• State
  • Locked Locked
  • Replies 7 replies
  • Subscribers 32 subscribers
  • Views 803 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.

DS100KR401: DS100KR401 vs SN65LVCP114

Michael Nycz
Intellectual 835 points
Part Number: DS100KR401
Other Parts Discussed in Thread: SN65LVCP114, , DS125BR820, DS125BR401A, DS125BR401

I have a media converter design that converts between 10G-KR and 10G-SR (copper to fiber).

I was having issues getting the 10G-KR side to link/pass data. I obtained an SN65LVCP114 development board and by adjusting settings I was able to improve the signal enough that the converter now would link and pass data at 10G. This was great! However, I then discovered that the SN65LVCP114 was only rated for -10DegC and my application requires a device that will operate at -40DegC.

I then looked at the DS100KR401 and it looked like it would perform similar signal conditioning functions to the SN65LVCP114 WITH a temperature rating down to -40DegC. This seemed like it should solve my problem. I obtained a development board for the DS100KR401 and hooked it up in place of the SN65LVCP114. I can now get my converter to link at 1G data rate but when I try at 10G, none of the settings are producing results like I had obtained with the SN65LVCP114.

Does anyone have any ideas what might be wrong? Is it in error to conclude that the SN65LVCP114 and the DS100KR401 would provide a similar function from a signal conditioning standpoint? The other thing we were going to try is running the SN65LVCP114 down to -40DegC in a temperature chamber. I know that it is not rated there but is it likely that it would function without problems at -40DegC?

Any help with this would be greatly appreciated.

Thank you,

Mike

  • 0
    TI__Mastermind 43795 points

    Hi Mike,

    The DS100KR401 and the LVCP114 are both linear equalizers, but they can react differently in systems.  The DS125BR820 is closer to the LVCP114 in architecture and performance.

    Can you tell me what is the system topology?

    What is the approximate loss in each channel segment?

    Are the system Tx FIR drivers using de-emphasis and pre-shoot in your testing?

    With this information I can help you define settings which will work at 10G as well.

    Best Regards,

    Lee

  • 0 in reply to Lee Sledjeski
    Intellectual 835 points

    Hi Lee,

    Thank you for the response. I see that the DS125BR820  is a uni-directional part. The DS100KR401 is a bi-directional part. Would the DS125BR401A be the cousin to the DS125BR820? Just bi-directional?

    We are using the redriver/conditioner as part of a media converter. The redriver is on the same PCB as a 10G-KR to 10G-SR converter module (4 channels). The setup that we were running is the following....

    Loop the copper side of two of the channels on the converter (or in this case, two channels of the redriver). The redriver then connects to our optical transceiver. The two optical channels are then connected to off the shelf media converters. The copper side of the media converters are then connected to PC's. We can get the setup to link to the 2 media converters at 1G and we can pass data (ping test). We then switch to 10G and can never achieve a link with the DS100KR401. If we switch to the SN65LVCP114 then we achieve the link and are able to ping.

    So in this case, on the external cable side of the converter, the redriver is talking to itself.

    We are in the process of setting up a scope to observe the eye diagram output of the redriver. We did that for the SN65LVCP114 but then we stumbled upon settings that would allow us to pass data so when we switched to the DS100KR401 the presumption was that we could find settings to get the redriver/converter to pass data. The oscilloscope is borrowed from another group but we should have it today to do some testing.

    We were also going to test the SN65LVCP114 at -40DegC (since it worked at room temperature). I am being asked to do this as a backup. Let me know if you have any further insights. If the DS125BR401A is a close cousin to the SN65LVP114 then that would be cool as well. I have a PCB laid out with the DS100KR401.... The DS125BR401A is nearly a drop in footprint!

    Thanks again,

    Mike 

  • 0
    Intellectual 835 points
    Also, Does the DS125BRXXX redrivers support 10G-KR?
  • 0 in reply to Michael Nycz
    TI__Mastermind 43795 points

    Yes the DS125BR401A supports 10G-KR as it is designed to be linear and therefore transparent to the TX FIR information which is going through it to the 10G-KR receiver.

    The A-side inputs have the best linear behavior with larger signals - so it is best to align the A-side inputs with the closest 10G-KR transmitter.

    I would need some additional information on channel segment attenuation to come up with approximate settings.  Tuning multiple re-drivers cascaded together can be frustrating.  I would start with the lowest CTLE settings and slowly work higher if needed.  Try to minimize the DEM as well to start with.

    Regards,

    Lee

  • 0 in reply to Lee Sledjeski
    Intellectual 835 points
    Lee,
    Thank you again for the information. With regard to chip settings.... The DS100KR401 has a pin MODE (pin 21) which sets either 10G-KR Mode Operation or 10G Mode Operation. For the DS125BR401 pin 21 is MODE_B which is set in these ways....
    Tie 1kΩ to GND = GEN 1,2 and SAS 1,2
    Float = Auto Mode Select (for PCIe)
    Tie 20kΩ to GND = SAS-3 and GEN-3 without De-emphasis
    Tie 1kΩ to VDD = SAS-3 and GEN-3 with De-emphasis

    For my case, which of these would be the proper choice?

    Thanks,
    Mike
  • 0 in reply to Michael Nycz
    TI__Mastermind 43795 points

    Hi Michael,

    I would just let the MODE_B pin float.  It will give you "linear" behavior at 10G-KR rates and act as a "limiting" amplifier for the 1G rate.  

    Regards,

    Lee

  • 0 in reply to Lee Sledjeski
    Intellectual 835 points

    Lee,

          Thank you again for all the guidance. I was working with the guys who were actually doing the testing with the DS100KR401 eval board. After digging through the setup we were now able to get the device working with similar settings to the SN65LVCP114. There were some setup issues and once we fixed those things seem to work as they should. We will be moving forward with the DS100KR401!

    One last question, we were trying to monitor temperature on this PCB. Is there a register on the DS100KR401 that would indicate relative temperature changes? We will do something else if not, but if there is already a temperature being measured on the device that would make things easier. 

    Thank you,

    Mike