Oct 20, 2011 SiPM Electrical and Cooling

Bias Stabilization and Temperature Compensation Review

1. Presentation GlueX-doc-1853

Tentative Agenda

1. Overview
2. Objectives:
   • Review of the latest device data schematic and options
   • Agree on the bias range (external voltage nominal values)
   • Reach consensus on the resistor set (number of trim resistors)
   • Discuss how to deal with increased leakage current
   • Review Monte-Carlo modeling for component variation
   • Review proposed trim procedure
3. Analytical Results
   • Passive Bias Control Circuit
   • Interactions
4. Temperature Testing
   • Thermal test results
5. Sensitivities
   • Resistor Value
   • Voltage Drift
6. Variation Analysis
   • Monte-Carlo of component tolerance
7. Proposed Design
   • Tradeoff in Trim Resistors -vs- Number of VSUPPLY Values
   • Trim Procedure
8. Discussion

Minutes

Attending: Carl, Jack, Chris, Fernando, Chuck, Jim, Ivan, Yi, Elton, Andrei and George Lolos (phone connection)

1. Jack went over his slide presentations. Below are some notes on the discussion
2. Fernando: 5 supply voltages match the 3-3-4 configuration, and 4 supply voltages match the 1-2-3-4 configuration
   • Jim will check the connector to verify that sufficient number of pins have been allocated for supply voltages
3. The board was designed for a nominal leakage current of 10 micro A. [Note: production units measure about 2.5 microA]. Over the lifetime of the SiPM, the dark current may increase to 30 to 40 microA.
4. Impact of leakage current was discussed. For a fixed supply voltage, it has several consequences.
   • bias voltage to SiPM is changed. For a 40 microA change in leakage current it corresponds to about 240 mv.
   • It modifies the Temperature Coefficient of Voltage (TVC ) by 0.13mV/K microA. or about 5 mV/K for 40 microA relative to the nominal 56 mV/K (10% change over lifetime)
   • Most significantly, any variations from sensor to sensor of the previous effects will lead to gain spreads as the leakage increases.
   • This will have operational impacts and may require readjusting resistor taps at some point.
   • Reducing the resistors in the chain will reduce the sensitivity but increase power consumption. Fernando proposed to size the board for larger physical resistors (higher power, lower values) to allow flexibility in choosing components at a later time. It was suggested that the resistors values be halved. This would increase the power consumption from 27 mW to 54 mW per SiPM (from 1.1. W to 2.2 W per wedge).
5. Changing the resistor settings is not a trivial task because they are located on the bias side of the board facing the cooling plate. Therefore all connections to the sensors must be disconnected before there is access to this side. It will likely require removal of readout from detector and adjustment on the work bench.
6. Summary
   • Use 4 supply voltages per wedge for the 1-2-3-4 configuration (5 for the 3-3-4 configuration). This would result in about 7 different supply voltages to the bcal as a whole.
   • Use sets of 4 trim resistors
   • Failure scenarios should be thought through as one selects which SiPMs are assigned to particular wedges to minimize the impact of repairs and adjustment during degradation (due to radiation or other conditions)
   • Proceed with fabrication of 2 prototypes for complete testing of 2 wedges. After bench testing, these may be used for instrumenting the mini-Bcal module for beam tests in Hall B.