On Aug.20 there were some problems with the outer detector software (FSCC/DC2 count errors) which didn't seem to recover automatically. The latest OD software changes (automatic Fastbus controller reset) were implemented approximately one week earlier and had performed very well. Since then the "ONLY Anticounter HEADER" problem hadn't caused any run stops, until the 2nd shift of August 20. Coincidentally, the laser system between the center hut and outer hut 1 had been initialized at the beginning of the shift and was running under test conditions. After the laser tests had been stopped the shift members told me they hadn't observed further OD problems.

Thus, it was suspected that there's a correlation between this particular OD failure and the laser system powered up. It was specifically interesting to find out if electrical noises generally would affect the OD DAQ system at all. There had been earlier suspicions about arcing high voltages at the OD paddle cards or PMT connectors. So, the next day I was checking the 120V AC voltages for the OD electronics for spikes or other kind of unusual signals. But the scope measurements (see example, where channel 1 is the voltage measured between earth ground and the "hot" outlet pin and channel 2 the voltage at the "null" pin) didn't show any differences between having the laser running or powered off.

When I finally removed the scope probes from the AC outlets, I observed some somehow regular high-frequency glitches, even though the probe wasn't connected to anything. Thus, the probe tip (standard tektronix high-impedance probe, with 1:10 attenuator) acted like an antenna and was picking up electro-magnetic signals inside the center hut. I couldn't figure out the direction of the noise sources, but when the probe was closer to the floor of the hut the amplitudes became larger. This time the laser system was still powered on, and the colleagues told me they were planning to continue their tests for the evening. It sounded to me it would have been difficult to switch it off for a while and then switch it back on again. So, I didn't bother them and I decided to postpone my measurements for a better opportunity.

This opportunity then came on August 23. While the laser was not powered, and with the exact same scope setup and the open probe lying on the floor of the center hut, there were no noises at all, except typical white (or sort of) noise at around 15...25mV amplitudes.

As soon as the laser system was powered on again, almost identical noise glitches appeared on the screen again as seen 2 days earlier. This time I did some more studies. First, the noise bursts seemed to be repetitive, but changing in amplitude with different trigger threshold settings. The frequencies of the signal inside the burst envelope (about 2-4 usec long) were rougly between 6...8 MHz.

Setting the scope to "peak detect" showed that there were at least three different repeating burst amplitudes overlapping at independent rates. As seen on the picture with 2msec/div scope setting there are ~100mVpp glitches at a rate of approx. 800-900 Hz, and pulses with somewhat higher amplitudes at approx. 200 Hz. At a setting of 20msec/div there's evidence of a 3rd class of bursts at a rate of approx. 10 Hz. The latter one is pretty consistent with the laser pulse rate setting.