HF Electronics

1
HF Electronics

• Status Report
• E. Hazen, J. Rohlf, E. Machado
• Boston University

2
HF Electronics InstallationPreliminary
HF Readout Boxes 24 PMTs each
Cable Route 5m through shielding
Electronics Housed in standard racks (3U card
cages)
3
HF QIE Crate Proposal
Calibration module can occupy unused ROB site (up
to 5 slots - 10cm)
QIE Board 6 Channels
Standard 3U subrack (custom backplane)
4 Boards 1 PMT box
Clock and Control Module (3 boards)
Typically only 3 of 4 ROBs used
ROB 1
ROB 2
ROB 3
ROB 4
CCM
4
PMT Base PCB
Base Supports 8 PMTs with HV distribution
Rear detail showing dual coax (signal plus
reference) for each PMT This is fine for current
testing but should probably change to a
solder termination for production.
5
HF PMT Cabling
12-pair Skew-Clear
Readout Box (24 PMTs)
Patch Panel
Octal PMT Boards
16 (short) coax or other option?
6-channel QIE Boards
6
PMT to QIE connectionPreliminary
Multi-pin shielded connectors
PMT Box
QIE Signal Input
PMT Anode
93O coax
100O pair (or multi-coax)
QIE Reference Input
Two pairs per PMT channel
Coax shield ground at each PMT base
Separate Ground for each PMT
Overall shield connected to QIE ground
7
PMT Box Patch Panel
MDR-36 Connector Site
We plan to make a small PCB which connects 6
PMTs to one multi-pin connector. These can be
retro-fitted to existing PMT boxes for the test
beam. The final box could use a similar PCB or a
larger one which would handle 24 PMTs.
Mini-coax compatible with existing PMT boards
8
HF QIE Digitizer Board
Change to 36 pin MDR Connector (12 pairs, each
with GND)
Electrically identical to HCAL version 3U
Eurocard form factor
Optical Outputs change to Honeywell ST
connector VCSEL (but still near rear)
9
Amphenol Skew-ClearTM Cable

• High-performance, compact
• Available in zero-halogen version (ATLAS, CMS)
• Terminates to convenient Mini-D connectors
  with overall metal shell

10
Skew-Clear Cable Drawing
OD 10mm (28AWG) or 12mm (24AWG)
Will supply zero-halogen version on request
11
MDR Connector Proposed Pinout
12-pair Skew-Clear cable handles 6 QIE
channels Each channel has Pair 1 Signal,
Signal Return, Shield Pair 2 Reference,
Reference Return, Shield Overall Shield connected
to metal connector backshell
Ref
Signal
Shield
Shield
Ref Return
Signal Return
12
PMT Cable QIEPedestal (noise) measurement
HV1100 V
50 O rms 0.74
100 O rms 0.53
As expected, 100 ohm cable improves noise by
sqrt(2)
13
Coax Cable Noise vs Length
14
Skew-Clear Cable Noise vs Length
Scatter due to inconsistent results (probably a
bad connection)
15
HF Noise Test Summary

• PMT Test setup comissioned with QIE/HTR/DCC
  chain.
• Extensive testing done with two cable types,
  lengths from 1-6m
• Dual RG-174 coax with overall braid shield
• Amphenol Skew-Clear(tm) 100O shielded pair
• Noise performance is excellent
• 9500e rms (1.5fC) for 6m skew-clear cable
• 12500e rms (2.0fC) for 6m RG-174 coax cable
• (single pe is 8fC for PMT gain of 5E4)

16
QIE Response to Pulser
4m RG-174
1m RG-174
17
PMT Pulse Examples
PMT pulsed using Nichia blue LED and avalanche
pulser Coupled through 2m RG-174 coax
125 pe
fC
35 pe
QIE Samples (30ns each)
18
Cable Matching Issues

• Problems with impedance matching
• RG-174 coax okay, but overall braid lowers
  impedance --gt reflections
• Paired cable (like skew-clear) is 100 ohm
  differential impedance only... single-ended
  impedance to ground is lower... not well matched
  to high-impedance setting of QIE.
• 93 ohm ribbon coax may be the best choice

19
HF Front-End Status Summary
20
Critical Issues for FE

• Cable / Connector
• Pair vs coax issue still open must settle ASAP
  so new QIE prototype can proceed. Need proper
  PCBs to make meaningful measurements of
  dispersion, crosstalk.
• ROBOX design follows PMT board, patch panel.
• Crate issues
• Backplane design
• Power Cooling
• LVPS
• How best to utilize 'common LV' schemewithout
  compromising performance

21
Web Page for HF Front-End
http//ohm.bu.edu/emachado/HF
New web page with cable/connector documentation.
Will be expanded as more documents become
available...
22
HF Electronics Meeting Proposal

• Start with video conferences
• Follow up with 'workshop' style meeting at Boston
  or Fermilab (late Jan or early Feb)