• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 15 subscribers
  • Views 509 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.

Using multiple THS788 Time Measurement Unit in parallel

Markus Rau
Prodigy 60 points
Other Parts Discussed in Thread: THS788

Hello,

i plan to use two THS788 chips in parallel, since i require 8 event channels. Both chips would share one Sync input via a fanout. My measurements will use relatively few Sync pulses (approximately one per second), but much higher rates on the event channels (>100k per second).


I'm afraid that the simple scheme of using a fanout to split the Sync pulse is not good enough to get the best possible performance. Due to the fanout's jitter and and the finite accuracy of the THS788, one chip might measure the Sync pulse 13 ps earlier or later compared to the other chip. Then all measurements until the next Sync pulse will be systematically off by +- 13ps.

Is there a possibility to avoid this problem? If one could access the current Sync time stamps it would be easy to correct for this problem, by checking that the lowest 6 bits of the two sync time stamps always have a fixed difference.

Regards,

Markus

  • 0
    TI__Intellectual 2845 points

    Hi Markus,

    There is not a way to read the internal Sync value on a part. You could use the calibration registers to make the differences in values as close as possible, but any sort of difference due to jitter on the fanout will be hard to adjust for on the fly. 

  • 0 in reply to Matthew Sullivan
    Prodigy 60 points

    Hi Matt,

    Thanks for your reply!

    I have one more question: Can one use the chip without ever sending a Sync signal? Would the chip output time stamps using a sync timestamp of "0"? This would be usefull for one of my applications where i don't have "start" signal and only look for very short range (<1ns) correlations between the event channels.

    Regards,

    Markus

  • 0 in reply to Markus Rau
    TI__Intellectual 2845 points

    Markus,

    It's hard to say what your results would look like in that situation. Your sync value is in reference to the free running counter, and the math done to calculate the difference between and Event and Sync is in reference to that as well. 

    When the part is initially start up, the value in the ALU for Sync should read "0", but that is in reference to nothing. The size of any events calculated will make no sense compared to your "0" sync value because the sync value has no relation to the counter whatsoever.