[time-nuts] SR620 binary dump

[Tom Van Baak]

> Using XREL exposes an additional digit of resolution. Set XREL to 1E7 and then you get ascii readings like:
> 
>    1.E-4
>    -6.E-5
>    2.E-4
>    -1.4E-4
>    .0E-6
>    -1.9E-4

I should clarify. Yes, it appears that when using XREL, one can get ~2.7 ps timing resolution/granularity for frequency readings <10 MHz. But when the frequency reading is >10 MHz the timing resolution/granularity is 10 ps.

For example, here are 1000 measurements using XREL 1E7 (columns are occurrence, value):

      1  9999999.999480
      1  9999999.999510
      1  9999999.999560
      1  9999999.999590
      4  9999999.999620
      3  9999999.999640
      9  9999999.999670
      6  9999999.999700
      8  9999999.999720
     22  9999999.999750
     24  9999999.999780
     28  9999999.999810
     50  9999999.999830
     71  9999999.999860
     45  9999999.999890
     73  9999999.999910
    122  9999999.999940
     54  9999999.999970
    294 10000000.000000
    152 10000000.000100
     28 10000000.000200
      2 10000000.000300
      1 10000000.000500

Notice how all the readings less than 10 MHz differ by 20 or 30 uHz (mean 27), but all the readings greater than 10 MHz differ by 100 MHz. I will continue to look into this, but my hunch is that this is a bug in the SR620 firmware; the only solution is using binary mode.

/tvb