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.

LMH3087 against LMH0344 and LMH0302 combined

Other Parts Discussed in Thread: LMH0387, LMH0344, LMH0302

 

 

Team,

Can the LMH0344 and LMH0302 together can perform the same functions as the LMH0387? If so what are the possible setbacks and loop holes to look out for.

 

thanks

Onyx

  • Onyx,

    Hypothetically, yes, this is possible.  But in reality, it is a very tricky layout exercise to create a bidirectional I/O using a discrete equalizer and cable driver (such as the LMH0344 and LMH0302).  Performance is inherently compromised when combining the two functions.  There are several concerns:

    1. Return Loss:  The biggest problem is meeting the SMPTE return loss specification with good margin.  You have to compensate for 2x the circuit capacitance.  This requires a complex return loss network that must be tweaked and fine tuned – a time consuming and costly process likely requiring several board spins.  By integrating this network inside the LMH0387, we are able to optimize the return loss and provide 7-8 dB of margin to the SMPTE 424M specification.  This allows extremely easy board layout for customers – all that is required between the BNC and the LMH0387 is the AC-coupling capacitor, and its placement is not critical.  This saves on BOM costs and conserves board space.
    2. Cable Driver Output Eye:  You now have a complex network and added capacitance on your cable driver output, and this will negatively affect the cable driver output eye.  This can cause jitter, overshoot, ringing, and too slow (or not matched) rise/fall times.  We’ve taken care of all of this inside the LMH0387 and already optimized the cable driver output eye.
    3. Equalizer Performance:  This complex network affects the equalizer performance as well.  This requires a very careful layout to avoid impedance discontinuities and stubs that will severely impact cable length performance, especially at 3 Gbps rates.  Again, we already took care of this inside the LMH0387, and you only see a slight reduction in cable length performance vs. using a dedicated equalizer.

    If you attempt this discrete solution, you will have a very difficult layout exercise as described above.  Once (if) you are happy with the design after your efforts, you will have to support a board design with these many critical passive components.  Assembly issues such as bad solder joints will be a nightmare to troubleshoot.  This is another benefit of using the LMH0387: one device replaces two ICs and many (critical) passives.

    Regards,
    Gary Melchior

  • Thank you Gary. I will relay this information over to the customer.

    Regards

    Onyx