IceRay Integration 2007

1. DOM-MB setup
   1. Rev1 DOR card
   2. External daisy-chained power supplies (98V)
   3. 3 km cable emulator
   4. Custom iceboot (new commands)
      
2. Power setup
   1. External power supply (96V)
   2. 10 Ohm / 10W load resistors on +/-
   3. 5V / 12V generated in Vicor power distribution box
   4. Power consumption: load+ICRR+TRACR, 22W; 1 amplifier module, +10W.
      
3. Amplifier test setup
   1. Channels BAT1, BAT2 connected to surface antenna amplifiers (input terminated)
   2. Channels BAT5-8 connected to RFCM antenna amplifier module (input terminated)
   3. All other channels unconnected (terminated at ICRR input)

Photos of various components and the test setup can be found here.

3. Results

All integration tests to this point have passed, with the following notes / caveats:

• During early integration, rare mainboard hard resets were observed, as if power were interrupted. This behavior could not be reliably reproduced and stopped after the first couple of days.
• Occasionally the ICRR-MB transfer will lose an event byte (approximately 1 out of 2E7 bytes).
• Due to firmware / interface timing issues, ICRR DAC channels must be written in a particular order.
• The calibration pulser has not been integrated / fully tested yet.

The following plots show a DAC threshold scan. The difference between the various input configuration and their gains result in the different threshold curves.

Figure 1: Threshold scans for BAT channels (both polarities).
BAT4, BAT5 are unconnected to amplifiers.

Forced-trigger waveforms were also acquired and examined. Waveform analysis is extremely primitive, with no proper pedestal subtraction or time-base correction performed (see figure 2).

Figure 2: Sample waveform (no calibration other than rough baseline subtraction).

Even so, the band-pass characteristics of the filters on the amplifier chains can be seen by coadding the FFTs of the noise, as shown in figure 3.

Figure 3: Frequency spectra of BAT1 and BAT8 channels under test.
BAT1 surface amplifiers have a pass band of 25-300 MHz, BAT8, 100-400 MHz.

All RFCM and surface amplifier test channels (BAT1-2, and BAT5-8) show the expected bandpass structure. The unconnected BAT3, BAT4 spectra are relatively flat. The 40MHz clock input to BAT9 can be seen clearly in the FFT for that channel in Figure 4:

Figure 4: BAT9 frequency spectrum, showing 40 MHz clock peak.

Here are the scripts used to produce these plots:

• ROOT script for waveform parsing, plotting, and FFT using wf.bin files from IceRay.py
• Gnuplot script for DAC scan plots from headers.txt file from IceRay.py

4. DAQ

The DAQ components consist of a custom iceboot (to interface with the ICRR), driven by a Python script (an extension of Kael's ibidaq library). We have successfully captured ICRR events (noise waveforms digitized by the Labrador ASICs on the ICRR + header information, like scalers and DACs), including a long run of 100K events over 9.5 hours (about 3 Hz). The data readout is not optimized (uncompressed), so the speed could improve there, but we need to add ATWD data and the mainboard header to the events, which will increase the event size.

A mini-HOWTO explaining how to run the proto-DAQ testing software, IceRay.py, can be found here (PDF).

Still to do:

• Add ATWD readout and mainboard headers
• Integrate RapCal / timing calibration
• Add cal pulser functionality
• Optimize readout speed

5. Document Collection

• Integration test spreadsheet (Excel)
• ICRR virtual address definition and event data structure (Excel)