[etherlab-users] Question about cascading EK1100

[Thorsten Schnebeck]

Hi,

old thread, but somehow the same problem:

> >Michael Sagner schrieb:
> >
> > I’m planning to use several EK1100 couplers on one ethernet line.
> > The first EK1100 will be connected directly to a PC’s Ethernet port 
> > (the PC is running Etherlab),
> > The second one will be connected via the slave Ethernet port of the 
> > first EK1100 and so on.
> > Q: What must be done to initialize the slave EK1100’s? Is there a 
> > C/C++ code sample for this, someplace?

> Andreas Stewering-Bone schrieb:
>
> There is no need to initialize the EK1100,
> but it is possible.
> You have to act on it like a normal slave without pdo configuration
> #define BusCouplerPos 0, 0
> #define Beckhoff_EK1100 0x00000002, 0x044c2c52
> ...
> sc = ecrt_master_slave_config(master, BusCouplerPos, Beckhoff_EK1100);
> if (!sc)
> return -1;
> and that is all

We also want to use cascading couplers but as part of ML/Simulink generate code.
When we want to use two EK1100 boxes the Matlab code generator creates wrong code :

ulser3.c: In function 'MdlStart':
ulser3.c:3988: error: 'sync_manager_2_044c2c52_00000000_0' undeclared  
(first use in this function)
ulser3.c:3988: error: (Each undeclared identifier is reported only once
ulser3.c:3988: error: for each function it appears in.)
gmake: *** [ulser3.o] Error 1

This is the generated code for the two couplers:

       /* S-Function Block: <Root>/Infrastructure
        * Registering EtherCAT block EK1100 with Driver
        */
       if ((errstr_rc = ecs_reg_slave( 0, /* TID */
         /* MasterId, DomainId, SlaveAlias, SlavePosition */
             0U, 0U, 0U, 1U,
         /* VendorId, ProductCode */
             2U, 72100946U,
         /* SdoConfigCount, SdoVar */
             0, NULL,
         /* SoeConfigCount, SoeVar */
             0, NULL,
         /* SyncManager Configuration */
             NULL,
         /* Distributed Clock Configuration */
             NULL,
         /* Input/Output Configuration */
             0, NULL))) {
         snprintf(etl_errbuf, sizeof(etl_errbuf),
                  "EtherCAT slave ifumPulser3/Infrastructure register "
                  "failed: %s", errstr_rc);
         rtmSetErrorStatus(ifumPulser3_M, etl_errbuf);
         return;
       }

       /* S-Function Block: <Root>/Infrastructure1
        * Registering EtherCAT block EK1100 with Driver
        */
       if ((errstr_rc = ecs_reg_slave( 0, /* TID */
         /* MasterId, DomainId, SlaveAlias, SlavePosition */
             0U, 0U, 0U, 20U,
         /* VendorId, ProductCode */
             2U, 72100946U,
         /* SdoConfigCount, SdoVar */
             0, NULL,
         /* SoeConfigCount, SoeVar */
             0, NULL,
         /* SyncManager Configuration */
             sync_manager_2_044c2c52_00000000_0,
         /* Distributed Clock Configuration */
             NULL,
         /* Input/Output Configuration */
             0, NULL))) {
         snprintf(etl_errbuf, sizeof(etl_errbuf),
                  "EtherCAT slave ifumPulser3/Infrastructure1 register "
                  "failed: %s", errstr_rc);
         rtmSetErrorStatus(ifumPulser3_M, etl_errbuf);
         return;
       }
     }

If I change manually the sync_manager_2_044c2c52_00000000_0 -> NULL everything works as expected and both couplers do enter OP-mode.

Any hints how this can be fixed?

Bye

  Thorsten

-- 
etherlab 1.3pre670