schmersal Praxi 100, Руководство пользователя

Страница: 11 / 52

11
Losing one packet is not a big issue because the same request can be sent again after timeout. In TCP, the packages
always reach their destination but if network congestion causes retransmissions, the resent packages will most likely
contain old data or instructions when they reach their destination.
The Praxi 100 family AC drive supports 10/100Mb speeds in both Full- and Half-duplex modes. However, real-time
process control requires the Full-duplex mode and the 100-megabit speed. Drives must be connected to the Ethernet
network with a Shielded Twisted Pair (STP) CAT-5e cable (or better). Use only industrial standard components in the
network and avoid complex structures to minimize the length of response time and the amount of incorrect dispatches.
The maximum length of an RS-485 cable depends on the bitrate used, the cable (gauge, capacitance or characteristic
impedance) and the number of devices in the bus. The Modbus RTU specification states that for a maximum 9600 bits/
second bitrate and AWG26 or wider gauge, the maximum length is 1000 meters. The actual cable length used in an
installation can be lower than this number depending on the aforementioned parameters.
The Praxi 100 family AC drive implements the ACD algorithm (IETF RFC 5227).
The ACD algorithm tries to actively detect if the IP address configured to this device is used by another device in the
same network. To accomplish this, the ACD sends four ARP request packets when the device’s Ethernet interface goes
up or when its IP address changes. The ACD prevents the use of the Ethernet interface until the ARP probing finishes.
This delays the startup of fieldbus protocols about one second. During the delay or after it, the ACD passively checks
incoming ARP messages for use of the device’s IP address. If another device with the same IP address is detected, the
ACD will try to defend its IP address with a single ARP message. If the other device with the same IP address also
supports ACD, it should stop using the address. If not, the ACD will close the Ethernet connection and indicate the
situation with an Alarm. This is done according the “DefendWith- PolicyB”. Acknowledging of the Alarm is not possible
if the problem is active. The ACD opens an Ethernet connection if the other device with the same IP address disappears
from the network. The alarm can be acknowledged after this. Other policies are not supported. If the fieldbus protocol
has been active, a fieldbus fault may be activated (depends on the fieldbus and drive application configuration).
The ACD functionality can be enabled and disabled with Duplicate IP Detection panel parameter (see Chapter 5.3.1
Ethernet common settings (M5.9.1)).
3.4 Connections and Wiring
3.5 ACD (Address Conflict Detection) in Ethernet Network