• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 2 answers
  • Subscribers 101 subscribers
  • Views 2928 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

AM3352: RGMII timing issues

yimin bai34
Genius 4755 points
Part Number: AM3352
Other Parts Discussed in Thread: STRIKE

Hi Sir 

we used AM3352 for development and have some issues which need your suggestion.

a. for item4 - TXC, we measured the time is 1.5ns but the max. time of spec is 0.75.

  any improvement method or could you confirm it is ok? 

2. In RGMII spec, the TCTL signal will change with the data signal at the same time. 

   We measure the signal of TCTL is high level during data transmission.

    you can see below photo. the green line is the TCTL signal. Could you confirm it is ok or not?

  • 0
    TI__Guru 59615 points

    You are correct the RGMII specification defines a max rise/fall of 0.75ns. It is very difficult to achieve this fast rise/fall with the output buffer used in the AM3352 device. This may be possible if the combined capacitive load from the PCB trace and attached device is very small. However, if you achieve the faster rise/fall time on these signals there is a potential this will create another issue. Many customers have found fast RGMII signals create significant radiated emissions which makes it difficult to pass CE and FCC testing. Therefore, the speed of the output buffer has been designed to provide optimum performance for both timing closure of the RGMII interface while minimizing radiated emissions.

     

    TI recommends you perform a timing analysis of this interface taking into account any PCB delays and specific timing requirements/switching characteristics of both AM3352 and attached RGMII device. It may be necessary to adjust PCB delays of individual RGMII signals relative the each other to provide proper timing margins.

     

     

     

    If you read the description of TXERR and RXERR Coding in the RGMII specification, you will find the following description.

     

    TXERR is treated in a similar manner. During normal frame transmission, the signal stays at a logic high for both edges of TXC and during the period between frames where no errors are to be indicated, the signal stays low for both edges.

     

     

    Regards,
    Paul

  • 0
    Genius 4755 points

    Part Number: AM3352

    Hi Sir 

    for RGMII TXC falling/Rising time, we measure SK-EVM and found the same issue. 

    the max. time of spec is 0.75ns but SK-EVM doesn't meet it. 

    Q1: Is it the limitation for TI AM3352 which cannot be under 0.75ns falling/rising time? 

    Q2. If above issue is true, does it cause any issue for RGMII timing?

    Thanks for the reply.

    BR

    Yimin

  • 0 in reply to yimin bai34
    TI__Mastermind 27890 points
    Hello Yimin,
    As Paul outlined in his response to this topic, this device was designed to strike a balance between interface timing and radiated emissions. Driving a (relatively) high-speed signal to meet a 0.75ns rise time would cause excessive radiated emissions that would likely not pass FCC and/or CE certification tests. I think it also important to note that the 0.75ns requirement outlined in the RGMII specification doesn't mention signal loading at all, which could be interpreted as the value being measured out of circuit.

    This family of devices has been in high-volume production for many years now with no RGMII timing issues reported. TI does recommend that a timing analysis of all high-speed interfaces be performed prior to PCB fabrication as no two boards share the variables that impact interface timing.