September 13, 1996   08:00 to 22:00
GZ KG EM WB PR SB SA MC TP KS

DAQ
	Much of the shift was devoted to debugging DAQ hardware. The problem
was that some DYCs were receiving gates/stops while others were not for very 
few events (<10 per 10,000 triggers). The non gated/stopped DYCs would report
no data, including the TDC DYC serving as a latch for the trigger signals
which in turn would cause the filter program to report an error for that event.
Occasionally, a DYC would not be readout at all for the event and would lead 
to a mis-match of synchronization of the two DYC streams. After eliminating 
many modules and replacing others, it was seen that the gatemaster live gate 
outputs were not all firing for each trigger (connected to scalers, the counts 
per live gate output would not be the same). The problem was only present for 
interaction triggers and not for pedestal/pulser triggers.
	Two Physics Section electronics engineers came out to look at the
module. They believed that the interaction trigger width would occasionally be
to small, <8ns, and only some of the gatemaster saw these small pulses. The
MLUs are to reproduce the strobe as output. However, the 10ns strobe provided
from the neutral logic at the output side of the MLU was seen to be fluctuating
with some of the output pulses being too small. The strobe width was increased
to 15ns resulting in a wider MLU output signal to the gatemaster. The problem
was not as frequent; more study will be needed as well as a re-look at the 
MLUs.
	To ensure that we were no longer affected, only one live gate output of
the gatemaster was used and a series of discriminator channels were used to
distribute the gates/stops; a few ns was added to the arrival of the 
gates/stops. Data was written to two disk files (~80,000 events in each) and
one tape of (~60,000 events). There was NO pedestal subtraction. Everyone is
invited to look at the data files: run92.dat and run102.dat in the directory
/rta1 on the back end/rapid turn around systems and /dump1 on the front end.

DETECTORS
        Some time was spent in investigating noise problems on the inner 
tracking detectors. We traced back to excessive heating (both from silicon 
barrel and straw chambers front-end electronics) most of the problems. Access 
time will be needed in order to assess the performance of the existing 
air-blower cooling system. 
        Some online plots from hodoscopes, Cherenkov and SCIFI detectors 
were collected, showing sensible results. We experienced difficulty in 
printing out these plots from the monitor program.  
	Some jet off time was dedicated for measuring noise in the luminosity
monitor.