NEXT
User name gilman,korsch,averett
Log entry time 12:59:02 on July27,2001
Entry number 67955
keyword=Phillips time asymmetry (partial) explanation
We (TA, RG, WK) go to the spectrometers to look into the 10^-3 time
asymmetry in the Phillips. We think we understand it on the left arm now.
In the counting house, when you look at the scope for the pseudorandom mode glitches, you see them predominantly in one phase:
h+ glitches TRUE for ~50 ns
h- glitches FALSE for ~25 ns
It seems very asymmetric. (Today anyway!)
From past studies, these glitches happen perhaps every several seconds,
in the middle of a 66 = 2*33 ms pulse, when the helicity does not change.
When the signals get down to the left arm, we see:
h+ glitches TRUE for ~50 ns
h- has no glitch usually, perhaps 20% of time there is 0.3-0.5 V ~5-10 ns glitch
Estimated rate is 11/4 minutes ==> 1 every 20 seconds.
In this mode, when a glitch happens in the left arm,
The h- gets cleared by the h+, the h+ 500-us delay starts, but does not get
latched due to h- being TRUE. Since there is almost no signal on h-,
there is (usually) no h- 500 us delay and thus no new latch. You lose
33 ms of h-, every 20 seconds, and 0.033/20 ~ 1.5e-03.
When the signals get to the right arm, we see:
h+ glitches TRUE for ~10 ns
h- glitches FALSE for ~25 ns
Estimated rate was 10 / 2 min ==> 1 every 12 seconds. Probably at this
level consistent with left arm. In this case, the h- signal going true again
should reestablish the latched signal, and there is no time asymmetry.
I am tempted to suggest the problem is fully understood. The problem is that
the different lines from counting house to spectrometers disperse signals differently,
resulting in slightly different behaviors. Possible exceptions to this being a full
understanding / things to do:
1) I thought we saw glitches in both phases before, not just the one mentioned.
I think the other phase was rarer; perhaps we got "lucky" today.
2) We saw a visual scaler asymmetry with Phillips on right arm. Perhaps the
h+ glitch was too narrow to dependably trigger it, and the asymmetry was
reversed, but we did not notice it? I *think* we tok a test run several days ago,
maybe someone can go back and check the phase of the asymmetry.
3) We should probably test the triggering of the Phillips / LeCroy boxes
to be sure of their response to short signals.