Digital signal processing for superheavy element studies D. Miller 1 , K. Miernik 2,3 , D. Ackermann 4 , R. Grzywacz 1,2 , S. Heinz 4 , F.P. Heßberger 4,5 , S. Hofmann 4,6 , J. Maurer 4 , K. Rykaczewski 2 , and H. Tan 7 1 Univ. of Tennessee, Knoxville, USA; 2 ORNL, Oak Ridge, USA; 3 Univ. of Warsaw, Warsaw, Poland; 4 GSI, Darmstadt, Germany; 5 HIM, Mainz, Germany; 6 Goethe-Universit¨ at, Frankfurt, Germany; 7 XIA LLC, Hayward, USA Digital acquisition systems bring a host of benefits to nu- clear physics experiments. Primarily, they provide a plat- form that is highly flexible and compact with the capabil- ity of extracting additional information from the details of the pulse shape. For recent investigations at SHIP in the first part of an experiment aimed at the detection of the su- perheavy element Z = 120 [1], digital signal processing techniques enabled detection of alpha decays as prompt as ∼100 ns after implantation. During 33 days of beamtime in April-May 2011, we de- ployed six Pixie-16 (Rev. D) modules [2] manufactured by XIA LLC in a single 6U compact PCI crate capable of in- strumenting all signals including solid-state detectors and logic signals relevant for SHIP analysis with digital sam- pling at 100 MHz. The digital acquisition system ran in parallel with the traditional analogue setup reproducing all essential spectra. Due to the relatively low rates of inter- esting decay chains, cross-correlation of the two systems was made using the internal 10 ns timestamp combined with an occasional readout containing the internal clock of the data acquisition computer to produce a human readable time within several microseconds. Due to the FIFO buffering of the latest revisions of the Pixie-16, the system operates with effectively zero dead- time allowing the detection of all incident events. Data throughputs as high as 72 MB/s have been separately ob- served in high rate experiments including overhead as- sociated with online analysis. With its simple single- crate/single-computer operation, the system also showed exceptional stability and reliability with no operative issues exhibited for the extent of the experiment. The standard Pixie-16 firmware determines event-by- event timing and energy onboard reducing the burden of offline analysis. Several new features were also developed in the firmware. For short decay times where the signal from the radioactive decay falls on top of the tail of the implantation signal, complete trace information is needed to extract the contribution from the implantation and de- cay components. The newly developed firmware records traces only in the situation where such a pile-up is detected (see Figure 1). For other events, solely the timing and en- ergy information is recorded. For maximum sensitivity to the predicted Z = 120 half-lives [3, 4], we recorded 9.5 μs traces for each piled-up trace. The lower limit of resolution for the pile-up inspector (∼ 100 ns) depends on the con- volution of the Si strip detector drift time with the pream- plifier response rise time and the analogue stage Nyquist filter in the Pixie-16 modules. Recording energy and tim- ing data for all events allows for the continuous monitoring of detector charateristics associated with the implantation 400 500 600 700 800 900 1000 0 1 2 3 4 5 6 7 8 9 Amplitude (ADC units) Time (μs) Figure 1: Trace capture selectively triggered from the pileup inspector with two successive implants followed by a low-energy alpha decay. residues as well as remaining sensitive to longer half-lives. The reduction of data greatly simplifies the offline process- ing and storage. Over the month long run, a total of 397 GB of data (150 GB compressed) was recorded. If traces were recorded for every event, this would become a much more cumbersome 46 TB of raw data to analyse. Several other features were also important for the anal- ysis. Since traces were not captured for every event, it no longer is possible to directly determine which signals ex- ceeded the input range of the 12-bit ADC such as those induced by fission events. These events were thus flagged onboard as “saturated” events. For saturated events, length of time out of ADC range provides an approximation of spontaneous fission pulse height. Total time out of range and several other important runtime statistics were also recorded onboard and occasionally dumped to the data stream. Initial tests were also done using the onboard QDC capabilities of the Pixie-16 to selectively integrate up to eight time intervals for each incoming signal. This al- lows position reconstruction from the information recorded for the position-sensitive silicon detector, though a detailed calibration and parameter optimization are still needed. This work is supported by HI Mainz. References [1] S. Hofmann et al., GSI Scientific Report 2011. [2] http://www.xia.com/DGF_Pixie-16.html, updated 16 Nov. 2011, accessed 16 Dec. 2011. [3] A. Parkhomenko and A. Sobiczewski, Acta Phys. Pol. B 36, 3095 (2005). [4] A. Sobiczewski, Acta Phys. Pol. B 42, 1871 (2011). PHN-NUSTAR-SHE-16 GSI SCIENTIFIC REPORT 2011 220