This page outlines some of the efforts underway towards evaluating the performance of an array of NaI(Tl) detectors for use as a CEvNS detector at the SNS

Slides for phone meeting, 2015-05-14

This discussion presently focuses on low-level details concerning, largely, the DAQ configurations used to collect different data sets.

Signals are recorded with a Struck 3316 digitizer. Manual available here

Binary files produced by NGMdaq are converted to ROOT files, with a format described in documentation available here

Detector numbering: 8 detectors are stacked in 3 layers; the bottom 2 layers have 3 detectors each, while the top has only 2. Counting from 0 at the top, left-most detector, the detector number increases from left-to-right across each row, and then from the top row to the bottom. This gives the central detector number 3.

A comment on the “accumulators” on the 3316

In addition to the ability to digitize waveforms, the 3316 allows for the determination of integrals of up-to-8 user-defined subregions of a waveform, referred to as the accumulators. The values of these integrals for each trigger are present in the data read off of the digitizer. The boundaries of these accumulators, in terms of samples relative to the start of the “active gate window”, must be predefined at the start of acquisition. For events where signals will occupy a well-defined region of the waveform relative to the channel trigger, these accumulator values can dramatically reduce the amount of data produced by the digitizer. Some information is, of course, lost when compared to waveform recording.

For internally-triggered runs, the accumulators were configured in such a way as to provide a quick, disk-space-efficient way to extract the spectra from detectors. Depending on the run, for each event the “signal” would generally be contained within accumulator 1 or 2, while the “baseline” would be represented by accumulator 0; these regions were of the same length (340 samples = 1360 ns), which will not necessarily fully contain NaI(Tl) signals, and the signal region was positioned so that it included only a few samples prior to the rise of the signal in the waveform.

In numerous cases, the accumulators can be used to produce very quick spectra, but it is important to consider potential drawbacks.

The 6 remaining accumulators are configured in various ways in the data. Little (i.e. no) effort has been spent in exploring their utility in the collected data.


Dropbox link to data

Descriptions of these data sets can be found below, along with information about all of the data sets available.


Internally triggered data

2015-04-23 - Datestamp 20150423011214

This run, very approximately 18 hours long, was intended to get a reasonably high-statistics background spectrum. To keep the file size manageable, the accumulators are relied upon to reconstruct the energy; waveforms are not recorded. Detectors 0-7 are in channels 0, 2, 4, 6, .., 14.

Detector biases (+ V)

0 - 1670

1 - 1550

2 - 1640

3 - 1700

4 - 1700

5 - 1575

6 - 1800

The channels are triggered solely by the internally generated triggers. Peaking time = 24, gap time = 2, threshold = 150.

Accumulator 1 provides an integral of the "signal" region, while accumulator 0 provides a background sample of the same length.

noWaveforms_allAccum_24hours - Datestamp 20150410220624

This run was intended to provide some insight into the backgrounds present with these crystals. All channels were internally triggered with low threshold settings. Accumulator 1 represented signal region, accumulator 0 represented baseline.

Identification of coincident events is possible through timestamp comparisons.

Detector biases (+ V)

0 - 1670

1 - 1500

2 - 1640

3 - 1580

4 - 1700

5 - 1520

6 - 1800

7 - 1750

Channels 0, 2, 4, 6, 8, 10, 12, 14: detectors 0-7, respectively. Triggered internally and independently.

waveforms_allAccum_5minutes - datestamp 20150410215822

Same settings as run with datestamp 20150410220624 (above), but with waveforms saved and a short duration. The intent is to allow comparison of waveform-based analysis approaches with accumulator-based approaches and to build familiarity with the latter.

Additional runs need to be copied to the dropbox. As this is done, I will update this text. Some runs are likely irrelevant.

Dual-triggered data

These runs have events which are triggered through some combination of an external 60-Hz pulser and the internal trigger logic on the digitizer: in some cases, the trigger is an OR of these two sources; in others, the trigger is an AND; finally, since variety keeps people interested, in still other runs, different channels are triggered in different ways.

The spirit of the combination of asynchronous (i.e. “internally triggered”) and solicited triggers is to enable an inspection of solicited-trigger backgrounds (as could be obtained by trigger on the 60-Hz timing signal of the SNS) with additional information related to events occurring in the crystals outside of these time windows. The additional information, gleaned from the asynchronous-trigger events, could be used to flag dubious regions of time during which crystals might be prone to see signals related to afterpulsing or other sources.

Numerous runs need to be copied to the Dropbox. As they’re made available, I will update this text with some descriptions.

Runs with datestamp 20150506010810 and later - dual_triggers/tagged-solicited

Internal and external triggers are both recorded for channels corresponding to detectors. Channel 15 is triggered only externally, and records the trigger-out signal for the detector channels; this channel provides an unambiguous flag that an event was caused by a solicited trigger.

Detector biases (+ V)

0 - 1670

1 - 1550

2 - 1640

3 - 1700

4 - 1700

5 - 1575

6 - 1800

7 - 1750

Channel 0, 2, 4, 6, 8, 10, 12, 14: detectors 0 - 7, respectively. Triggered on either internal trigger (parameters: peaking time 24, gap time 2, threshold 20000).

Channel 15: triggered by external trigger. Signal is trigger-out pulse from channels corresponding to detectors 0-7.

Run 20150506021517: overnight run, no waveforms recorded.

Runs with datestamp 20150507074628 and later - dual_triggers/coincidenceWith60Hz_waveforms

This configuration is intended to provide the insight available through waveform digitization of solicited-trigger events, while keeping file size manageable and storing information about “asynchronous” events in the central detector. Channel 8, which does not save waveforms but which does provide internally-triggered events for the central detector, is intended to allow rejection of afterpulsing-contaminated solicited-trigger events.

Detector biases (+ V)

0 - 1670

1 - 1550

2 - 1640

3 - 1700

4 - 1700

5 - 1575

6 - 1800

7 - 1750

Channels 0 - 7: detectors 0 - 7, respectively. Triggered if internal trigger generated within 10 microseconds of 60-Hz external pulser. Internal trigger parameters: peaking time 4, gap time 2, threshold 200 (channel 3, threshold 100). Waveforms recorded.

Channel 8: Copy of central detector; internal trigger, peaking time 24, gap time 2, threshold 20000. No waveforms recorded.

Channel 15: Triggered on edge of external 60-Hz pulser.

Runs categorized as "untagged-solicited"

These runs took data with both internal triggers for individual channels and solicited 60-Hz triggers common to all channels (excepting, potentially, missed events due to narrowly-separated trigger arrival times, or pileup). Unlike the "tagged-solicted" data sets, there is no completely-unambiguous tag identifying solicited-trigger events: a solicited-trigger event should include all active channels with an identical timestamp (+/- 1 clock), but one or more channel could be missed (due to pileup, for instance). Also, there is some very small chance that multiplicity-8 events could be generated by backgrounds. Due to the slight ambiguity in identifying solicited-trigger events, the "untagged-solicted" runs were intended to be improved upon with the "tagged-solicted" approach, where an additional channel is activated on the digitizer and triggered exclusively by the external trigger, providing a timestamp for all solicited triggers against which detector channels can be compared. Despite potential drawbacks, the untagged-solicted runs do afford high statistics.

For the internal triggers, the thresholds were intended to be not-exceptionally low: the idea is that the asynchronous internally triggered events could be used to keep track of high-amplitude events occuring soon before solicited-trigger events, and thereby provide some ability to reject events contaminated by afterpulsing. The peaking time was 24, gap time 2, threshold was 5000 for runs with datestamps 20150423221421 and 20150424192220 (these runs can be chained together - no settings changed) and the threshold was increased to 20000 for run 20150424195132.

Accumulator region 2 was intended to serve as the "signal" region, while accumulator 0 provided baseline. No waveforms are recorded. Detectors 0-7 are in channels 0, 2, 4, .., 14.

For all of these runs, the detector bias voltages were unchanged.

Detector biases (+ V)

0 - 1670

1 - 1550

2 - 1640

3 - 1700

4 - 1700

5 - 1575

6 - 1800

NaI(Tl) array testing (last edited 2015-05-14 17:57:06 by GraysonRich)