The 1996 run period was mainly devoted to the installation and
commissioning of the LKr calorimeter and to the implementation
of the Neutral Trigger in the experiment. Most of
the other components of the NA48 detector were already in their
final configuration and tested in 1995.
A data taking period of 4 days with 
beams at nominal
intensity took place towards the end of the run for physics studies.
Since only 8 
of the LKr electronics was available
in 1996, a strategy had to be adopted in order to measure the intrinsic
performance of the electromagnetic calorimeter and to study
decays with the detector. In a first step,
a calibration of 384 individual 2 
2cm 
instrumented cells
with an electron beam was performed
at different energies (25, 50 and 100GeV/c). In the second step, the entire
calorimeter was read-out by instrumenting "supercells" of 2 8
single cells. In this way, many aspects of the Neutral Trigger could
be tested and physics data could be taken with 
and 
beams.
The LKr calorimeter was successfully operated at 1500V. The maximum
driftime was measured to be 3.5 
s in agreement with measurements
performed in the laboratory. After a first purification of the
liquid krypton, the lifetime of the electrons was found to be greater than
45 s for several weeks without any additional purification.
The temperature differences inside the cryostat were measured to be smaller
than 0.2 
K and were very stable as a function of time.
Figure 5: Energy response of the LKr calorimeter to 50 GeV/c electrons.
The performance of the LKr calorimeter was measured during the
calibration period with an electron beam. Figure 5 shows the energy
deposition
in a shower box of 11 11 single cells for 50GeV/c incident electrons.
The resolution obtained is .69 , in close agreement with prototype results.
The contribution of the noise, dominated by the incoherent component, was
less than 8MeV per cell. The low energy tail in
the distribution is due to losses in the material in
front of the calorimeter. The position resolutions
measured at the same energy are 780 m and 970 m in the x and y
projections
respectively. The time resolution at 100GeV is better than 150ps in the
cell containing the maximum energy deposition of the shower.
Figure 6: Invariant mass distribution of two photons measured
with the LKr calorimeter.
Data taken with the LKr calorimeter fully read out in the "supercell"
configuration have permitted to check the different functionalities
of the Neutral Trigger. In particular, the total energy and peak counting
performed as expected yielding trigger rates slightly above 1kHz. Moreover,
decays
with four clusters in the photon detector were clearly identified
and their rates were found to agree with the predicted ones.
The 
mass resolution obtained from 2-body 
decays
is about 3.5MeV (see figure 6), mainly limited by the
poor granularity of the "supercells" and
their associated electronic noise. This result is compatible with
a mass resolution of about 1MeV that could be reached with
instrumented 
cm single
cells.
Very good performances were also obtained with the magnetic spectrometer
for the 
decays. The four drift chambers worked reliably and provided high detection
efficiency and position resolution at rates close to 1MHz. The kaon invariant
mass resolution obtained off-line from fully reconstructed
events
is 3MeV in both and beams.
Moreover, the good
resolution on the decay vertex position provides a clear separation of
the and sources, allowing a measurement of the tagging
efficiency in the charged decay mode.
The Charged Trigger operated satisfactorily in rejecting by a factor of
about 70 the 3-body charged decays. The on-line reconstructed
kaon mass has a resolution better than 7MeV in
decays and the reconstruction
efficiency of the trigger is close to 98 .
Analysis of both neutral and charged kaon decays is in progress
in order to better understand the performance and acceptance of the NA48
detector, the tagging efficiency and the
contributions from several background sources.