r/+/-

The status of the logical input specified by the input a ia parameter is processed by algorithm b ab and the target block's am auto/manual flag is changed in accordance with table sw1 .

List processing

These algorithms permit the execution of the block's algorithm over a contiguous list of blocks instead of individual blocks. The target list is specified by sa and sb . The first output block is specified by sa, the last by sb . The input list is of exactly the same length as the output list, thus it is only necessary to specify the first input block with ia. Indirection is allowed on ia, sa and sb. The direction of processing is always from the first to the last, and cannot be reversed; thus the block specified as the first output block must have a lower or equal block number as that specified as the last. If ia is zero then an imaginary input list of blocks all having a 0 (nonviolated,reset, false) logical result is used.

The only restriction to the size of the lists is that they must both be accommodated within the constraint of the system.

aa = ab = 6 List processing (inverting)

The switch type of both sa and sb must be the same. For each block an intermediate result is calculated and the corresponding target block is modified as if the algorithm were 0 (inverting).

aa = ab = 7 List processing (noninverting)

This algorithm functions in the same list processing way as above, except that for each block an intermediate result is calculated and the corresponding target block is modified as if the algorithm were 1 (noninverting).

Table sw1

algoritm	description	logical previous	logical current	intermediate result	switch type s	switch type r	switch type +	switch type -
0	off (inverting)	-	0	1	a	m	a	m
		-	1	0	-	-	m	a
1	on (noninverting)	-	1	1	a	m	a	m
		-	0	0	-	-	m	a
2	on-going	0	1	1	a	m	a	m
		1	1	0	-	-	m	a
		1	0	0	-	-	m	a
		0	0	0	-	-	m	a
3	off-going	1	0	1	a	m	a	m
		0	0	0	-	-	m	a
		0	1	0	-	-	m	a
		1	1	0	-	-	m	a
4	any-change	0	1	1	a	m	a	m
		1	0	1	a	m	a	m
		0	0	0	-	-	m	a
		1	1	0	-	-	m	a
5	no-change	0	0	1	a	m	a	m
		1	1	1	a	m	a	m
		0	1	0	-	-	m	a
		1	0	0	-	-	m	a
6	list processing (inverting) see text
7	list processing (noninverting) see text

Pattern Block

The pattern block is used to switch up to five blocks auto or manual. The block has the following parameters in addition to the common parameters tg sn ra am vn and ss.

sa	switch a	block_number s/r or indirect_block_number s/r
sb	switch b	block_number s/r or indirect_block_number s/r
sc	switch c	block_number s/r or indirect_block_number s/r
sd	switch d	block_number s/r or indirect_block_number s/r
se	switch e	block_number s/r or indirect_block_number s/r

Each time the block is executed the blocks specified by switches sa to se are switched auto or manual depending on the specified switch type s or r respectively.

Note unless continual and repeated operation is desired it is usual to the set the single shot flag ss to s.

Process Control Blocks

The p.i.d. block is designed to perform P, PI, or PID control functions, with standard control constants, executing a variety of alternative control algorithms, and calculating control performance data at the same time.

The motor block is designed to allow comprehensive control of the starting/stopping of electric motors, pumps, and opening/closing of valves. The connections are made through digital inputs and outputs. Facilities are included to initiate alarm signals if the motor/pump fails to respond to control signals in allowed time periods. The accumulation of run time is also included.

PID Controller block

The p.i.d. block is designed to perform P, PI, or PID control functions, with standard control constants, executing a variety of alternative control algorithms, and calculating control performance data at the same time. In addition to the common parameters the p.i.d. block has the following

sh	setpoint hl	floating_point_value
sl	setpoint ll	floating_point_value
mh	measured variable hl	floating_point_value
ml	measured variable ll	floating_point_value
ir	input run	block_number
ip	input proc	block_number
sr	future use	floating_point_value
ar	future use	integer
db	dead band	floating_point_value
gf	goodness factor	floating_point_value
ag	goodness factor algorithm	integer
kp	proportional constant	floating_point_value
ti	integral time	floating_point_value
td	derivative time	floating_point_value
ap	proportional algorithm	integer
ai	integral algorithm	integer
ad	derivative algorithm	integer
an	answer	floating_point_value
op	output block	block_number

The am flag of a pid block may be considered as an îauto-requestî flag. When the am flag is auto the loop may be regarded as requested auto. The vn flag may be regarded as that flag which indicates that the loop is in fact operating in automatic.

When conditions are right for the controller to start operating in automatic the vn falg is set, and previous stored values of er are set to zero, and previous values of mv are set to the current mv value. If the se flag is set ( e equalise) then the setpoint value is made equal to the measured variable. This is done to ensure bumpless transfer.

ir input run

If the vn violation status of any block connected here is v then the normal processing of the block may proceed. The vn of this pid block is made the same as that of the pointed to block if this pid block is auto. If this pid block is inactivated because of this connection it will be initialised when reactivated. This allows interlocking of control loops.

ip input process

If the vn violation status of any block connected here is v then the normal processing of the block may proceed. No re-initialisation is performed nor is the vn status changed as a consequence of this connection. This parameter is intended for use with the ai=1 dead time controller option.

ia input a

The measured variable (mv) in the following is derived from the analogue result ra of the block specified here.

mh input hl, ml input ll

These limits are used to constrain the value of the measured variable (mv).

is input setpoint, sp setpoint

If the is value is that of a valid block number, then the sp value is obtained from the analogue result ra of that block. The sp value is the same as the ra of this block.

sh setpoint hl, sl setpoint ll

These limits are used to constrain the value of sp. Upon execution of this block the ra of the block pointed to by is will also be constrained, That is to say the constrained sp is fed back to the block pointed to by is on each scan.

op output block, an answer

an answer is the result of the control equation below which is then added to the op output block. If the op output block is an analogue output, pulse duration output or a control block the value of an is added to the sp of that block, other wise it is added to its ra.

Note that because of the fact that an incremental value is added to the op output block it is possible to alter manually, or with other blocks, that value.

Control Equation Normal operation (ai=0)

er = sp - mv

er_1 = er from previous scan

er_2 = er from two scans previous

mv_1 = mv from previous scan

mv_2 = mv from two scans previous.

an = kp * ( prop_value + int_value + deriv_value )

where prop_value, int_value and deriv_value are internal temporary variables depending on the algorithms and constants as follows

td derivative time, ad derivative algorithm, deriv_value

if ad = 0 then

deriv_value =(mv - 2*mv_1 + mv_2) * td / scan_time_in_minutes

if ad = 1 then

deriv_value =(er - 2*er_1 + er_2) * td / scan_time_in_minutes

if ad = 3 then

er = error squared, (note if error is -ve, er is also made -ve)

deriv_value =(er - 2*er_1 + er_2) * td / scan_time_in_minutes

Derivative time (in minutes) can be set to zero to disable derivative action. The derivative action may be configured to operate on the measured variable (ad=0) or the error ( ad=1). Even error squared control is available with (ad=3).

kp proportional constant, ap proportional algorithm, prop_value

if ap = 0 then prop_value = er - er_1

if ap = 1 then prop_value = mv - mv_1

The proportional action may be configured to operate on the error ( ap=0) or the measured variable ( ap=1)

ti integral time, ai integral algorithm, int_value

if ti is not equal to 0 then int_value = er * scan_time_in_minutes / ti

To disable integral action simply set ti to 0, otherwise ti represents integral time in minutes. ( ai integral algorithm is reserved for future use.)

Control Equation dead time controller ai=1

• an = kp * er - ti * er_1 - td * an_1

• er = sp - mv
• er_1 = er from previous execution
• an_1 = an from previous execution
• sr
• ar



These parameters are reserved for future use.

• db dead band (time since last setpoint change in minutes)
• gf goodness factor
• ag goodness factor algorithm

if ag = 0 gf = gf + abs( er ) absolute error

if ag = 1 gf = gf + er * er error squared

if ag = 2 gf = gf + abs( er ) * db absolute error * time

if ag = 3 gf = gf + er * er * db error squared * time

After a step change in demand, one can look at the rising value of gf and compare with other tuning constants.

Motor Block

The motor block is designed to allow comprehensive control of the starting/stopping of electric motors, pumps, and opening/closing of valves. The connections are made through digital inputs and outputs. Facilities are included to initiate alarm signals if the motor/pump fails to respond to control signals in allowed time periods. The accumulation of run time is also included.

Each of the logical inputs and outputs has associated with it an algorithm for inverting the significance of that input or output. In each case this algorithm follows the normal ABACUS convention where 1=noninverting, and 0=inverting.

In addition to the common parameters the motor block has the following

ir	input ready	block_number
ar	algorithm	(0=inverting, 1=noninverting)
iu	input run	block_number
au	algorithm	(0=inverting, 1=noninverting)
it	input start	block_number
at	algorithm	(0=inverting, 1=noninverting)
ip	input stop	block_number
ap	algorithm	(0=inverting, 1=noninverting)
ia	input auto	block_number
aa	algorithm	(0=inverting, 1=noninverting)
c1	control input 1	block_number
a1	algorithm	(0=inverting, 1=noninverting)
c2	control input 2	block_number
a2	algorithm	(0=inverting, 1=noninverting)
af	algorithm	(0=AND, 1=OR)
sr	start output	block_number
nr	algorithm	(0=inverting,1=noninverting)
sp	stop output	block_number
np	algorithm	(0=inverting, 1=noninverting)
sa	alarm output	block_number
na	algorithm	(0=inverting, 1=noninverting)
t1	time limit	floating_value
d1	scratchpad for above	block_number
t2	time counter	floating_value
d2	scratchpad for	above block_number
t3	run time min.sec	floating_value
d3	scratchpad for above	block_number
t4	run time	hours floating_value
d4	scratchpad for above	block_number

Motor block flow diagram

Calculation Timing and type conversion

Index Block	This block type is used for counting, timing and integer arithmetic.
Input code conversion block	The input conde conversion block is used to reference a continuous range of logical block results and produce an analogue result depending upon their status.
Output code conversion block	The output conde conversion block is used to convert an analogue result into a bit pattern and switch a continuous range of blocks.
Control Block	The control block is a multifunction block capable of many different types of calculation, control functions, accumulation functions, and list processing.
Sequence Block	This block causes the execution of curtain other blocks independently of their scan parameters in a sequence determined by an input list to the sequence block. Alternatively a serial programme can be executed.
Scratchpad Block	All blocks which have no other block type are classified as scratchpad blocks. Scratchpad blocks are nonexecutable, and only have the common parameters. They are used for analogue and data storage only.

Index Block

The index block is used for counting, timing and integer arithmetic. The block has the following parameters in addition to the common parameters tg sn ra am vn and ss.

ia	logical input	block_number or indirect_block_number
aa	algorithm a	0 - 7
ib	analogue input	block_number or indirect_block_number
ab	algorithm b	0 - 7
hl	high limit	floating_point_constant
ll	low limit	floating_point_constant
k1	constant	floating_point_constant

ia input a, aa algorithm a

The status of the logical input specified by the ia parameter is processed by aa in accordance with table IX1. If the intermediate logical result is 1 then further processing is carried out as described below, otherwise processing is discontinued, except for the initialisation described for ab of 1.

ib input b

The analogue result of the block specified by this parameter is used as the measured variable in further processing of this block.

hl high limit, ll low limit, k1 constant

The value of these parameters are used in the various calculations performed in accordance with ab.

Table IX1

ab = 0 Integer addition

ra = measured variable + k1

ab = 1 cyclic progression

If ib is nonzero ra = measured variable

If ib is zero and k1 = 0 ra = ll

If ib is zero and k1 < 0 ra = hl

if ra+k1 hl then

ra = ll, and vn = v

else

ra = ra + k1 , and vn = n

If k1 < 0 then

if ra+k1 < ll then

ra = hl, and vn = v

else

ra = ra + k1 , and vn = n

In this way the value of ra steps from one limit to the other with steps of k1. When arriving at the limit the violation flag is set for one scan.

ab = 2 Integer subtraction

ra = k1 - measured variable

If ( ra < ll ) or ( ra hl ) vn = v

else vn = n

ab = 3 Integer multiplication

ra = k1 * measured variable

If ( ra < ll ) or ( ra hl ) vn = v

else vn = n

ab = 4 Integer division

ra = measured variable / k1

If ( ra < ll ) or ( ra hl ) vn = v

else vn = n

ab = 5 Integer division

ra = k1 / measured variable

If ( ra < ll ) or ( ra hl ) vn = v

else vn = n

ab = 6 Delay Timer

Input Code Conversion Block

The input code conversion block is used to reference a continuous range of logical block results and produce an analogue result depending upon their status.

In addition to the common parameters the input conde conversion block has the following

aa = 0 binary transfer

aa = 2 displacement

Output Code Conversion Block

The output conde conversion block is used to convert an analogue result into a bit pattern and switch a continuous range of blocks.

In addition to the common parameters the output conde conversion block has the following

This algorithm transfers the binary pattern of the measured variable to the am flags of the specified range of blocks. The block specified by sw is regarded as the least significant bit in the word, and is represented by 1 in the measured variable, the next block by 2 then by 4 etc. Should a bit be set in the measured variable then the corresponding output block will be set auto, otherwise it will be set manual.

Control Block

The control block is a multifunction block capable of many different types of calculation, control functions, accumulation functions, and list processing. The ab algorithm b parameter specifies the manner in which the control block is processed. The algorithms are divided into three main categories.

0 to	9	Calculation and control
10 to	13	List calculation
14 to	19	Accumulation
20 to	25	List processing

The block has the following parameters in addition to the common parameters tg sn ra am vn and ss.

ia	analogue input	block_number or indirect_block_number
aa	algorithm a	0 - 3
ab	algorithm b	0 - 32
sp	setpoint	floating_point_value
is	inout setpoint	block_number or indirect_block_number
in	increment	floating_point_value
er	error	floating_point_value
eq	equalisation flag	e/n
ds	data scratchpad	block_number or indirect_block_number
hl	high limit	floating_point_value
ll	low limit	floating_point_value
k1	constant k1	floating_point_value
k2	constant k2	floating_point_value
ca	cascade a	block_number cascade_type or indirect_block_number cascade_type
cb	cascade b	block_number cascade_type or indirect_block_number cascade_type

Calculation and control algorithms ab 0 - 9

ab	description	calculation(s) performed
0	simple transfer	result = sp + in - mv (mv = ra of ia block)
1	simple ratio	result = (sp + in - mv) * k1
2	flow rate conditioning	result = k2 * sqrt( (mv - k1)*(sp + hl)/(in + ll) )
3	general mathamatics	result = k2 * (mv - k1) * (sp + hl)/(in + ll)
4,5,6	two term control	error = sp + in - mv (error is an internal variable)
		result* = k1 * ( error - k2 * er )
		er = error
*Note
4	er is used for aa calculation
5	ra is used for aa calculation
6	sp is used for aa calculation
7	natural logarithm	result = k2 * ln( (mv - k1)*(sp + hl)/(in + ll) )
8	exponential	result = k2 * exp( (mv - k1)*(sp + hl)/(in + ll) )

List Calculation algorithms ab 10 - 13

These algorithms operate on the contiguous list of block defined as those blocks between is and ia. ia should refer to a block with higher block number than that pointed to by is.

ab	description	calculation performed
10	sum	result = k2 + k1 * sum ( is .. ia )
11	ave	result = k2 + k1 * ave ( is .. ia )
12	max	result = k2 + k1 * max ( is .. ia )
13	min	result = k2 + k1 * min ( is .. ia )

Accumulation algorithms ab 14 - 19

Algorithms are either pre-initialised or post-initialised.

Pre-initialisation involves resetting the accumulations and setting the result to zero the first time the block is processed after having been set auto.

Post-initialisation processes the accumulations to obtain the result prior to resetting the accumulations, on the first scan after the block has been set auto. This having the advantage that the block may be left manual whilst accumulating, it only being necessary to set it auto to obtain the result, and to reset the accumulations for the next period. (It is normal to set the single shot flag ss in this case.)

ab	description	calculation(s) performed
14	integration pre-initialised	result = acc / ( k1 * 100 )
15	integration post-initialised	result = acc / ( k1 * 100 )
16	average pre-initialised	result = acc / er (er = number of samples)
17	average post-initialised	result = acc / er (er = number of samples)
18	std. deviation pre-initialised	result = k1 * standard deviation (-1 if less than 10 samples)
19	std. deviation post-initialised	result = k1 * standard deviation (-1 if less than 10 samples)

List processing algorithms ab 20 - 25

ia

ca	cascade a	block_number cascade_type or indirect_block_number cascade_type
cb	cascade b	block_number cascade_type or indirect_block_number cascade_type

The ca parameter specifies the first block to receive a cascade from the list processing algorithm. The cb parameter specifies the last block to receive a cascade. The number of target blocks is thus defined by the cascades, and therefore the number of input blocks is the same. cb must point to a higher block number than ca.

The cascade type must be the same for both ca and cb, and must be consistent with the type of target block.

When the algorithm is processed the measured variable is taken from each input block in turn. The measured variable is processed and the result cascaded to the corresponding target block.

For each item of data thus processed, limits are checked to produce an intermediate violation status, and result which are taken into account by the cascades.

ab	description	calculation(s) performed
20	simple transfer	result = mv
21	multiplication	result = (sp - mv) * k1
22	division	result = (sp - mv) / k1
23	division	result = k1 / (sp - mv)
24	multiply input	result = mv * A / k1 A = initial value of by output result to be cascaded to
25	simple transfer	result = mv mv = result of block whoseindirect input number is the result of the current input. If this number is invalid no cascade is performed at this stage.

aa Logical algorithm

The analogue result of the block calculated by the various algorithms ( ab) is further processed by aa. (In the case where ab = 4 it is the value of er, where ab = 6 it is the value of sp) The list processing algorithms have the result processed by aa at each step.

aa	ra hl		ll <= ra <= hl		ll ra
	ra	vn	ra	vn	ra	vn
0	result	n	result	n	result	n
1	result	v	result	n	result	v
2	hl	v	result	n	ll	v
3	result	v	0	n	result	v

Cascades

• sp - Setpoint add The recipient block may be of any type, but if it is a control block, pulse duration output block, or an analogue output block it is the sp of that block which is affected, otherwise the analogue result ra which is affected. (In the case of the PID block the ra and sp parameters are equivalent). The affected parameter has the result of this block added to it.
• so - Setpoint overwrite The recipient block may be of any type, but if it is a control block, pulse duration output block, or an analogue output block it is the sp of that block which is affected, otherwise the analogue result ra which is affected. (In the case of the PID block the ra and sp parameters are equivalent). The affected parameter has the analogue result of this block written to it, overwriting its previous value.
• k1 - k1 overwrite The recipient block must be a control block whose k1 parameter is overwritten by this blocks analogue result.
• k2 - k2 overwrite The recipient block must be a control block whose k2 parameter is overwritten by this blocks analogue result.
• hl - hl overwrite The recipient block must be a control block whose hl parameter is overwritten by this blocks analogue result.
• ll - ll overwrite The recipient block must be a control block whose ll parameter is overwritten by this blocks analogue result.
• in - in add The recipient block must be a control block whose in parameter is added onto by this blocks analogue result.
• dp - double precision add The recipient block must be a control block whose accumulator parameter is added into by this blocks analogue result. If the recipient block has ab of 16 or 17 (averaging) the er parameter is also incremented by one each time.
• sv - Switch if violated The recipient block will by switched on (its am made auto) if this block is in violation.
• sn - Switch if not violated The recipient block will by switched on (its am made auto) if this block is not in violation.
• rv - Reset if violated The recipient block will by switched off (its am made manual) if this block is in violation.
• rn - Reset if not violated The recipient block will by switched off (its am made manual) if this block is not in violation.

Sequence Block

Scratchpad Block

All blocks which have no other block type are classified as scratchpad blocks. Scratchpad blocks are nonexecutable, and only have the common parameters. They are used for analogue and data storage only.

Operator Communications

The operator communicates with ABACUS4 via an operators station. This is usually a PC running the Linux operating system and the X window system. This may be the same system that is running ABACUS4 or another reachable over a network.

Text based operator interfaces may be constructed, but these are not documented here.

All normal operator access to the ABACUS4 block database is done via the Alarm Code Block, although all blocks are accesable to the Tck/tk language.

The Alarm Code Block permits the monitoring of the analogue or logical results of blocks, and the generation of process alarms when limits are violated, or upon changes.

The block permits the operator access to its high and low process limits, and to the result, setpoint, auto/manual and auto/manual request flags of the loop being monitored.

Alarm Code Block

The Alarm Code Block permits the monitoring of the analogue or logical results of blocks, and the generation of process alarms when limits are violated, or upon changes.

The block permits the operator access to its high and low process limits, and to the result, setpoint, auto/manual and auto/manual request of the loop being monitored.

In addition to the common parameters the Alarm Code Block has the following

si	signal name text	text_string (24 bytes)
ia	input a	block_number
is	input setpoint	block_number
ar	auto/man request	block_number
st	status	block_number
hl	high limit	floating_point_value
ll	low limit	floating_point_value
xh	high high limit	floating_point_value
xl	low low limit	floating_point_value
aa	algorithm	integer
hy	hysteresis value	floating_point_value
un	units text	text_string (9 bytes)
mo	ok text	text_string (9 bytes)
ml	high alarm text	text_string (9 bytes)
mh	low alarm text	text_string (9 bytes)
lx	low low alarm text	text_string (9 bytes)
hx	high high alarm text	text_string (9 bytes)
ma	auto text	text_string (9 bytes)
mm	hand text	text_string (9 bytes)
ur	auto request text	text_string (9 bytes)
hr	hand request text	text_string (9 bytes)
s1	status 1 (alarm)	text text_string (9 bytes) (read only)
s2	status 2 (hand/auto)	text text_string (9 bytes) (read only)
fn	file name text	text_string (9 bytes)

This is the loop name which is used in all ABACUS journaling. (It should be copied over the the descriptor of any Enterprise Server database points - see later.)

ia input a

The ra of the block referenced here is taken as the measured variable of the loop. Subject to the value of aa it is this block's result which is tested against the various limits.

is input setpoint

The ra of the block referenced here is taken as the setpoint of the loop.

ar auto/man request

The vn of the block referenced here is taken as the auto/man request flag of the loop. That is to say the flag which is set when the operator requests that the loop go into auto.

st status

The vn of the block referenced here is taken as the auto/man status flag of the loop. That is to say the flag which is set when the loop is in auto.

hl high limit

ll low limit

xh high high limit

xl low low limit

hy hysteresis value

Each time the measured variable passes one of the above limit values a message is recorded in the journal file, excepting that if the hy value is nonzero a return excursion is ignored until a movement greater than hy occurs. This can be used to avoid multiple alarms when a noisy signal goes into alarm.

For correct functioning of the limit test facilities the following should be true

xh > hl > ll > xl, and hy should be positive and less than the difference between any two of the other limits.

s1 status 1 (alarm) text

The text string written by the block processor here depends on the relationship between the measured value and the alarm limits. If hy = 0 the following is carried out

if ( measured value < xl ) then s1 = lx (low low alarm text)

else if ( measured value < ll ) then s1 = ml (low alarm text)

else if ( measured value > hl ) then s1 = mh (high alarm text)

else if ( measured value > xh ) then s1 = hx (high high alarm text)

else s1 = mo (ok text)

If however the hy value is nonzero the above becomes

if ( measured value < xl ) then s1 = lx (low low alarm text)

else if ( (xl + hy ) > measured value >= xl ) then s1 is unchanged

else if ( measured value < ll ) then s1 = ml (low alarm text)

else if ( (ll + hy ) > measured value >= ll ) then s1 is unchanged

else if ( measured value > hl ) then s1 = mh (high alarm text)

else if ( (hl - hy ) < measured value <= hl ) then s1 is unchanged

else if ( measured value > xh ) then s1 = hx (high high alarm text)

else if ( (xh - hy ) < measured value <= xh ) then s1 is unchanged

else s1 = mo (ok text)

aa algorithm

This algorithm determines which limits are used to define an alarm state and set the vn of the Alarm Code Block or if the st logical value should be used. The following tables shows the effects of alternative aa values. (aa values greater than 9 inhibit the printing/logging of setpoint changes)

Analogue alarms

aa	mv<xl	mv < ll	mv>ll&mv<hl	mv>hl	mv>xh
0,10	vn = n	n	n	n	n
1,11	vn = v	v	n	v	v
4,14	vn = v	n	n	n	v

Digital alarms

aa	st -> vn = n	st -> vn = v
2,12	vn = n	v
3,13	vn = v	n

s2 status 2 (hand/auto) text

The text string written by the block processor here depends on the vn of the block referred to in st. If this block's vn = v then ma is copied, otherwise mm.

All changes to limit values, setpoint value, auto/manual request flag, and status changes are logged in the journal file if the Alarm Code Block is auto. One line is written for each event, and has the following format

YYMMDD HH:MM:SS signal_name message ( old_value ) new_value

where

signal_name is the si of the Alarm Code Block

YYMMDD HH:MM:SS is the time and date of the event

message = "setpoint changed"

"hi-hi lim changed"

" high lim changed"

" low lim changed"

or "lo-lo lim changed" as appropriate.

When an alarm limit is exceeded the format is as follows

YYMMDD HH:MM:SS signal_name alarm_text value units

where alarm_text is the appropriate alarm text string ( ml mh lx hx or ok ), value is the measured value, and units is the text un.

The ABACUS4 - Tcl/tk interface

The easiest way to get to see a list of the commands is to run the xman program, and look into the n section.

Displays are stored in a subdirectory of /home/abacus. This directory has the name of the application, and also contains the datadump.bin file(s). If the project is called smith for example then these files will reside in the directory /home/abacus/smith. It is usual to have one 'display manager' file which includes code for menus and a frame for the displays to be shown in, and a number of 'display files'. In addition the background diagrams and dynamic symbols are stored in gif format.

ABACUS4 programs which Tcl/tk programs may use

eng2

eng2 b1234,p

Here block 1234 is printed, something like the following being returned:

Block 1234 () type index 

B1234 () 
nn 
de 
nm 
tg spare;sn 0(0.0 sec);ra 0.000 
am m ;vn n ;ss r 
en 0;dd 0 
aa 0;ia 0( 0.000 m-rn) 
ab 0;ib 0( 0.000 m-rn) 
k1 0;hl 0;ll 0 

eng2 requires that all the commands come together in its first argument, so that if spaces are part of the syntax, then the argument should be enclosed in quote marks. If you use this program from an xterm window for example beware of the command line conversions that your shell does, in particular the * character is translated into all the filename of files in the current area, before being passed onto, in this case, the eng2 program. This can be avoided either by enclosing you arguments in quotes, or by prefixing the * with the \ character.

eng2 is sometimes used with a unix pipe. For example if we ish to list all the si paramters of the blocks in the range 8001 to 8099 we could use

eng2 b8001,xb8099,.si

This would produce something like

Block 8001 () type alarm code 
Xblock 8099 type alarm code 
b 8001;si 
b 8002;si# 02 GAUGE ON/OFF SHEET 
b 8003;si# 03 GAUGE STATIC 
b 8004;si# 04 GAUGE STANDARDIZE 
b 8005;si
...

eng2 b8001,xb8099,.si | sed -e "s/^b //" -e "s/;si/ /" 

Block 8001 () type alarm code 

Xblock 8099 type alarm code 
8001 
8002 # 02 GAUGE ON/OFF SHEET 
8003 # 03 GAUGE STATIC 
8004 # 04 GAUGE STANDARDIZE 
8005 
8006 

eng2 b8001,xb8099,.si | sed -e "s/^b //" -e "s/;si/ /" | \ 
grep -v "^B\|^X" 

eng4

ra

frank(abacus):~$ ra 14
0.250931frank(abacus):~$
frank(abacus):~$

ra2

frank(abacus):~$ ra2 14 51 52 53
0.15332 0.1 0.2 0.58
frank(abacus):~$

ra3

vn

vn2

vn3

am

si

si3

sp3

s13

s23

un3

ABACUS 4

System blocks, Configuration etc.

Data dumping is discussed, as is logging in and out of the system.

ABACUS4 processes are listed with a short description of their function.

Starting and Stopping ABACUS4

To start ABACUS manually, log in as abacus (initial password abacus) and enter the command abacus. You will be asked if you wish to stop and restart ABACUS, pressing enter with the yes field selected will first stop any running ABACUS processes, then after a short pause startup the ABACUS system.

To stop ABACUS manually, log in as abacus (initial password abacus) and enter the command kill.abacus.

On systems configured with the PI-Bus driver, and possibly others, you may be asked for your password before ABACUS will start. This is because the processes which access I/O ports on the PC must be run with root privileges.

Installing ABACUS4

README.1ST

This is the README.1ST file for abacus-4.9807 this first public version. 


Hello everyone, this is just a bit of information about ABACUS.
ABACUS runs under Linux, the free Un*x operating system found on a large number
of Internet servers around the world. 

The minimal hardware requirements are an IBM compatible 386 machine or better
with at least 4 Mb RAM (preferably at least 8 Mb). A hard-disk with 30Mb has
been used although >200Mb is required for the display system (X). On a network
(Ethernet or even serial) a disk-less system should be possible, although this
cannot be considered a secure arrangement.
 
To install ABACUS 4 on your Intel based Linux machine you will need to
consider the following.

You need a user called abacus with his/her home at /home/abacus. He/She must have sudo
rights (I put 'abacus ALL=ALL' at the end of /etc/sudoers, but you might want
to work out something more sophisticated, let me know what you find works)

As root you can install the package from the / directory with

	cd /
	tar -xzvf where_ever_it_is/abacus-4.version.tar.gz
	install/doinst.sh
	
Or (if you are using Slackware) you can use the installpkg

	installpkg where_ever_it_is/abacus-4.version.tar.gz



The ABACUS 4 manuals are in html format in the abacusmanual/ directory.

To start the system log in as abacus, and start the command abacus. The
file abacus is a bash shell, and is where the main configuration of
which parts of abacus are started (i.e. which process I/O drivers, and
which block scan processors)

ABACUS 4 expects '.' to be in the PATH environment variable, other wise it
won't find everything.

Process I/O
===========

Drivers exist is this version for pibus, indata and das08. The source for the
das08 and a number of h files are stored in the src directory.

It is expected that new drivers will appear in the not too distant future for
MODBUS and COMLI protocols. These may however require modifications to the
database structure (i.e. new block types).

DISPLAYS
========

The displays are made using Tcl/tk. I've used version 8.0, some however will
work with earlier versions. The tk_openfile type widgets have been used in the
trend page, and 8.0 is required for that.

LICENSING
=========

ABACUS 4 is a commercial product and a license is required to run it. However
it will run for an evaluation period of 24 hours before shutting down
without a valid /home/abacus/abacus.computer.id.dat file.

At the time of writing the pricing has not been fixed, so if you want to buy
a license mail to abacusabacus@hotmail.com

STARTUP
=======


Here is a log of how it went when I started abacus

	frank(abacus):~$ abacus
	starting kill of abacus
	kill 8461: No such process
	kill_smem:mem_id 0, shmctl returned 0
	Starting abacus


	Warning - id_check failed -1
	The identity of your harddisk does not match that expected
	This ABACUS system will run in demonstration only
	The file /home/abacus/abacus.computer.id.dat contains a coded
	version of the identity of the disk for which this software is
	licensed. Please send the following information to the supplier
	of your ABACUS system, and arrange for a new 
	/home/abacus/abacus.computer.id.dat file to be sent to you.

	MODEL="WDC AC31600H"
	FW_REV="23.16U23WDC AC31600H"
	SERIAL_NO="WD-WT2893535299"


	This information has been saved in /home/abacus/hard.disk.id.txt

	Abacus will start shortly in demonstration mode.
	It will require restarting after a period of 24 hours


	key = 0
	shmget(key, TOTAL_SPACE  , IPC_PRIVATE | IPC_CREAT ) errno = 0
	mem_id = 134

	Loading from datadump file ...Database load done

	          system start      1, stop     20
	        pibus ai start    101, stop    420
	        i/p code start    801, stop    810
	           index start   1001, stop   1199
	       checklist start   2001, stop   2199
	          switch start   2501, stop   2699
	         pattern start   4001, stop   4199
	       sequencer start   9001, stop   9050
 	        control start   5001, stop   5999
	          p.d.o. start   6001, stop   6080
	    analogue out start   6301, stop   6364
	        o/p code start   6501, stop   6550
	      alarm code start   8000, stop   8999
	           motor start   9501, stop   9699
	       dig input start  14001, stop  14320
	      dig output start  14501, stop  14820
	            disc start  17501, stop  17510
	       data link start  17901, stop  17910
	    scan control start     51, stop     65
	          p.i.d. start   7001, stop   7299
	      indata I/O start  18001, stop  18511
	at: pluralization is wrong
	Job 68 will be executed using /bin/sh
	Subroutine start   7901, stop   7999
	Date                    Owner   Queue   Job#
	08:52:00 07/08/98       root    c       62
	10:10:00 07/08/98       root    c       63
	11:24:00 07/08/98       root    c       64
	11:28:00 07/08/98       root    c       65
	15:02:00 07/08/98       root    c       68
	Start journal program
	Start dcctsk ?
	Start X windows stuff
	Starting sxserv
	Starting blktsk ...
	Starting /home/abacus/dmptsk
	Starting /home/abacus/rep 1
	Starting trendmem program
	/home/abacus/trendmem /home/abacus/trenddatafiles/trend.000.002.dat update
	datadump to be done every 300s approx 
	/home/abacus/trendmem /home/abacus/trenddatafiles/trend.001.002.dat update
	/home/abacus/trendmem /home/abacus/trenddatafiles/trend.002.002.dat update
	/home/abacus/trendmem /home/abacus/trenddatafiles/trend.003.002.dat update
	/home/abacus/trendmem /home/abacus/trenddatafiles/trend.063.002.dat update
	/home/abacus/trendmem /home/abacus/trenddatafiles/trend.080.002.dat update
	/home/abacus/trendmem /home/abacus/trenddatafiles/trend.080.060.dat update
	Starting engdrawserver.sh



If you get

	/home/abacus/faucet: Trying again . . .
	/home/abacus/faucet: Address 9000 in use, sleeping 10.
	/home/abacus/faucet: Trying again . . .
	/home/abacus/faucet: Address 9000 in use, sleeping 10.
	/home/abacus/faucet: Trying again . . .
	/home/abacus/faucet: Address 9000 in use, sleeping 10.

then it means that there is still a faucet running in the system using
address 9000. 

	frank(abacus):~$ ps ax | grep fauc

will find it, and then use kill -9 on it.




Bugs, complaints, suggestions and complements to abacusabacus@hotmail.com
please.

System Scratchpad Block

These blocks may be used to synchronise events with the system clock.

• block 2 minute of the day
• block 5 second of the current minute (0 - 59)
• block 6 minute of the current hour (0 - 59)
• block 7 hour of the day (0 - 23)
• block 8 day of the week (Sunday = 0)
• block 9 day of month (0 - 31)
• block 10 month of year (1 -12)
• block 11 year

• block 14 random value (0 - 1)

• block 20 DATADUMP control

am is set by the dump command in engineers mode, the datadump programme sets vn = v during the dump process. This only applies if the dmptsk line in the startup shell script ABACUS was configured with the -b 20 switch.

Scan Control Block

In addition to the common parameters the Scan Control Block has the following

• sp scan time in seconds floating_point_constant
• f1 first block to be scanned ABACUS_number
• l1 last block to be scanned ABACUS_number

Only blocks with number from f1 to l1 inclusive are processed.

Configuration

The format of the text file /home/abacus/abacus.sys is as follows

20000 Max Block Number

1 20 system

0 0 PiBus ai

201 250 K48

401 410 N45

0 0 siox3

0 0

0 0 pulse counter

801 810 input code

..

The first line has the highest block number in the system. Each line after that has the first and then last of each successive block type. These are in the order as shown with the look command. Subsequent text is not used and only serves as commentary.

All blocks of a particular block type have contiguous block numbers. A block number cannot refer to two different blocks at the same time, and so first and last block ranges must not overlap. Since every block number from 1 up to the maximum block number is allocated the common parameters, and is designated a scratchpad if not of any defined block type, the higher the Max Block Number the more space consumed.

Unused block types are marked with first and last block numbers equal to zero.

To initiate a new block configuration the command /home/abacus/initsk i is used. This should olny be done on a system where ABACUS is NOT running.

After initialising the database initsk spawns engineers mode to load up the applications in file /home/abacus/abacus.eng. This file usually contains engineers mode commands to set up scan rate values etc.

The total size of the database may not exceed a predetermined amount. When INITSK initialises the database it prints the total size.

Datadump files are stored in binary format, and are a straight copy of the memory into the file. It is therefore impossible to convert directly from one shape of ABACUS system to another without using a 'symbolic' dunp. eng2 may be used to do this before you change the block configuration of a system. Something like

eng2 b1,xb999999,short,fp > /tmp/symbolic.dump

will create a file called /tmp/symbolic.dump with all the eng commands required to re-establish you database. Check with an editor that the file is complete, before you destroy your old system!

When you've got your new system up and running you might want to use something like

eng2 xb99,@/tmp/symbolic.dump

to resotre your data. (xb99 overides the requirement for connecting to each block's tag)

Data dumping

In order to preserve the applications programme the entire ABACUS database is saved periodically. This process may be automatic depending on the dmptsk line in /home/abacus/ABACUS.

. This section covers communications between ABACUS4 and other systems, and other ABACUS4 systems.

MODBUS Communications

MODBUS Master

Each block which is intended for use as a MODBUS signal must have its nm setup with the following format

MODBUS <channel> <address> <R=register, or C=coil> <reg/coil reference> <IN/OUT>

Note that this software was developed against an ABB controller (commander series) where <reg/coil reference> is one greater than the 'offset address'.

Example: To send the ra of block 10002 to the 'Local Set Point' register (42) of a Commander 100 controller whose address is 1 on the modbus channel 1 the block should look like this...

Block  10002  () type scratchpad
nn.b.name                                
de.descr                                                                                  
nm.name   MODBUS 1 1 R 42 OUT      
tg.tag       spare sn.scan     0(0.0 sec) ra.rslta      123.000 
am.au/mn    a  vn.vltn     v  ss.sshot    r  
en.A Entpr          0 dd.D Entpr          0

Block  10001  () type scratchpad
nn.b.name                                
de.descr                                                                                  
nm.name   MODBUS 1 1 R 2 IN        
tg.tag       spare sn.scan     0(0.0 sec) ra.rslta      -18.000 
am.au/mn    a  vn.vltn     n  ss.sshot    r  
en.A Entpr          0 dd.D Entpr          0

The modbus program

modbus port baudrate channel sys-block-number

Here follows a trial run, using the above two blocks only. Note that the sys-block-number is left undefined, the default being 20+channel, i.e. 21 in this case.

frank(abacus):~$ /home/abacus/modbus /dev/cua1 9600 1
stty sane ispeed 9600 < /dev/cua1 
stty -ixon -ixoff -crtscts -parenb cs8 cstopb -echo raw ispeed 9600 < /dev/cua1 

Starting cycle...
MODBUS Block 10001 (n    -18), chan 1, adr 1, creg R, cregnr   2, direction IN
MODBUS Block 10002 (v    123), chan 1, adr 1, creg R, cregnr  42, direction OU
Starting cycle...
MODBUS Block 10001 (n    -18), chan 1, adr 1, creg R, cregnr   2, direction IN
MODBUS Block 10002 (v    123), chan 1, adr 1, creg R, cregnr  42, direction OU

frank(abacus):~$ eng2 b21,p
Block     21  () type scratchpad

B21 ()
nn                              
demodbus: Port /dev/cua1 at 9600 Baud, Channel 1. Found 2 blocks of interest     
nmmodbus: /dev/cua1       
tg   spare;sn  0(0.0 sec);ra      7.000
am  m ;vn  v ;ss  r 
en          0;dd          0

$ABACUS_HOME/modbus /dev/cua1 9600 1 > /dev/null 2>&1 &

Warning

MODBUS Slave

frank(abacus):~$ grep UC71:A720 analog.dat 
 2880 UC71:A720..               C  1   0    0   0        0      0.000      0.000 JARNHALTSMATARE - GRUVV.
 2881 UC71:A720..HL             C  0   0    0   0        0      0.000      0.000 JARNHALTSMATARE - GRUVV. HOG
 2882 UC71:A720..LL             C  0   0    0   0        0      0.000      0.000 JARNHALTSMATARE - GRUVV. LAG
 2883 UC71:A720..XL             C  0   0    0   0        0      0.000      0.000 JARNHALTSMATARE - GRUVV. INGET STROM