[Paper Review] A Method to check the Connectivity for the ATLAS TRT Detector

N. Ghodbane, X. Pons+1 more|ArXiv.org|Oct 14, 2005

Particle Detector Development and Performance3 citations

TL;DR

This paper presents a dual-method connectivity check for the ATLAS TRT endcap detector, combining noise scan rate integral analysis and DTMROC accumulation mode testing to identify non-functional readout channels. The method detects 99.6% operational channels, with only 0.4% non-functional due largely to WEB metallization defects.

ABSTRACT

We report on a technique developed to diagnostic for non working readout channels of the Transition Radiation Tracker (TRT), one of the three inner detectors of the ATLAS experiment. From a detailled study, we show that 99.6 % of the readout channels of the TRT endcap-detector are perfectly operational.

Motivation & Objective

• To develop a reliable method for diagnosing non-working readout channels in the ATLAS TRT endcap detector during integration.
• To address the challenge of detecting intermittent or latent failures in readout channels not caught during initial acceptance testing.
• To improve detector quality assurance by identifying faulty channels linked to WEB vias, blind holes, or metallization cracks.
• To ensure high operational efficiency of the TRT by minimizing undetected channel failures before installation.
• To validate the effectiveness of noise scan and accumulation mode techniques in detecting low-signal or disconnected channels.

Proposed method

• Use of noise scan rate integral analysis to detect channels with abnormally low noise rates, identifying those below μ - 3σ as potentially faulty.
• Employment of the DTMROC chip's accumulation mode to measure drift time signal rates, with channels below 0.01 Hz flagged as problematic.
• Application of a pF-meter to measure capacitance at Fujitsu male connectors, detecting values more than 5 pF below average, indicating via or blind hole defects.
• Utilization of a voltmeter's diode check function to inject a small DC current and measure forward voltage drop, identifying abnormal diode behavior at signal paths.
• Correlation of results from multiple techniques—noise scan, accumulation mode, capacitance, and diode check—to isolate faults to specific locations (via, blind hole, or straw).
• Implementation of a MySQL-based database and PHP web interface to store and retrieve test results for systematic analysis and fault localization.

Experimental results

Research questions

• RQ1 How can non-functional readout channels in the ATLAS TRT endcap detector be reliably detected after acceptance testing?
• RQ2 What techniques can effectively identify intermittent or latent faults in WEB vias, blind holes, or signal paths not detected during initial qualification?
• RQ3 To what extent do noise scan rate integral and DTMROC accumulation mode testing improve fault detection sensitivity compared to standard acceptance tests?
• RQ4 What is the contribution of WEB metallization defects to channel failures, and how can they be isolated using electrical measurements?
• RQ5 How effective are combined electrical diagnostics (capacitance, diode check) in localizing faults to specific components (via, blind hole, or straw) in the readout chain?

Key findings

• The connectivity check method identified 487 non-working readout channels in the TRT endcap detector, with 231 disconnected after acceptance tests due to failed performance.
• Of the 231 disconnected channels, 99.6% of the total readout channels (122,880) remained operational, indicating a 99.6% operational rate.
• Only 0.19% of total channels were disconnected after acceptance tests, with the remaining 0.21% of non-functional channels detected during integration.
• The noise scan rate integral method detected 216 of the 256 new problematic channels, while the accumulation mode method detected 252, showing high overlap and complementary sensitivity.
• A total of 256 new faulty channels were found during integration, primarily due to WEB via issues, with 48% of them having undergone prior repairs during acceptance testing.
• A-type WEBs required significantly more repairs than B-type WEBs, and some repaired channels failed again over time, indicating long-term reliability issues with certain WEB designs.