This 2nd general VTX workshop mostly focused on mechanical integration aspects. But all other VTX activities were also discussed to some extent.
Here is a very brief summary of the meeting, followed by minutes, briefly summarizing each talk and discussions
Main points and open questions
==> Progress
- 2 validations! TPG heat-conductance and lpGBT operation after irradiation
- oVTX services design close to ready
- iVTX services design partly done
- iVTX ladder concept: plan for validation
- Excellent discussion with PXD
- iVTX, oVTX module & ladder assembly: first detailed considerations
- first test on iVTX patch-panel prototype
- continued discussion and plans on back-end boards and DC-DC converter board
- WG6 started
- Plan for TDR consolidated, writing to start shortly
==> Decisions
- oVTX-ladder supporting end rings with slanted based adapted to cone shape
- Margherita's presentation defines our oVTX status -> to be translated to documentation
- all current dock-boxes (serving PXD+SVD) will be used by VTX (preliminary guess)
- for installation related topics quote 'like VXD' in TDR
- new chat from DESY: start with testing and management channels
- create VTX group in Gitlab==> Questions requiring a rather urgent answer
- WG1, BG hit-rate for L1
- WG1, cross-check #sensors needed per layer with reduced FWD acceptance 17->22 degrees
- WG1, shortened L1 and L2 to 3 sensors (see above)?
- WG2, final GDS of pad rings expected by WG4 (4-sensor flex)
- WG2, update of power dissipation expectation, awaited by WG4
- WG4, build sharing proposal of dummy-chips (OBELIX-1 pad wafer) between iVTX, oVTX
- WG4 plan for thinning pad-wafers
- WG4 for end-flex: stress relief, line-matching, length rather 50 cm?
- KEK-mechanics: FEM analysis by to validated new design of support-rings with extended empty shapes
- WG5, constraints on trigger jitter from VTX read-out
- WG5, bandwidth of links from VTX-backend to trigger-board
- WG6 to connect with WG3,4,5==> Open question (not necessarily urgent)
- WG1 list of benchmark physics channels to be studied
- WG2, will side pad positions for OBELIX-2 change / OBELIX-1?
- WG4, final thickness of Si block
- WG4, need of iVTX sliding mechanism or not?
- WG4, plan for iVTX end-flex attachement
- WG4, how is the X-Y aligment done of the iVTX ladders / mounting rings?
- WG5, how to provide front-end to Chinese group for test-bench with back-end boards==> Documentation
- WG4, implement service doculentation, build list of step files, with file location and date
- WG5, documentation on patch-panel, specs for back-end, ...
- Need for help page allowing everybody to contribute
Minutes
* WG2 sensor design, slides from Hung
- OBELIX-1 submission: still in the last step to correct issues with design rule
- OBELIX-2 design preparation, see https://vtxupgrade.pages.desy.de/obelix1/obelix2_plan/
- 3 open questions from users:
- stop read-out & flush matrix: is it implemented in OBELIX-1 already or only for OBELIX-2?
=> to be checked with digital designers
- detailed power with hit rate still missing => TBD after submission
- plan for thinning pad-wafers ?
* WG3 sensor testing, slides by Christian
- testing of TJ-Monopix2 phasing out
x last open question for consistency between MPV (charge) level wrt efficiency at low bias voltage (< 10V)
- Workshop on DAQ developments
x 1st in 4-5 May
x 2nd just before OBELIX-1 arrival
x first test scripts ready by this Summer
- list of very initial tests
x discussion already started during TJ-Monopix2
x plan should be ready for when OBELIX-1 arrive
- For test beams
x possibly MedAustron October-November already
x then KEK November-December* WG4-oVTX design, slides by Maurizio
- overall CAD progressing, but some missing bits
x for example manifolds missing
- end rings supporting ladders
D x decide for solution B on page 6 with slanted based adapted to cone shape
- lenght of ladders and nb of sensors per flex documented
- 1st ideas on the cooling system (plant) design
x quite some questions open to progress on this design
- progress on omega-shape support
x stiffness reached (low sagita) with re-inforcement
x matrerial budget +0.04% XO => reaching 0.613% X0 fo L6
- thermal test on-going with kapton heaters, to be compared with simulation
- procedure: layer N + services of layer N then layer N+1 + services ...* WG4-oVTX, dummy flex design, slides by Alessandra/Massimo
- 4-sensor being prepared
x assume 4 TTT output aggregated on 1
- special attention about bounding clearance
x in principle, no need for deep access tool
- connector prepared for OBELIX-emulator on a FPGA board
- Reminder of plan: 4-sensor-Cu -> 4-sensor-Al -> L6length-Al
x lpGBT and VTRX+ will be integrated for L6length proto
- Electrical tests for 4-sensor-Cu planned in Vienna
- Questions:
x why call it dummy flex? indeed connexion to actual sensors
x nb of layers? 6 metals, already for 4 sensors
x testability? would need a temporary test boards rather than specific stuff on flex* oVTX assembly, slides by Stefano
- List of preparation work for production:
x specs, procedures, QA group, tooling, facilities, shipping/export
x eventually leading to establish a schedule
x 2027: sharing of the work should be established
- underline variety of design (flex, modules, ladders)
x 7 flex types
x 50 discrete components on flex for each sensors => 500 components on ladder
=> try to identify facility/person-power available for R&D
- proess proposal for assembling sensors on flex, i.e. module assembly
x use xyz-stage + CMM control
x gantry machine for automatic placement also possible
x underline critical aspect of gluing spots for bounding stability
- process proposal for assembling modules on ladder, i.e. module mounting
- Assembly of 4-dummy-chip proto module starting in Pisa
- Note Pisa has a laser cutter now (same as in MBI)
- New program by Pisa
x goal 1 = L5A with dummy-flexes 12+4 flexes
x goal 2 = 4 OBELIX-1 on Cu-dummy-flex
x goal 3 = L5A with modules
- Questions
x will side pad positions for OBELIX-2 change / OBELIX-1?
x should assembly site be specialized for 1 or 2 types of flex?
x how many dummy-chips (OBELIX-1 pad wafer) do we need?* iVTX design & tests with TPG, slides by Julien
- Reminder on main concepts (ladders, supporting pieces, assembly of iVTX halves on beam pipe)
x Thicknesses: Si not clear what is driving the thickness
x assembly of detector assumed to be similar as PXD
x support rings need specific cooling (taken into account in cone design)
- OPEN questions:
x detail of the end-flex attachement
x grounding of TPG
x sliding mechanism needed? wrt to heat transfer
- About sliding mechanism
x sliding expected at 10 um level from naive CTE computation
x "hope" is sliding not needed
x dedicated tests needed -> possibly long process
- thermal mock-up
x design of set-up done, progressing to realisation
x TPG cutting, parylene coating, TPG glueing, copper arms: all planned, realisation starts soon
x results hoped for before this Summer
- Discussion
x wafer thickness?
o thermally not important
o the thinner the better for material budget
o decision will depend on assembly yield and stiffeness
x TPG is not stiff at all => Si will bend the TPG
o do we need a stiffener somewhere?
x which measurement to control the movement?
o not planed
x statistics needed to confirm result / sliding / bending
o yes important
x glueing technique of arm over TPG?
o suggestion by Franz,
- News from Shuji: availability of a TPG-like material by Kyocera
x looks stiffer than TPG, 2 pieces obtained 14x1 cm,
x what are the exact properties?* Diced block out of CMOS processed wafers, slides by Mathieu
- made of MIMOSIS-pad wafers
- dimensions very close to iVTX
- Available now in two thicknesses 100, 150 um
x working toward 50 um
- available now BUT no plan yet to use them yet
- Questions
x alternative are the IZM samples but rather thick: 400-700 um
x such Si ladders will be used at some point but after first test with TPG alone* iVTX development in QMUL, slides from Dom, Adrian, Ali and Dhyey
- Reminder of overall design, Dom
x various parts of all pieces quite detailed, still in-progress
copper arms, inner/outer mounting rings, manifold or cooling block, mount blocks
x services in-progress as well
- design of sliding mechanism getting simpler
x use of graphite TIM by Panasonic
x next step with mechanical prototype
- next steps include prototyping, thermo-mechanical testing with IJClab
- Questions on design
x test-try installation as soon as possible
x how is the X-Y aligment done of the ladders / mounting rings?
x TODO: get mockup of end pieces as soon as possible??
- Conductivity measurement, Adrian, Ali
x 1st measurement on pure material: 4.7 W/m.K
x setup ready with NTCs read-out to measure long TPG samples from Julien
- Simulation on water cooling option, Dhyey
x parafin solution for ladder excluded
x water solution works with some ideal assumptions though
x depending on the need for this solution, further dvpmts can be done* CBM Irradiation of TPG samples, slides by Franz
- irradiated samples 70x10x0.5 mm2 at IPHC-CYRCé
SI fluence n_eq Si: 1e13, 1e14, 5e14 & 1e15 neq/cm2
- very robust study (calibration, measurement, analysis)
- no influence of radiation on heat conductivity over fluence explored
- absolute values of heat conductivity a bit overestimated but reasonable
=> conductivity 1500 W/m/K is stable with irradiation* iVTX-DRL process, slides by Andres
- new approches proposed by IZM
x RDL first: produce RDL, bonds single dies on it and glue in "epoxy-like"
x ladder first: dice block, then apply RDL
o require alignment structure bottom/up
=> iVTX prefers ladder-first approach
x now preparing quote and order with IZM
- bending issues from 1st IZM process
x 15g weight needed to flatten the thick ladder
x to be re-done with thinner layer
- additional measurements on existing prototypes
x perfect eye-diagram
x post-meeting NOTE: mail from Carlos (confirmed by Andreas) stressing that resistivity are changing depending on whether the wafer has a CMOS process or not
- home-process still in development,
x produced proto under tests
x next steps already planned
- end-flex cable first idea (aka kapton-flex)
x 3 layers, 40 cm long, 1.9 cm width
x ready to order soon
- Questions
x for end-flex: stress relief, line-matching, length rather 50 cm
x discrete components on RDL? Ni-Au pad coating
x aligment structures dimension ? ~200 um smaller that what we need
o important is: need structure like pads
x timeline for next IZM trial
x need to develop their process: October
x expect electrical properties should be identical
x IZM can only do 2 um thick Al, too thin for our transmission goal => stick to Cu
x why not continueing process like 1st-prototype
o too many (cost) masks which prevent adaptation to changing good-ladder position on wafers* Integration matters, slides by Shuji
- list of action items for KEK-IR, iVTX, QCS, BG and STR groups
- review of IR, QCS, STR activities
x note that decision timing for Nb3Sn or NbTi is still to be clarified
- review of activities during LS2
x list of main contact/responsible peoples for the various groups
x probably a new integration meeting will be decided
x rough succession of procerdures during LS2
x note VXD group sollicitaded from dismounting VXD including cables
- Space for sensor integration
x B1: oVTX, iVTX
x Fuji hall: need reservation
x VXD clean booth on B4: VTX integration
- First consideration for the integration of ITT detectors
- Questions / Discussions
x BG mitigation impact on VTX?
o anyway no room for additional metal shield
x ITT integration room too small?
a risk
x mechanical structure for iDC or ITT supporting the inner later
a strong constraint from VTX
x dock box for VTX:
need all PXD+SVD dock-boxes
x old VXD, possibly activated, will stay in B4
- when: floor of B4
- who: VXD group + VTX group
- maybe 2 to 3 weeks to remove
x gold coating reduction
- really scaring people (Carsten, Shuji)
- better decreasing radius since smaller beam pipe
x reduced beam pipe diameter => already decided
x Job on VTX
- oVTX nothing till TDR, for installation quote 'like VXD' in TDR
- iVTX needs work on service, then similar to oVTX "like VXD"Here is a supplemental picture album link, for iTT/oDC service space.
https://photos.app.goo.gl/7vESaeQLzPpbKTfq5* WG4-oVTX services study, slides by Magherita
- Decision path:
x FEM analysis by KEK mechanics to validated new design of support-rings with extended empty shapes
x if validation => we can do realistic tests with mock-ups
x if issues => remove insulations, decrease cable diameters against electronics rule
- Next steps?
x nothing more for TDR
x iVTX rough sketch for June (Dom)
- about installation cart: no worry for re-use because VTX lighter than VXD
- this presentation = documentation material for web page
- Brief discussion about diamonds
x sensors on beam pipe critical
x other sensors (SVD cone) not so much* Discussion with PXD gemba-style
- installation procedure on beam-pipe
x PXD did vertically to beenfit from gravity but required rotation
=> strongly disliked by KEK and worry that not possible next tile since Koriki-san will not be around
x Karsten thinks it is doable for the side
the key issue is alignment on both sides simulatenously and ability to push in place* WG1 Status of performance, slides by Vidya
- reminder of baseline geometry (well documented on gitlab)
x (beneficial) impact of beam pipe with smaller radius and thicker material already anticipated
- smaller radius of L1, current question is hit rate?
x Carsten using Eugenio formula for BG increase -> much less increase
=> work on-going to understand BG
x work is done and followed by Vidya
- Nb of sensors reduced along with reduced acceptance
x could be used to reduced variety of flex
x could reduce heat load
x BUT crosschecked absolutely needed
- not covered yet
x Ks efficiency
x FEI and tagged analysis (wait for large MC production)* Upgrade performance group, slides by Ale
- Some info on non-tracking studies (FWD-taggers, ECL, KLM)
- FTL benefits clear for PID and pt<300 MeV
- pt resolution very slightly worsed with FTL compared to empty ITT
- Benefits on tracking expected for connextion oDC to VTX and in reduced fakes
=> to be studied (Martin Schmied)
- Study (Ale) of boost reduction to recover physics acceptance with reduced geometry acceptance (FWD 17->22 degrees)
x optimal boost ~018 but only for +0.5% absolute efficiency
- reduced boost impacts vertex displacements, important mostly for TDCPV analsysis
x however possibly not the most critical parameter
- beam background under study by Bogdan Kutsenko
x VTX and FTL geometries ready, STL is expected
- list of benchmark physics channels to be studied* WG5 lpGBT test, slides by Markus
- goal reminder: qualification of lpGBT at 42 MHz operation frequency
- 15 lpGBT and 8 VTRX+
- doses: 0.21, 0.42, 0.87, 0.92 MGy
- results after irradiation: max frequency increased slightly (+1%)!
- supply current unchanged
- optical output power reduced by 15%
o OK if good fireflyer tranceivers used
- Questions
x no clear explanation for enhanced frequency
x NIEL irradiation needed? rather not, since no expected impact on frequency
x SEU ? triplication present, not a worry
x VTRX+ order: expected mid-2026
x lpGBT order: 500 units to be placed, no constraint on when to place the order* WG5 lpGBT test board for iVTX
- current patch-pannel techno HDI 6 metals
- first proto received in late February, test on-going
x power, PLL, up/down com OK on one board
x next is e-link and test 2nd board
- Discussion points
x test by another group welcome
x test for thermal behaviour to be done
x do not forget documentation!* WG5 activities of Chinese groups, slides by Jingzhou
- review of IHEP solutions for back-end board
- option 1 based on ATCA standard & Virtex Ultrascale, including trigger
x 1st version expected mid-2026
- option 2 based on MTCA standard & Kintex Ultrascale
x 2nd version in preparation, possibility to add requirements
- Organisation between IHEP, Shandong, Jilin to develop firmware
x testbench would need front-end
- review of specifications for power supplies and potential candidates
x DC-DC: CERN bPOL12V_v2 or bPOL48V_v6 match radiation hardness but warning on reliable issues@
x CAEN back-end supplies: various boards possible
- Discussion
x no financial arrangement decided yet on China side* Interface with trigger group, slides by Koga-san
- strixel granularity drives the bandwidth
- review of #links, #back-end boards and #trigger boards
x #trigger boards cannot be decided yet, depends on following question
x question to VTX group: bandwidth of links from VTX-backend to trigger-board
- no plan to update timing resolution
x currently maximum (worst case) of 50 ns
x request from Koga: which constraint on the trigger jitter from VTX* WG6 activity startup, slides by Carlos
- goal: demonstrator operating with VTX DAQ and modules (iVTX, oVTX)
x long-term and decorrelated from VTX: large area beam telescope with OBELIX
x "short-term": several single-OBELIX planes operated together in a telescope geometry
- Groups interested: IFIC, IGFAE, IFCA, Bonn, RAL, Göttingen, Vienna, Pisa, KEK
- IFIC group building up:
x so far followed other WG activities
x will soon start contacts
- Discussion point about back-end board
x difficult to get FELIX, renting possible
x limited amount especially for 155
x ALICE3 considers to produce such boards* TDR discussion, based on Jerome's slides
- Reminder on structure and timeline discussed in Pisa (Dec 2025)
- Reminder on list of Technology Demonstrators, as basis to validate design in TDR
- Proposal for main writers for 1st sections and milestones (see slides)
- Discussion:
x sensor contains part on test (Giuliana)
x mechanical structure includes services & integration
x power supplies in DAQ
x page count target? max 50 pages (preliminary estimate)
x be light on installation, use 'like VXD' as much as possible
x constraints from SKB and BII should be in 1.b) General Description
x renumbering of layers L1,2,3,45 (instead of L1,2,4,5,§) YES, no chance to revert back to 6 layers
- TO DO:
x start overleaf
x change documentation with layer numbering
x update logo (Alice?)* General discussion, based on Jerome's slides
- Chat:
x old RocketChat: organize archive
x new: create management and testing channels
- Database: still based on hephydb but renewed (Christian IPHC & Christian MBI on it)
- Gitlab: create a VTX group to allow write access by default to various VTX projects
- Documentation: one reason for lack of contribution: missing help => PEDAGOGY Needed
- Next general workshop:
x when? probably end of 2026/beginning of 2027
Note that MBI move to PSK from OCt-Dec
x call open for hosting
- Homework due in July 2026
x for institutions: build list of personnel, budget reachable and facilities accesible
x for VTX management: build Product/Work breakdown structure + start risk register
Pre-meeting practical information
The workshop will be held in:
- Monday & Tuesday: building 1B, 2nd floor, Seminar room 4A/b,
- Wednesday: building 5, auditorium
Remote participation is possible, use https://speakapp.link/to/I1Wn8A.
Eduroam and DESY-guest networks are available.
Starting and ending times should allow participants to travel on Monday morning and Wednesday afternoon.
Timetable is in preparation and current status is PRELIMINARY.
Fees will cover:
- coffee breaks and social dinner: 60 EUR
- coffee breaks only: 30 EUR
Social dinner on Monday 20th April at 7 PM at Hoppe's Restaurant on the Elbe bank: Övelgönne 6, 22605 Hamburg. Fish, meat and veggie options will be available. The dinner includes a main dish and one soft drink/beer.
Registration are now closed, contact us in case of issue.
