[etherlab-users] example code

[Graeme Foot]

Another thing you can look at is the "ethercat slaves -v" command.

If you have a linear topology then the last slaves "DC system time transmission delay" value tells you how long it takes for a frame to get sent to it, so if you double the value you get the approximate round trip time.

If you have a start topology (or various mixtures) then the round trip time will be less than double this value.

Then its down to allowing for your realtime jitter and EtherCAT master processing overhead etc.

G.

________________________________
From: etherlab-users-bounces at etherlab.org [mailto:etherlab-users-bounces at etherlab.org] On Behalf Of Jeroen Van den Keybus
Sent: Thursday, 21 November 2013 05:34
To: Curt
Cc: etherlab-users at etherlab.org
Subject: Re: [etherlab-users] example code

Using Xenomai and the specific patched drivers, I could easily go beyond 10 kHz with an EK1100 and 2 EL2008 on a Core 2 Duo 2.2 Ghz with e1000. I also did extensive measurement of delays and came up with about 6.5 us for the IgH master and something in the order of 10 us for the NIC from transmit start to first symbol on wire and back (last symbol on wire to receive start). So the NIC delays are certainly appreciable as well.

Are you running stock Linux (e.g. no RTAI, Xenomai or RT-PREEMPT) ?


J.


2013/11/19 Curt <cfiene at cybermetrix.com<mailto:cfiene at cybermetrix.com>>
Sorry for the newbie questions.

I've downloaded built and installed the ethercat master 1.5.2.

Used the generic driver to communicated with an Beckhoff EK1101 hub, EL1034
digital input and EL2034 digital output.

I then modified the ./example/user/main.c code to work with these modules.
Seems to work OK.   I then  looped the digital out to a digital in to some
basic performance testing.

At a update rate of 1kHz (1 millisecond), it can detect a change on average of
about 2 ms.    I increased the rate to 2kHz,  it can detect a change on
average of 1ms, BUT, I see an occasional, "EtherCAT WARNING: Datagram da69fecc
(domain0-0-main) was SKIPPED 1 time".   I don't see this warning at a 1kHz
update rate.

Are these reasonable results, and does this imply that the fastest rate I can
expect without errors/warnings is 1 kHz?

Would I get significantly better results using the Ethercat kernel module for
my chipset?


Thanks,

Curt Fiene
Cybermetrix
_______________________________________________
etherlab-users mailing list
etherlab-users at etherlab.org<mailto:etherlab-users at etherlab.org>
http://lists.etherlab.org/mailman/listinfo/etherlab-users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.etherlab.org/pipermail/etherlab-users/attachments/20131120/d63479bc/attachment-0004.htm>