PXL Electronics

Status update for HFT TC meeting on October 14, 2010 at BNL

1HFT TC 05/11/2010 - LG

Outline

• CD-2/3 cost and schedule – overview and methodology.

• 4th level walkthrough with changes in scope, major cost changes, schedule drivers

• Risk and contingency• Technical progress

HFT TC 05/11/2010 - LG 2

PXL Cost and Schedule - overviewI will present 1.2.2 and the electronics part of 1.2.4 (Infrastructure)

HFT TC 10/14/2010 - LG 3

1.2.2 Methodology

(re-)Estimate cost – Detailed excel spreadsheet with a breakout by project WBS #.Import rolled up costs into project document categories by WBS #.Assign contingency and risk.•Assign proper dependencies between subsystems•Level effort based on resources available (and/or locate more/different resources).•Integrate all subprojects into one schedule.•Add appropriate milestones.•Match project profile to DOE funding profile by reasonably adjusting the tasks durations and timing while maintaining a reasonable work flow.•Use the detailed excel workbook cost estimate and the project document to track progress and report at rolled up level.

HFT TC 10/14/2010 - LG 4

PXL Cost and Schedule - overview

1.2.2.4.2.1 PXL Final Sensor Prototype 1 Order

1.2.2.4.2.2 PXL Final Sensor Prototype 1 Receive - Milestone

1.2.2.4.2.3 PXL Prototype 1 Test

1.2.2.4.2.1 PXL Final Sensor Prototype 1 Ordermasks and processing To be revisited wrt high res si and other optionswafers (additional) To be revisited wrt high res si and other options

1.2.2.4.2.3 PXL Prototype 1 TestPCB design and LayoutPCB fabricationPCB partsPCB loading (contributed 60hr tech)(iteration) PCB design and Layout(iteration) PCB fabrication(iteration) PCB parts and loadingInterface PCB dicingPXL Sensor bench testing (includes firmware/software)PCB board telescope test design and layoutPCB board telescope test fabricationPCB board telescope parts and loading(iteration) PCB board telescope test design and layout(iteration) PCB board telescope test fabrication(iteration) PCB board telescope parts and loading(iteration) Interface PCBConnectors/cablingTelescope mechanical (allignment, box etc) update Ph-2 testing boxTelescope trigger detector and logic system update PH-2 systemTelescope firmware/softwareTelescope test Beam RunData Analysis and feedbacktravel

Project WBS

Cost estimateWBS

HFT TC 10/14/2010 - LG 5

PXL Cost and Schedule – overview – walkthrough at high level

Level 4 WBS categories

•1.2.2.3 Phase-2 Sensor•Development testing and production procurement, dicing, thinning and probe testing of Phase-2 sensors

•1.2.2.4 PXL Sensor•LU testing•Development testing and production procurement, dicing, thinning and probe testing of PXL sensors.

•1.2.2.5 Ladder Cable•4 stage development of Ladder cable design and prototypes including Al conductor

•1.2.2.6 Phase-2 Ladder Assembly•Assembly and testing of Phase-1 Ladders including cable procurement.

•1.2.2.7 Read-out Electronics•Development, design, production and testing of RDO electronics and slow control system including DAQ RDO PCs and DDL procurement. 2 full RDO systems.

•1.2.2.8 PXL Sensor Ladder Production•Assembly and testing of PXL Ladders including cable procurement.

HFT TC 10/14/2010 - LG 6

PXL Cost and Schedule – overview – walkthrough at high level

1.2.2.3 Phase-2 Sensor

WBS Item CD-2/3 cost

CD-1 cost

diff Cause

Phase-2 production 168,749 65,400 +103,3496 additional Phase-2 wafers ordered => 42k$, 26k$ probe station development is included here, dicing and back-thinning 8k

We currently estimate to have enough wafers for a 3 sector prototype detector. If we wish to make more it takes just over 1 wafer / sector.

HFT TC 10/14/2010 - LG 7

PXL Cost and Schedule – overview – walkthrough at high level

1.2.2.4 PXL Sensor

WBS Item CD-2/3 cost

CD-1 cost

diff Cause

PXL sensor 876,620 880,000 -3,380 Added scope – Latch-up testing

Note that task 158 (1.2.2.4.3.1 Respin sensor design 3) is a contingency item and that the duration is schedule contingency. We need a good way to integrate this.

HFT TC 10/14/2010 - LG 8

PXL Cost and Schedule – overview – walkthrough at high level

1.2.2.5 Ladder Cable

WBS Item CD-2/3 cost

CD-1 cost

diff Cause

cable development 133,989 153,475 -19,486 some progress (ITB), cheaper Al cable procurement

We have made progress on this item, as indicated in the schedule.

HFT TC 10/14/2010 - LG 9

PXL Cost and Schedule – overview – walkthrough at high level

1.2.2.6 Phase-2 Ladder Assembly

WBS Item CD-2/3 cost

CD-1 cost

diff Cause

Phase-2 Ladder Assembly 161,720 118,982 Assembly of more ladders in scope

Note that sector fabrication is a mechanical WBS item – put as a placeholder to tie to mechanical schedule.

HFT TC 10/14/2010 - LG 10

PXL Cost and Schedule – overview – walkthrough at high level

1.2.2.7 Read-out Electronics

HFT TC 10/14/2010 - LG 11

WBS Item CD-2/3 cost

CD-1 cost

diff Cause

RDO electronics 583,386 452,849 +130,537new RDO board (139.3k$), 19.5k$ for PH-2 prototype installation, some effort on firmware has been accomplished

We have made progress on this item, as indicated in the schedule.

PXL Cost and Schedule – overview – walkthrough at high level1.2.2.7 Read-out Electronics - continued

HFT TC 10/14/2010 - LG 12

PXL Cost and Schedule – overview – walkthrough at high level

1.2.2.8 PXL Sensor Ladder Production

WBS Item CD-2/3 cost

CD-1 cost

diff Cause

PXL sensor ladder production 407,355 543,845 -136,490

246k$ for Al conductor cables in CD-1 vs. 54k$ for CD-2/3.Sector/detector now included in CD-2/3 for ~15k$.

This starts based on sensor delivery. There is a 3 month schedule contingency left in the schedule as part of a third redesign for the sensor. This task can start 3 months earlier.Note – no installation costs have been added. Is this a BNL task?

PXL Risk and Contingency

HFT TC 10/14/2010 - LG 13

WBS Item CD-2/3 cont

CD-1 cont

Cause

Phase-2 production 23% 41%In the CD-1 schedule 6 extra wafers for the prototype detector were labeled as contingency.

PXL sensor 46%New technology, new designs. One full mask fabrication held in contingency.

cable development 42%Driven by a 64% contingency on getting the Al conductor cable out of the CERN shops at cost estimated. Other contingencies sit at ~25-35%

RDO electronics 25% This seems reasonable for a new design based on a working prototype.

PXL sensor ladder production 29% This seems reasonable.

Note – the rolled up values for % contingency in the project document seem to be off. What is shown is cont$/basecost$.

• Risk and Contingency items for incorporation into existing cost and schedule.– Schedule risk for 3rd iteration of PXL sensor design. This is already

incorporated as a task with a 60 day duration.– If a re-design of the memories in the sensor is required for latch-up

tolerance, it will take ~6 months. This is not currently fully in the schedule, I assume currently one redesign (plus above) with a duration of 20 weeks.

– Cost risk – many of our procurements are sourced from Europe. Currency fluctuations are a risk. This needs to be explicitly addressed.

– Schedule risk associated with meeting a cd-2/3 approval date is not yet explicitly addressed.

HFT TC 10/14/2010 - LG 14

PXL Risk and Contingency

• Current Milestones – – Q2FY11 - PXL Final Sensor Prototype 1 Receive– Q2FY12 - PXL Final Sensor Prototype 2 Receive– Q3FY12 - Install Prototype PH-2– Q4FY12 - PXL Final Sensor Receive– Q4FY13 - Install PXL Detector

• Reviews – – 12/2010 Sensor design review – external– Prior to all production efforts – internal

HFT TC 10/14/2010 - LG 15

PXL Plans and Activities

Technical Progress

HFT TC 10/14/2010 - LG 16

We have made significant progress on the PXL detector this month. We had an in-person meeting with Rui de Olivera (head of the CERN Printed Circuit Board (PCB) shop). The CERN PCB shops appear to be fully capable of fabricating an aluminum conductor flex cable that would meet the requirements of the PXL readout cable as currently envisioned. A proof of principle is shown below in the form of cable developed for the ALICE pixel detector.

Figure 1 Aluminum conductor flex cable developed for the ALICE PXL detector. This cable is four layers with micro-vias and a 50 um feature size. The top side of the cable is gold plated to allow for wire-bonding and soldering. The lower cable in the picture shows one of the inner layers in aluminum conductor.

The timescale for delivery of boards can be negotiated but is similar to standard industry timescales. To produce the 200-300 cables that we need would require about 3-4 months. It would be better to work out a delivery schedule that allows for slower production if it can be arranged with the PXL schedule. The cost for the ALICE cables (50 um trace and space and microvias) was 170 SF per cable. This is a much more complex cable than the STAR PXL cable. He estimates that, based on my description, the cost would be about half, or 85 SF per cable. We can pay him just as we would any vendor. He will give a written quotation and can invoice LBNL. Rui recommends gold as a finish for wire bonding and component attachment (soldering or conductive adhesive). This would consist of 5-6 um of nickel followed by 0.1um of gold. They have not had good luck trying to get bond wires to stick to the amorphous aluminum that is the product of vacuum deposition. Since the CERN PCB shop also makes standard PCBs, we will do fabrication for the next 3 phases of ladder cable development there. This will serve as a way to familiarize the CERN shops with our designs, gain their advice and work out any communication or procurement problems before the aluminum conductor cable production fabrication. We now have a well defined path to the lower radiation length PXL detector as well as a more reasonable cost estimate for the cable fabrication.

HFT TC 10/14/2010 - LG 17

Meetings were held with IPHC in Strasbourg with Marc Winter, Christine Hu and other members of the PXL sensor design team. The main focus was to feed back information gained from the testing that we have done on sensors in a ladder configuration. These tests have resulted in us requesting several small changes to the final PXL sensor design. These include converting the sensor “START” signal from a single-ended to a differential LVDS signal, removing the requirement of bonding to the “RSTB” bonding pad, and the implementation on the “SPEAK” signal over LVDS or defined in a fixed state by using an internal “pull-up” resistor. IPHC has agreed to these modifications and will implement them in the initial design for the PXL sensor. The net result of these changes will be a full differential sensor interface for running conditions, one less trace in the bonding area of the sensors on the cable, and fewer bonding wires.

Technical Progress

HFT TC 10/14/2010 - LG 18

Technical ProgressIn addition we have made progress in the probe testing, ordering a new probe testing card and in the planning for a set of latch-up measurements to be done at the LBNL 88” cyclotron in early November. We plan to test new MAPS prototypes and all PXL detector electronics components, which are planned to be located in high and moderate radiation areas in STAR. MAPS prototypes to be tested for latch up and SEU cross-sections are Mimosa 26 and a dedicated chip for latch up tests designed at IPHC. The LU test chip is composed of shift-registers that are built with digital cells with different spacing between NMOS and PMOS transistors. The latch up test results will help to assess the need for designing new, latch up resistant digital cells for use in sensors for the PXL detector. As soon as the full testing setup is assembled, which is expected to happen in the second week of October, we intend to be on-call and ready to fill in empty time slots that may appear due to cancellations or earlier terminations of user runs at the 88” cyclotron. In this scenario, there is a possibility that latch up (and SEU) test results would be available before the end of October. A more likely outcome, however, is that we will use free beam slots in the 88” cyclotron in early November. A complete plan may be found here http://rnc.lbl.gov/hft/hardware/docs/latchup/ Latchup_plans_2010_(draft2).doc

backup

HFT TC 05/11/2010 - LG 19