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.

LMH0384 SDI Rate

Other Parts Discussed in Thread: SD384EVK, LMH0384, LMH1219

Hello,

I'm using the SD384EVK to evaluate the LHM0384 and had a quick question regarding the SDI rate of the LMH0384.

I just wanted to ask per the datasheet that the maximum input data-rate was specified at 2.97 Gbps. This is ~1.5 GHz for falling and rising edge. Does this mean I can feed into the input of the LMH0384 a signal of up to 1.5 GHz?

I am asking this because I can only go as high as 950 MHz into the input of the LMH0384 (SDI). There is no mentioning of maximum input frequency in the datasheet (just maximum DR of 2.97 Gbps), so I may be misinterpreting things.

Regards,

Hayden

  • Hi Hayden,
    As you noted LMH0384 works up to 2.97 Gbps. Carrier detection circuitry does not declare carrier detect if you input a clock frequency in the range you noted. Once carrier detect is inactive, the signal path within the device is shut off or muted. LMH0384 carrier detects works fine with data pattern up to 2.97 Gbps.
    Regards,,nasser
  • Thanks Nasser!

    I just want to clarify quickly that the device should be able to function with up to 1.5 GHz input?

    I don't have a bit analyzer to create a Pseudo Random Bit Stream (PRBS). I have a signal generator and am currently inputting square waves up to 1.5 GHz (not Gbps). I assumed 2.97 Gbps =~ 1.5 GHz, which is a conversion ration of 2 when talking about data rate vs. the raw frequency of the information waveform.

    So far, it works up to 950 MHz, but with anything higher there is no output.

    Regards,

    Hayden

  • Hi Hayden,

    The LMH0384 operates up to 2.97 Gbps, which is the equivalent of sending a 1.485 GHz through the device if you are sending a '1010' clock pattern through the device. Your calculation is correct.

    The LMH0384 device is a cable equalizer only, and therefore does not have a reclocker that locks to specific data rates (compared to, say, the LMH1219 with an integrated reclocker). I would have guessed that you should have no trouble equalizing and outputting a clock signal going up to 1.485 GHz. However, as this device is also only characterized for functionality at SMPTE data rates, I would use 1.485 GHz as a data point as opposed to 950 MHz.

    What are you seeing once you get past 950 MHz? If the output shuts off, I wonder if there is a Mute Threshold that is going into effect. How much cable are you using before the LMH0384, and what is your JP5 pin setting? Please double-check that it is left unconnected so that the LMH0384 does not auto-mute.

    Thanks,

    Michael

  • Hi Michael,

    I currently have JP5 open/disconnected for normal operation.

    The cable length is only a couple of inches. The output of the signal generator is transformed from 50 ohms to 75 ohms via a transformer, who is then connected to the 75 ohm BNC of the LMH0384. I can see the signal on the scope up to 950 MHz, anything higher than that, no more waveform, i.e. a flatline.

    Also, I've attached a picture of the board with the jumper configurations.

    Regards,

    Hayden

  • Hi Hayden,

    The settings look reasonable to me. I'm curious about what is going on here, because the device should not be shutting off like that. Can you try the following:

    1. Can you take a scope shot of the SDO output for 135 MHz (the equivalent of 270 Mbps, i.e. SD data rate), then compare with a scopeshot of the SDO output for 950 MHz?

    2. What is the amplitude of the output right at the 75-ohm converter output? Do you have a scopeshot you can share of this for the same two frequencies as requested above? SDI calls for 800 mVp-p (single-ended) +/- 10%. If the amplitude is too low, it could cause the LMH0384 to mute. Just a sanity check.

    Thanks,

    Michael
  • Hi Hayden,

    To close out this thread, we will handle detailed customer questions via e-mail.

    Thanks,

    Michael