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.

Keystone II SGMII signal levels

Other Parts Discussed in Thread: 66AK2H12

We are connecting an SGMII port of the 66AK2H12 to a 1000Base-KX port across a VPX backplane thru AC coupling caps.   The 1000Base-KX is a 1.25Gbps SERDES version of gigabit Ethernet (per IEEE 802.3-2008).   The 1000Base-KX has transmitter characteristics of 800 to 1600 mV differential peak-to-peak output voltage.   The IEEE standard also recommends using only 4.7nF AC coupling caps (to reduce inrush current in the receiver).

My questions are:

1) What is the absolute voltage ratings and recommended operating voltage range for SGMII differential receiver inputs?     The K2 Serdes guide says it conforms to LVDS.   But I'd like to know the specific K2 ratings.

2) Is 4.7nF AC coupling caps sufficient for the K2?    The K2 documentation recommends 100nF.

3) When will the K2 Hardware Design guide be released?

thanks.

  • The K2H is only designed to support SGMII. SGMII and 1000Base-KX are similar but not identical standards. Specifically the SGMII interface can support 10, 100 & 1000 Mb/s operation through the SGMII auto-negotiation while 1000Base-KX is 1000Mb/s only and doesn't support negotiation.

    1) The K2H follows the SGMII specification which allows for Vi to range from 675mV to 1725mV.

    2) All operation on the SGMII is characterized using the 100nF specified. I don't know if 4.7nF is sufficient.

    3) The KS2 HDG is currently in review and should be released soon.

    Regards, Bill 

  • Hi Bill,

    Thanks for your reply.     We understand that K2H SGMII interface doesn't officially support connecting to 1000Base-KX  (SGMII is at the MAC layer, the 1000Base-KX is at the PHY layer).   However, we've heard others have made it work by forcing the link settings (and disabling auto-negotiation), which is totally acceptable for our application.    

    Please let us know if you know that it cannot be made to work this way.

    Scott

  • Hi Scott,

    What your suggesting appears to solve the auto-negotiation issue and I didn't see any discrepancies in the signal levels. I am not aware of any customer that has tried this. If you have success please let me know.

    Regards, Bill