AM3359: EtherCAT Slave clock architecture

varun Rajamanickam

Part Number: AM3359
Other Parts Discussed in Thread: AMIC110, , TMDXICE110, TMDSICE3359

Whether AM3359/AMIC110 (EtherCAT Slave Controller) and Ethernet PHY should be sourced from same Clock Source for EtherCAT Communication?

Implementation in TMDSICE3359 and TMDXICE110 are quite different w.r.t clock architecture.

Do you have any suggestion / guideline related to clock architecture for EtherCAT Slave controller utilizing AM3359/AMIC110 ?

Note:- I am seeing guideline available from Beckhoff for their EtherCAT Slave Controller - page 5

Regards

Varun R

over 8 years ago

0 Biser Gatchev-XID over 8 years ago

TI__Guru**** 393215 points

Hi,

Both TMDSICE3359 and TMDXICE110 boards support EtherCAT Slave, so you can choose the clock generation scheme that suits your design.

0 varun Rajamanickam over 8 years ago in reply to Biser Gatchev-XID

Prodigy 240 points

Hi Biser

Thanks for your reply. I understand both TMDSICE3359 and TMDXICE110 boards support EtherCAT Slave controller.

Is there any difference in performance ?

Beckhoff suggest one way of clocking, Whether TI recommends any one preferred way of clock generation?

Regards
Varun R

0 Paula Carrillo over 8 years ago in reply to varun Rajamanickam

TI__Mastermind 40580 points

Hi Varun, you are right, it is recommended to replace the three individual crystals for the PHY and SoC, with a single clock distribution circuit (such as CDC913E), in order to ensure dual PHY and minimize phase jitter.

This is typically accomplished using external clock buffer.

We have a reference TI design (TIDA-00299) that might be a good reference for AMIC110 and EtherCAT.

Thank you,

Paula

Processors

Processors forum

AM3359: EtherCAT Slave clock architecture