• State Resolved
  • Replies 2 replies
  • Subscribers 33 subscribers
  • Views 32 views
  • Users 0 members are here
Support feedback
Options
Options
Related

DS90LVRA2-Q1: Question on Maximum Skew

Nathan Rabenold
Intellectual 755 points

Part Number: DS90LVRA2-Q1

Hi,

I am using this part on high speed ADC interface signals going into an FPGA (at 2.5V). We are extremely tight on timing with the +/-300psec channel-to-channel skew variation. We aren't as concerned with the actual skew itself (which we can compensate for) but rather the variation in potential skew. We will operate very near 2.5V with temperatures realistically between 0-75C.  So I have two questions:

  1. Given our temperature range, is there anything that would allow us to assume a smaller skew variation? I'm just wondering how likely it is we actually see +/-300psec channel-to-channel.
  2. On a given part, do you have any data on how much skew variation there could be over the lifetime? We are looking at calibrating for the channel-to-channel skew, but this would require some evidence that a given part won't drift over days of runtime to either extreme.
  • +1
    TI__Genius 12796 points

    Hi Nathan, 

    Q1: Yes you can expect to see smaller skew with a smaller temperature range. However, I cannot guarantee how much of a skew reduction you will see since we characterize these devices for their full operating temperature range, in this case -40 to +105C. 

    Q2: For most TI devices, we typically provide ~10 years of operating lifetime with no spec variation outside of the datasheet limits. However, for this device, the assumption is that the ambient temperature is a constant 55C (see below from the device's quality page). In reality, the ambient temperature will fluctuate over the device's operating life (i.e. the mission profile will change). The constant 55C mark is a general approximation of most common mission profiles, but not all. To learn more about this, I recommend looking at the reliability section of our website: https://www.ti.com/support-quality/reliability/reliability-home.html 

    Regards,

    Matt 

  • 0 in reply to Matt Smith
    Intellectual 755 points

    Okay, thanks. I was hoping I could find some justification to assume a worst-case skew skew variance of something less (and concrete). It doesn't sound like we will be able to do that.