Large-scale modular PLC with redundant CPUs (dual-rack hot standby)

01

Check switchover, task scan and loop response times

HollySys describes its LK large-scale PLC as designed for medium to large-scale control systems, offered with a single CPU or a dual-rack redundant CPU design, with redundancy in CPU, power supply and network. It states less than 100 ms CPU switchover time, less than 100 μs minimum task scheduling time and less than 200 ms loop response time. [1] [2]

The two racks are tied by the LK240 redundant synchronisation module with two LC optical-fibre interfaces; HollySys lists hot-plug optical modules for 2 km (multi-mode) and 20 km (single-mode) links, and the LK922 redundant DC power module accepts 19.5 to 60 V DC. [2]

Rockwell Automation, describing its ControlLogix 5590 high-availability system, states that switchovers occur as fast as 20 ms. Its list of switchover causes covers loss of power, a major controller fault, removal, insertion or failure of any module, loss of connection between redundancy modules, and loss of an EtherNet/IP connection that leaves the module unable to see any device on the network. It also warns that FactoryTalk Linx software can lose communication with the redundant pair for up to 30 seconds during a switchover. [3]

Editorial note: a controller switchover in tens of milliseconds and an HMI or historian gap measured in seconds are different quantities; ask the supplier for both, and for how outputs behave during the transfer.

Give the supplier the process to be controlled, the I/O count by signal type, which parts must be redundant (CPU, power, network, I/O), the tolerable switchover time, the HMI and historian software, fieldbus types, and the installation environment.

02

Choose CPU capacity, I/O and programming languages

HollySys's technical data for three LK CPUs: the LK220 runs at 600 MHz with 24 MB integrated program and data memory, the LK222 at 667 MHz with 32 MB and the LK224 at 766 MHz with 64 MB. Floating-point instructions take 25, 18 and 12 ns, and bit operations 15, 10 and 7 ns. Each has 512 KB of MRAM for power-loss retention. [2]

Fieldbus capacity differs by model: up to 128 Powerlink slaves on the LK220 and 256 on the LK222 and LK224, 124 Profibus-DP slaves on each, and 16, 32 or 64 Modbus TCP master/slave connections. Each CPU has two 10/100M RJ45 ports with dual-network-port redundancy. The CPU, power and communication modules sit on 4-, 6- or 7-slot control backplanes (LK130, LK132, LK133). [2]

HollySys programs the LK in its FA-AutoThink engineering software, which it says aligns with IEC 61131-3 and supports LD, FBD, CFC, SFC and ST; tasks can be cyclic, event-triggered, state-triggered or run at full speed, and protocol configuration covers Profibus-DP, Modbus, Powerlink and EtherCAT. [2]

Editorial calculation: moving from the LK220 to the LK224 multiplies integrated memory by 64 ÷ 24 ≈ 2.7 and doubles Powerlink slave capacity (256 ÷ 128 = 2), while floating-point time falls from 25 to 12 ns, about 52% shorter. [2]

For contrast, Chitic's PCS1200 is an all-in-one compact PLC whose single CPU module accepts up to 10 extension modules, with RS232 and RS485 ports and a stated instruction time of 0.1 µs. [4] Editorial note: that is the class of controller covered on the Programmable logic controller page; the large-scale form adds backplanes, extension racks and redundancy.

03

Dual-rack redundancy in HollySys and ControlLogix controllers

The supplier linked to this page lists a large-scale PLC with dual-rack redundancy; its card links to the page reviewed. The table compares the published controllers from the sources.

Ask each supplier for the switchover time with its test conditions, which modules can be hot-swapped, and the maximum I/O per rack and per system.

04

Backplane, paired CPUs and sync module that make the controller redundant

05

I/O points, scan time and ports to size the redundant CPU racks

Count the digital and analog points per rack and for the whole system, and state the scan time the redundant pair must hold.

06

Send the redundancy, I/O and network data together

Paste it into your enquiry below or into an email to a maker, then replace every bracket.

Redundancy: [CPU; power; network; I/O; tolerable switchover time]
I/O: [digital and analog counts; signal types; isolation]
CPU: [program size; task times; memory]
Networks: [Powerlink; Profibus-DP; Modbus TCP; Ethernet]
Software: [HMI; historian; batch]
Environment: [temperature; cabinet; EMC]
Documents: [IEC 61131-2 test reports; redundancy test record]

See manufacturer records Send an enquiry through CNFX