EtherCAT Slave on AM3359 using DC Sync0 mode

Hi,

I have a problem with the "System Time" (ECAT register 0x910-0x917) 64bit counter. The values you can read out from this register are pretty "confusing" during startup (INIT - PREOP - ...). Is the source code from the PRUSS firmware ecat_frame_handler_bin.h and ecat_host_interface_bin.h available (pasm Assembler)? I would like understand and may change the behavior of the "System Time".

I am using ESC firmware 1.2.96 from the sys/bios sdk version 1.0.0.8. So it is not the latest one. Master is TwinCAT 3.1.

Best regards

over 10 years ago

0 Biser Gatchev-XID over 10 years ago

TI__Guru**** 393215 points

Hi,

I will forward this to the ISDK team.

0 Frank Walzer over 10 years ago

TI__Mastermind 44306 points

Hi VS,

no, we do not release the source of our EtherCAT ESC. We are not even allowed due to legal reasons. Now why do you need to analyze the System Time directly? Usually this controlled/managed by the master system. We are following the standard and as such there should be no changes needed. Otherwise you need to tell us what is wrong and we can file it as an issue.

Best regards,

0 PratheeshGangadhar over 10 years ago

TI__Mastermind 49951 points

> I am using ESC firmware 1.2.96 from the sys/bios sdk version 1.0.0.8. So it is not the latest one. Master is TwinCAT 3.1.

This recommended to update to the latest version 1.1.0.8 or even better to 2.1.0.1 - on fixed errata info please refer to http://processors.wiki.ti.com/index.php/AM335x_EtherCAT_Slave_Errata

0 VS over 10 years ago in reply to Frank Walzer

Prodigy 40 points

Hi,

I found the problem in the Errata (problem SDOCM00103985). I don't think it is a problem of the EtherCAT master (TwinCAT), because I checked the values from the System Time (0x0910-0x0917). During startup (INIT - PREOP) the high word and the low word of the System Time operate independent. For instances, if the low word stops already, the high word is still able to count. When TwinCAT read the System Time, it is likely to get an invalid value (high word and low word) and calculate from that value an useless Start Time.

Who owns the legal rights on the PRUSS firmware, may be Beckhoff?

Best regards

0 Frank Walzer over 10 years ago in reply to VS

TI__Mastermind 44306 points

VS said:

Who owns the legal rights on the PRUSS firmware, may be Beckhoff?

TI owns the firmware but due to Beckhoff IP we can't open this.

regards,

Processors

Processors forum

EtherCAT Slave on AM3359 using DC Sync0 mode