Proposed Hybrid / Module Electrical Test Procedures
This is a PROPOSAL
for guidelines to be advertised by the
TestDAQ group for use in the immediate future, not necessarily longer-term or production testing
: please comment
-
Why use common chip settings & optimisation rules ?
- Comparison of results different modules and/or test systems
- Help new groups
- Design test systems & software
-
Optimisation
- Operating modes (throughout):
- Edge Detection ON (more aggressive)
- Compression mode: Anyhit ( or 01X - shouldn't matter with edge detection on)
- Strobe Delay Optimisation & Delay Calibration:
- At nominal FE Bias and Shaper settings (e.g., ABCD1: FEB 202uA, FESh 18uA)
- Strobe delay scan (Occupancy vs Strobe delay) at Qi 4fC, threshold 200mV
- -> optimum setting; delay calibration (PLOT)
- FE Bias & Shaper Optimisation:
- 9 points at and around (+/- 1 unit) nominal FE Bias & Shaper
(more points if necessary)
- "Quick" threshold scans at two Qi settings: 2.5, 3.5fC (Vcal 25, 35mV)
100-500mV, 25 steps, at least 100 events /step
- Extract VT50 and output noise from S-curve fits for each channel (PLOT)
- Simple linear fit -> Gain slope and offset for each channel (PLOT)
- Use Gain and Output noise at 2.5fC to extract Input Noise (ENC) (PLOT)
- Histogram offset spread and ENC for each chip (PLOT)
- Choose optimal FE Bias and Shaper setting for each chip
-
Calibration & Noise Measurement
- At optimal Strobe delay, FE Bias & Shaper for each chip
- Detailed threshold scans at many Qi settings: (1.5),2.0,2.5,3.0,3.5,4.0,(4.5,5.0)
100 to 500 mV threshold in 25 steps, at least 100 events /step (more if faster system)
- Extract VT50 and output noise from S-curve fits for each channel (PLOT)
- Linear fit of "selected linear region" -> Gain extrapolated to 1.0fC for each channel (PLOT)
- Use Gain and output noise (at lowest Qi with reasonable S-curves) to extract Input Noise (ENC) (PLOT)
- Why use common chip settings & optimisation rules ?
- Comparison of results different modules and/or test systems
- Help new groups
- Design test systems & software
-
Optimisation
- Operating modes (throughout):
- Edge Detection ON (more aggressive)
- Compression mode: Anyhit ( or 01X - shouldn't matter with edge detection on)
- Strobe Delay Optimisation & Delay Calibration:
- At nominal FE Bias and Shaper settings (e.g., ABCD1: FEB 202uA, FESh 18uA)
- Strobe delay scan (Occupancy vs Strobe delay) at Qi 4fC, threshold 200mV
- -> optimum setting; delay calibration (PLOT)
- FE Bias & Shaper Optimisation:
- 9 points at and around (+/- 1 unit) nominal FE Bias & Shaper
(more points if necessary)
- "Quick" threshold scans at two Qi settings: 2.5, 3.5fC (Vcal 25, 35mV)
100-500mV, 25 steps, at least 100 events /step
- Extract VT50 and output noise from S-curve fits for each channel (PLOT)
- Simple linear fit -> Gain slope and offset for each channel (PLOT)
- Use Gain and Output noise at 2.5fC to extract Input Noise (ENC) (PLOT)
- Histogram offset spread and ENC for each chip (PLOT)
- Choose optimal FE Bias and Shaper setting for each chip
-
Calibration & Noise Measurement
- At optimal Strobe delay, FE Bias & Shaper for each chip
- Detailed threshold scans at many Qi settings: (1.5),2.0,2.5,3.0,3.5,4.0,(4.5,5.0)
100 to 500 mV threshold in 25 steps, at least 100 events /step (more if faster system)
- Extract VT50 and output noise from S-curve fits for each channel (PLOT)
- Linear fit of "selected linear region" -> Gain extrapolated to 1.0fC for each channel (PLOT)
- Use Gain and output noise (at lowest Qi with reasonable S-curves) to extract Input Noise (ENC) (PLOT)
Why use common chip settings & optimisation rules ?
- Comparison of results different modules and/or test systems
- Help new groups
- Design test systems & software
Optimisation
- Operating modes (throughout):
- Edge Detection ON (more aggressive)
- Compression mode: Anyhit ( or 01X - shouldn't matter with edge detection on)
- Strobe Delay Optimisation & Delay Calibration:
- At nominal FE Bias and Shaper settings (e.g., ABCD1: FEB 202uA, FESh 18uA)
- Strobe delay scan (Occupancy vs Strobe delay) at Qi 4fC, threshold 200mV
- -> optimum setting; delay calibration (PLOT)
- FE Bias & Shaper Optimisation:
- 9 points at and around (+/- 1 unit) nominal FE Bias & Shaper
(more points if necessary)
- "Quick" threshold scans at two Qi settings: 2.5, 3.5fC (Vcal 25, 35mV)
100-500mV, 25 steps, at least 100 events /step
- Extract VT50 and output noise from S-curve fits for each channel (PLOT)
- Simple linear fit -> Gain slope and offset for each channel (PLOT)
- Use Gain and Output noise at 2.5fC to extract Input Noise (ENC) (PLOT)
- Histogram offset spread and ENC for each chip (PLOT)
- Choose optimal FE Bias and Shaper setting for each chip
-
Calibration & Noise Measurement
- At optimal Strobe delay, FE Bias & Shaper for each chip
- Detailed threshold scans at many Qi settings: (1.5),2.0,2.5,3.0,3.5,4.0,(4.5,5.0)
100 to 500 mV threshold in 25 steps, at least 100 events /step (more if faster system)
- Extract VT50 and output noise from S-curve fits for each channel (PLOT)
- Linear fit of "selected linear region" -> Gain extrapolated to 1.0fC for each channel (PLOT)
- Use Gain and output noise (at lowest Qi with reasonable S-curves) to extract Input Noise (ENC) (PLOT)
- Edge Detection ON (more aggressive)
- Compression mode: Anyhit ( or 01X - shouldn't matter with edge detection on)
- At nominal FE Bias and Shaper settings (e.g., ABCD1: FEB 202uA, FESh 18uA)
- Strobe delay scan (Occupancy vs Strobe delay) at Qi 4fC, threshold 200mV
- -> optimum setting; delay calibration (PLOT)
- 9 points at and around (+/- 1 unit) nominal FE Bias & Shaper
(more points if necessary) - "Quick" threshold scans at two Qi settings: 2.5, 3.5fC (Vcal 25, 35mV)
100-500mV, 25 steps, at least 100 events /step - Extract VT50 and output noise from S-curve fits for each channel (PLOT)
- Simple linear fit -> Gain slope and offset for each channel (PLOT)
- Use Gain and Output noise at 2.5fC to extract Input Noise (ENC) (PLOT)
- Histogram offset spread and ENC for each chip (PLOT)
- Choose optimal FE Bias and Shaper setting for each chip
Calibration & Noise Measurement
- At optimal Strobe delay, FE Bias & Shaper for each chip
- Detailed threshold scans at many Qi settings: (1.5),2.0,2.5,3.0,3.5,4.0,(4.5,5.0)
100 to 500 mV threshold in 25 steps, at least 100 events /step (more if faster system)
- Extract VT50 and output noise from S-curve fits for each channel (PLOT)
- Linear fit of "selected linear region" -> Gain extrapolated to 1.0fC for each channel (PLOT)
- Use Gain and output noise (at lowest Qi with reasonable S-curves) to extract Input Noise (ENC) (PLOT)
100 to 500 mV threshold in 25 steps, at least 100 events /step (more if faster system)