• Main INDEX
  • Monthly INDEX
  • PREV
  • NEXT

    User name R. Michaels

    Log entry time 16:30:57 on June23,2001

    Entry number 64291

    keyword=re: "mysteries" of qasy

    Re halog 64196.
    First of all, the automatic "apar" entries seen by chkfeedback
    and used by autofeedback only use 50,000 events (27.8 min),
    and as I've checked several times but not the run in question,
    if you analyze the same number of events by hand you'll see
    identical results. Maybe I'll increase this to 1 hour, but
    in any case it's not a mystery.
    Second, if you compare "qasy" to "apar", note the errors.
    In 1/2 hour, the typical errors are ~8 ppm (apar) and 50 ppm
    (qasy). Do they agree within errors ? Admittedly, sometimes
    not. Finally, there are some funny oscillations in the apar
    histograms if the beam has low noise, as recently.
    Not sure about halog 64196, but what I've seen is 1 bit
    out of 3000 channel (signal) -> 300 ppm structures.
    Solution: turn on "DAC Noise" (It's a trick for smoothing
    this out, which is turned off at present.). Will be done soon.
    Not a problem for present level of accuracy.