• State
  • Locked Locked
  • Replies 5 replies
  • Subscribers 34 subscribers
  • Views 375 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

LMH0397: Poor EOM diagrams for HD-SDI and SD-SDI formats

jmgalceran
Prodigy 10 points

Part Number: LMH0397

Hello,
we are using LMH0397 to receive SDI signals, with equalizer and CDR enabled. We read/write registers of LMH0397 usign a picoblaze in a FPGA, and through SPI bus.

We read EOM with 3G-SDI signals, reasonably well, but we don't get expected results with HD-SDI and SD-SDI signals.
We follow procedure as explained in www.ti.com/.../5797181755001 and in Programming Guide.

For example, setting timer threshold to 0xFF (register 0x2A) we get EOM like this for 3G-SDI:

eye_3gsdi.png

HEO=55 (about 47ps jitter), VEO=150 (about 469Vpp).

 

However we get EOM like this for HD-SDI:

eye_hdsdi.png

HEO and VEO measures are zero. EOM seems to be totally open.


With SD-SDI we get this EOM:

eye_sdsdi.png
Really strange, and also HEO and VEO measures are zero.


Do you think EOM for HD-SDI and SD-SDI are the expected ones? Why HEO is zero? Cannot it be made with a totally open EOM? Can I get a bit closer to a nice EOM for HD-SDI and SD-SDI?

Thanks a lot,

Josep

  • 0
    TI__Mastermind 26771 points

    Hi Josep, 

    EOM should work HD and SD-SDI operation. 
    For HD-SDI, it looks like the eye diagram is being captured, but the voltage range isn't being set correctly. Could you try manually setting the voltage range on EOM? You can disable VEO scale by setting register 2C bit 6 to 0 and setting register 11 bit 7:6 to 01, 10, or 11 for greater granularity. 

    Best,
    J

  • 0 in reply to J
    Prodigy 10 points

    Hi J,

    When we set manually the registers 0x2C (VEO scale manual) and 0x11 (eom_sel_vrange), EOM diagrams are much better.
    For example, we get this for HDSDI (VEO scale manual 9.3mV):

    And for SD-SDI (VEO scale manual 9.3mV):

    We have seen that when VEO scale is auto, register 0x29 reads 9.3mV for 3G-SDI, but 3.125mV for HD-SDI and SD-SDI. It seems that auto VEO scale does not set the best value in case of HD-SDI and SD-SDI. But no problem, we can set this range manually.

    However, what we are worried about is that HEO and VEO mesures are always zero. Even with 3G-SDI format, VEO measurement is very variable (between 10 and 150). For HD-SDI and SD-SDI they are always zero.

    We follow Programming Guide (clause 3.11) to read VEO and HEO values.

    Is there any special trick to get HEO / VEO?

    Regards,

    Josep

  • 0 in reply to jmgalceran
    TI__Mastermind 26771 points

    Hi Josep, 

    Yes, the automatic scaling may not work at some times. 

    On HEO and VEO, the registers may overflow if the eye height and the width is too big compared to the scale. Could you try with the highest scale setting and see if there is any value on HEO and VEO?

    Best,
    J

  • 0 in reply to J
    Prodigy 10 points

    Hi, J, 

    It seems that changing vertical scale does not help.

    We have also tried to reset Cable Eq Acquisition and/or CDR, but same behaviour occurs.

    However, we can make these mesurements graphically, based on EOM diagram, assuming that the real measurement will be higher that the graphic one, right?

    Regards,

    Josep

  • 0 in reply to jmgalceran
    TI__Mastermind 26771 points

    Hi Josep, 

    It also could be that the VEO and HEO are not triggering at the right places. I know in some cases the VEO and HEO may measure the wrong eye height/width causing incorrect value to be stored in the register. 

    The measurements can be made using EOM diagram, too. 

    Best,
    J