• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 33 subscribers
  • Views 1602 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.

SN75LVDS387: LVDS for I2S signal transmission

Stephan P
Prodigy 220 points
Part Number: SN75LVDS387

Hello TI forum, 

We are currently working on a digital audio application (I2S - 1xSDATA In / 16xSDATA Out) for which it is necessary to transmit the digital audio signals over a distance of about 5m.

For this we are considering to use LVDS for the transmission from board 1 to board 2 as shown in the attached block diagram. Board 1 provides the clocks (MCLK, BCLK, FCLK) and 1x SDATA which is then transmitted via LVDS to the MCU on board 2. The MCU on board 2 is a slave to these clocks and provides 16x SDATA which is transmitted back to board 1 via LVDS. The input interface of the host on board 1 is a slave to these clocks coming from board two to account for the delay in the system.

The clocks in the system are: MCLK: 36.864MHz, BCLK: 12.288MHz, FCLK: 192kHz and we are using 3.3V signal levels

 Could you please help me out with following questions:

  1. Do you see any issues with using a Samtec EQCD or HQCD cable (50 Ohms single ended) for the transmission? The two lines next to each other would be used for an LVDS pair (similar as mentioned in your application note SLAA284 but without the extra GND line between the LVDS pairs).
  2. It is planned to connect board 2 directly to a backplane. The connection cable would then go from board 1 to the backplane and then to board 2. Is this extra connector an issue in this application?
  3. Would the SN75LVDS series be an appropriate choice for this application or would you suggest other parts?
  4. Is there any benefit from using 3x 8channel LVDS devices instead of 1x16 and 1x4 channel devices as shown by the red blocks in the block diagram (skew or other issues)?

Thanks a lot,

StephanClocking scheme.pptx

  • 0
    TI__Mastermind 35130 points

    Hi Stephan,

    1. That should be fine, as the two 50-ohm single-ended conductors results in 100-ohm differential. 

    2. It could be. The extra connectors as well as the backplane will introduce impedance mismatches to the signal, which could cause reflections and result in signal integrity issues. 

    3. It's an appropriate choice if you need 16 channels.

    4. Skew, as you mentioned, is the main benefit of using 1x16 and 1x4 channel devices as opposed to 3x8 channel. To minimize skew, the less parts you use the better.

    Regards,

    I.K. 

  • 0 in reply to Ikechukwu Anyiam
    Prodigy 220 points

    Hi I.K.,

    Thank you for your response. After checking the SN65LVDS series it seems to be a better choice compared to the SN75LVDS due to the improved ESD rating, would you agree?

    I have one further question: are there any issues expected when the devices are enabled in case the connection cable is not plugged in? Is it necessary to first check if the cable is connected before enabling the devices?

    Thanks,

    Stephan

  • 0 in reply to Stephan P
    TI__Mastermind 35130 points

    Hi Stephen,

    If ESD rating is a care-about for your design I agree SN65LVDS series is a better choice.

    I believe your question is about hot-plugging, which shouldn't be an issue as long as the device absolute max ratings are not exceeded when plugging in the cable if the device is on. To best avoid damaging the device you should ensure GND/Power are connected before the signals are connected when hot-plugging the cable.

    Regards,

    I.K.