• State TI Thinks Resolved
  • Locked Locked
  • Replies 11 replies
  • Answers 1 answer
  • Subscribers 33 subscribers
  • Views 323 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.

DS15BR400: DS15BR400

Goran Fiolic
Intellectual 420 points
Part Number: DS15BR400

We are using DS15BR400TSQ in our new project and we are having some issues. While input signal looks perfectly OK, the output , however, looks like modulated input. Pre-emphasis was turned off in this measurement. We triple-checked everything (Schematics, Soldered components, Layout) but can't really find anything that would be causing this modulation. Could you provide any explanation for this modulation? Output is connected to terminated FPGA  LVDS input. We have an older version of the board without DS15BR400TSQ where FPGA output drives FPGA input directly - no issues here. The new PCB version has DS15BR400TSQ inserted in between FPGA output and input. Connection between inputs and outputs is made using RJ45 connectors and 50cm STP cable (On both PCB versions). The idea behind using DS15BR400TSQ was to extend communication range (FPGA inputs and outputs are located on the same FPGA).

  • 0
    TI__Guru* 93670 points

    Greetings,

    Agreed this looks somewhat modulated. Was this measurement done single ended or differential? I am thinking this could be a setup or a measurement issue the device does not have any intelligence to modulate the waveform. Please let us know.

    Regards,, Nasser

  • 0 in reply to Nasser Mohammadi
    Intellectual 420 points

    Measured was done single ended at 'p' leg of differential pair. Input was measured first (Input from FPGA directly at DS15BR400 input pin), and then the output (again, directly at the pin). In both measurements oscilloscope was connected to GND at the same PCB location. Input looked OK, but output was modulated.

  • 0 in reply to Goran Fiolic
    TI__Mastermind 35130 points

    Hi Goran,

    I confirmed this behavior in the lab. It seems it only occurs if the outputs are terminated differentially. When they are terminated with single-ended 50-ohms to ground the modulation disappears. Not quite sure why this is the case (100-ohm differential termination should be okay) but ultimately it will not have an effect on the system since the modulation gets canceled out when measured differentially. 

    Regards,

    I.K.

  • 0 in reply to Ikechukwu Anyiam
    Intellectual 420 points

    Hi I.K.

    Thanks for the answer. I assumed that modulation would not be affecting differential signal reception. However, since this modulation looks to me like some kind of buffer instability I have some doubts about continuing to use this device. Could you confirm that modulation would not change within buffer's operating temperature range in a way where differential signal transfer would become corrupted? Also, if up to 5 meters of UTP or STP cable is connected to buffer's output, can you guarantee buffer's norma operation?

    Regards

    Goran

  • 0 in reply to Goran Fiolic
    TI__Mastermind 35130 points

    Hi Goran,

    I can set up a test for this in the lab to confirm, though it will have to be some time next week. Have you seen any issues with the differential signal transfer in your system?

    Regards,

    I.K. 

  • 0 in reply to Ikechukwu Anyiam
    Intellectual 420 points

    Hi I.K.

    So far no, but we did not perform all the tests yet so I don't have any definitive answer to this right now. We are proceeding with our tests, and I will have some more data to share on Monday.

    Regards

    Goran

  • 0 in reply to Ikechukwu Anyiam
    Intellectual 420 points

    Hi,

    We were testing our device and fortunately everything worked as expected. We tried LVDS link with pre-emphasis turned on over 5 m, and with pre-emphasis turned off over 2 m and 1 m UTP cable. We also performed the same tests while cooling and heating LVDS buffer. Anyhow, we are currently unable to do these tests using a larger number of devices so if you could do the same thing on your side (or anything else to boost our confidence in this buffer) that would be helpful.

    Regards

    Goran

  • 0 in reply to Goran Fiolic
    TI__Mastermind 35130 points

    Hi Goran,

    That's good news. I'll see if I can set up a test this week. I don't have the capability of manually testing a large number of devices on the bench, but every device is tested with automatic testing equipment before being shipped. This device has also been on the market for over 14 years without any major reported issues so I'm not expecting to see any problems with the test.

    Regards,

    I.K.

  • 0 in reply to Ikechukwu Anyiam
    Intellectual 420 points

    Hi I.K.

    OK. Thanks for your effort. Let me know the results of your tests. If everything checks out we will proceed with production (currently is on hold)..

    Regards

    Goran

  • 0 in reply to Goran Fiolic
    TI__Mastermind 35130 points

    Hi Goran,

    Ran some tests today with 2 EVMs (one as the TX and one as the RX) and didn't see any issues with transmitting/receiving differential signals across temp. You should be good to proceed with your production.

    Regards,

    I.K.

  • 0 in reply to Ikechukwu Anyiam
    Intellectual 420 points

    HI I.K.

    OK, Thanks again for your support..

    Best Regards

    Goran