ECAT configuration

[koustubh_h]

Hi,

I am facing some problems with ECAT configuration when it comes to setting alias.

I am using ARM processor PPMAC 1020 CPU with firmware version 2.6.0.0

 

I tried doing the configuration with PMAC IDE vers. 3.1.4.0 and 4.4.0.99 but with the same results.

 

Now about the problem,

I am connecting 4 drives (ABB make Motiflex e180) and 2 encoders (Hengstler make) through ECAT switch (Beckhoff make CU1128) in star topology.

 

When i scan devices i can see all slaves and i m able to generate ENI file and load to PPMAC and run properly, no issues.

This is what i get when i type "ecat slaves" in the terminal window -

ecat slaves

0 VID=$00000002 PC=$04685432 0:0 OP + Slave_1001 [CU1128]

1 VID=$000000B7 PC=$000002C0 0:1 OP + Slave_1002 [MotiFlex e180]

2 VID=$000000B7 PC=$000002C0 0:2 OP + Slave_1003 [MotiFlex e180]

3 VID=$00000002 PC=$04685432 0:3 OP + Slave_1004 [CU1128-B]

4 VID=$000000B7 PC=$000002C0 0:4 OP + Slave_1005 [MotiFlex e180]

5 VID=$000000B7 PC=$000002C0 0:5 OP + Slave_1006 [MotiFlex e180]

6 VID=$00000002 PC=$04685432 0:6 OP + Slave_1007 [CU1128-C]

7 VID=$20041961 PC=$00000547 0:7 OP + Slave_1008 [ACURO]

8 VID=$20041961 PC=$00000547 0:8 OP + Slave_1009 [ACURO]

There is no mismatch and i am able to Activate EtherCAT etc., etc.,

 

But i need a configuration where i should be able to remove any device intermittently.

For that i set "identification value" to the station address expecting it to write to EEPROM.

it does, when i go to diagnosis mode and verify at EEPROM tab at 0x0004 i can see the station address assigned.

But, the same address is not replicated to ESC register no. 0x0012 which it must.

That is why it creates a network mismatch which when i observe from "Network Mismatch Analyser" i see configured alias as 1002 (the station address assigned) but the network alias keeps showing 0 (which is the value at ESC register 0x0012.

 

Even after making group by putting station address at 0x0012 or by putting offsets at input and output to "pinned group" it doesn't work.

 

Hence,

if i remove one slave from the configuration the next slave gets detected as the earlier one and the whole configuration goes for a toss.

 

The same phenomenon is observed in PMAC IDE 3.1.4.0 through EC-engineer or through PMAC IDE 4.4.0.99 with embedded EC-Engineer.

Also network mismatch analyser is absent in vers. 4.4.0.99.

 

Please help.

Maybe i am doing something wrong, so kindly help me pin-point where i could be.