In October of 2000, we performed a test of the prototype Front-End Electronics (FEE) boards at the KamLAND site in Kamioke, Japan. This test served two primary purposes. The first was to establish the running environment at the experimental site, in particular to measure the level of electronic noise induced in the mine. The second goal was to demonstrate the capability of the FEE board to record single photon signals from the KamLAND photomultiplier tubes (PMT's).

The key result of the on-site test is that the noise environment in the mine is very good. We measured noise rates in the range of 24-30 kHz with a threshold of 0.6 mV. This voltage corresponded to approximately 1/5 of the peak voltage of a single photoelectron pulse from the PMT. With the exception of the overhead crane, which not be run during normal data collection, we found no large sources of electronic noise. In particular, the MACRO electronics did not induce substantial noise to the ATWD FEE boards. A challenge for the KamLAND collaboration is to complete the detector construction and instrumentation without increasing the level of electronic noise.

We also made several important observations about the running environment during the electronics test. We found that proper grounding was essential in order to avoid ground-loop oscillations on the FEE signal inputs. In particular, we found that we can not allow the PMT signal cables to float, a configuration that we had considered. We also noted the importance of shielding high-voltage cable extension connectors, as the high-voltage ground for each PMT is isolated from the common experimental ground.

Finally, we discovered a discrepancy in the calibrations of PMT high voltage to gain. These calibrations were done before the phototubes were installed in the detector, but they appear to no longer be valid. We found that the tubes exhibited higher gains than expected for the high voltages we applied. As a result, the PMT gains will have to be recalibrated in situ. This discovery has implications for the high-voltage distribution system, which must be able to supply the correct voltage to each PMT to ensure a uniform distribution of PMT gains.

The on-site test also provided an opportunity to demonstrate the capabilities of the FEE boards developed at LBNL. Figure 1 shows a sample single-photoelectron pulse as captured by the FEE electronics. A multiple pulse waveform, which illustrates the ability of the system to resolve multiple pulses in close time sequence, is shown in the main FEE Annual Report. The single photoelectron pulse height distribution is shown in Figure 2. The on-site data is also useful for software development. We are currently using the on-site data to help develop algorithms to convert the raw waveform data to pulse amplitudes and arrival times.

Figure 1. A sample single-photoelectron-pulse waveform recorded during the on-site KamLAND electronics test. Both raw and pedestal-subtracted waveforms are shown, with the pedestal in dashed red, and a flat baseline in dashed green. (The horizontal scale is about 2.3 ns per sample, while the vertical is approximately 0.8 mV per count.)

Figure 2. A sample pulseheight distribution from the on-site electronics test. This histogram shows the pedestal-subtracted area of single pulses, most of which are due to single photoelectrons. These data were collected with a threshold of approximately one-fifth the pulseheight of a single photoelectron.