• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 33 subscribers
  • Views 350 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.

DS90LV049: DS90LV049 Noise Reduction

Jacob Powell
Prodigy 80 points
Part Number: DS90LV049

Hi,

I'm using the DS90LV049 to send a SPI bus across a ribbon cable to a remote display. I have the DS90LV049 chip on both boards to send and receive the SPI. I'm using a 3M 3365/20 ribbon cable with 100ohm terminating resistors. I'm testing with a noisy AC environment and it is causing issues for the data being sent to the display. When the cable is short ~5" the noise is not a big problem but when I go to a 12" cable then the display starts to get bad data. The schematics are attached, and I added a 0.001uF cap to the power supply pins of the chips which helped reduce the noise for the short cable, but it did not help with the longer cable. The 50ohm pull downs are DNI since our processor could not drive that load. We are running the SPI at 8MHz which works fine in a non-noisy environment for distances up to 10ft. Any suggestions on how to mitigate the noise coming in through the cable? Do I need to add more caps to filter the power pins? I read through the "How to Terminate LVDS Connections" TI app note and saw Figure 2:

Would this termination help get some of the noise out?

  • 0
    Guru 243820 points

    Power/ground lines between the signals help against crosstalk; you are already doing this. The best protection against external noise would be twisted pairs, and shielding.

    Can you show oscilloscope traces of the noise? Is it common-mode noise, which is what split termination would help against?

  • 0 in reply to Clemens Ladisch
    Prodigy 80 points

    Here are some captures from the scope. The scope is probing SCK+ and SCK- on the display side.

    Scope reading with Short Cable:

    Scope reading with Long Cable:

  • 0 in reply to Jacob Powell
    Guru 243820 points

    Quite a lot of ringing, which generates differential-mode noise.

    Is the characteristic impedance of the cable really 100 Ω? Try adjusting the termination resistors.

  • 0 in reply to Clemens Ladisch
    Prodigy 80 points

    Thanks Clemens,

    I changed the terminating resistors to 150ohm to match the "balanced" impedance of the cable:

    The ringing did go down, but there still seems to be a little noise on the line:

    The display is still getting some bad data, but I haven't ruled out that the noise could be coming in through the +3.3 and GND wires. The +3.3 is oscillating on the display board some so the data getting corrupted after it reaches the board. (+3.3 is green trace)

  • 0 in reply to Jacob Powell
    Guru 243820 points

    The noise that I can see is caused by the signal edges.

    Try other nearby resistor values. (If the characteristic impedance changes between the cable and PCB traces, it might not be possible to get rid of all reflections.)