[etherlab-users] DC transmission delay w. fibre optic junctions

[Matthieu Bec]

After reading the log, I checked out the top devel branch - that fixed 
my DC problem.
Is there a rough timeline for a stable-1.6 to come out?

Thanks -



On 1/9/12 10:51 AM, Matthieu Bec wrote:
>
> Hello all,
>
> I have an EtherCAT bus using fibre optic junction modules 
> (EK1521/EK1501) to access remote I/O. The DC system time transmission 
> delay reported by ethercat tool is giving me some odd numbers:
>
> # ethercat slaves -v| egrep transmission
>   DC system time transmission delay: 0 ns           // slave 0 - EK1100
>   DC system time transmission delay: 574771462 ns   // slave 1 - EK1521
>   DC system time transmission delay: 2147484367 ns  // slave 2 - EK1501
>   DC system time transmission delay: 2147484512 ns  // ...
>   DC system time transmission delay: 2147484657 ns
>   DC system time transmission delay: 2147484802 ns
>   DC system time transmission delay: 2147485346 ns
>   DC system time transmission delay: 2147485486 ns
>   DC system time transmission delay: 2147485631 ns
>   DC system time transmission delay: 2147485776 ns
>   DC system time transmission delay: 2147485921 ns
>   DC system time transmission delay: 2147486076 ns
>   DC system time transmission delay: 2147487336 ns
>   DC system time transmission delay: 2147487476 ns
>   DC system time transmission delay: 2147487631 ns
>   DC system time transmission delay: 2147488491 ns
>
>
> The offset values downstream of slave-2 resemble a lot to
>   2^31 =  2147483648
>
> I'm wondering it that might point to some signed-ness (or other) issue 
> in the DC offset computation in the ethercat master?
>
> The weird offset prevents me to control an EL2252 DC-module programmed 
> with time stamp. The device works fine if I don't use the fiber link.
>
> Has anyone encountered or knows how to fix this?
> I'm using master stable-1.5
>
> Many thanks -
>