Title: SPD ASIC design status'
1SPD ASIC design status.
- Present Requirements.
- The 3V3 prototype.
- Preliminary test results.
- Key points to be analysed.
- Future work.
S. Bota, A. Diéguez, D. Gascón - October 2002
Barcelona
2I. Requirements (Energy measurement)
Energy deposition in a SPD cell (electrons and
photons).
- Random signal shape (20-30 phe/MIP)
- Shaping methods discarded
Single event MIP signal
Integrate
BW?100 MHz
3I. Requirements (Timing)
Average MIP signal
- Only about 80 of signal in 25 ns
- No dead time on integration
?d ?12 ns
- Dual channel synchronous system
- Pile-up correction
Normalized integral of Cosmic ray signals in main
and secondary period
4I. Requirements (Other)
- PMT gain limited by aging (DC current) ?
100fC/MIP in central cells (HV650V). - Resolution ? 0.05 MIP (ltlt resolution given by
photostatistics). - 5-10 MIP range to perform tail correction.
- Bandwidth gt 100MHz.
- Linearity error lt 5.
- Robust to temperature variations (band gap
reference). - Small cavity ? Power consumption lt 1W.
5II. The 3V3 prototype
6II. The 3V3 prototype (Input stage Preamplifier)
7II. The 3V3 prototype (Input stage Integrator)
8II. The 3V3 prototype (Input stage OpAmp)
9II. The 3V3 prototype (Input stage Noise and
Offset)
Transfer function of a system that integrates for
?T?
10II. The 3V3 prototype (Pile-up compensation
variable Gm)
11II. The 3V3 prototype (Pile-up compensation
Track Hold)
12II. The 3V3 prototype (Comparison input stage)
13II. The 3V3 prototype (Comparison latched
comparator)
14II. The 3V3 prototype (DAC)
7 bits floating DAC 1b for sign 6b for
modulus R-2R architecture
15III. Preliminary Tests Results
- PRELIMINARY RESULTS (automatic set up in
progress...) - Supply voltage ?1.65 V
- Consumption 25 mA (19 mA channel 6 mA for
buffers). - Pile-up correction control VbiasH1.65V
VbiasL1.45V. - Threshold dispersion (OZE at integrator output)
65 mV r.m.s (for 10 samples) .
16III. Preliminary Tests Results
- Noise lt 2 mV r.m.s
- Synchronous noise effects have been observed
- Systematic test under different conditions
(Threshold input, Pile-up control ....) when
automatic set-up will be ready.
17III. Preliminary Tests Results
- Gain is about 3 times the gain of RUN2-3.
- Range for signal path is gt?1V (range for
compensation to be studied).
18III. Preliminary Tests Results
- Linearity error of the generator has been
measured calibration is needed.
19IV. Key points to be analysed
- Confirmation of offset measurements (20 chips
will be packaged). - Study of pick-up interference coming from
switching on offset and noise. - Sensitive points signal input, threshold input,
pile-up compensation control... - Dependence on board design...
- Study of pile-up compensation linearity
(Temperature!). - Matching of compensation factor (same control for
different channels).
20IV. Future work.
- Complete test of RUN4 (channel and individual
blocks). - Prepare radiation tests for beginning 2002
- EVOLUTION OF FRONT END
- 300fC/MIP? 100 fC/MIP??
- Study how to increase the gain (if needed) ?
offset cancellation is needed.
21IV. Future work.
- If the gain is not increased do we need to
cancel the offset? - The dynamic range will not be decreased because
we can use the positive quadrant. - 3?offset lt 500mV (to use the linearest region) or
even 3? lt 1V. - The DAC has a range of 320mV for 5 mV resolution
(LSB). Therefore