Difference: PixelWeekly (29 vs. 30)

Revision 302015-06-01 - JohnLipp

Line: 1 to 1
META TOPICPARENT name="Meetings__Wikis"

Pixel Weekly

Line: 8 to 8
  In general, indico pages are listed in: https://indico.cern.ch/categoryDisplay.py?categId=547
Date Indico page Other notes/docs (e.g. uploads to PUUKA Twiki)



  1. Module assembly of CMS Phase-I pixel upgrade: Kirk (20mins)
  2. Module mini-production: Craig (20mins including discussion)
  3. UK preparations for module assembly: Craig (30mins including discussion)
  4. Report on high rate bump bonding: Richard (20mins including discussion)

Items 3 & 4 will be reported at tomorrow's ITk module meeting. The indico page is at: https://indico.cern.ch/event/396129/

7th May 2015 https://indico.cern.ch/event/391195/
  1. Update on thinned module bump bonding, Richard 15mins
  2. Update on RAL BB, John L, 15mins
  3. Status of multimodule testing, Richard/Sneha, 15mins
  4. Hybrid to module assembly work, Ilya/Paul, 15mins
  5. AOCB
  1. Update on thinned module bump bonding, Richard Bates
    See Richard Slides.
    Richard presented an update on thinned module assembly. This focused on work at Advacam on flip chip of assembly. Several different techniques have been looked at: Au UBM to replace current Advacam transition metal UBM, solder on sensors to give a solder-solder bond and a new reflow method: "Advacam reflow method". 5 samples with 100um thick readout chips and 300um single chip sensors, some have SCL and some do not. Good bump yields with solder-solder and also Advacam reflow method.
    More samples being made.
    Need extend Advacam reflow method to quads and check it is compatible with large scale production.
    Samples will be subjected to thermal cycling but it was pointed out that glues had failed on strip samples at -50C. Test samples will be thermally cycled before the samples.
  2. Module production exercise
    See Richard slides (last slide)
    Richard reported that at the pixel design group large scale bump bonding was discussed (https://indico.cern.ch/event/392019/). The bump bonding capacity needs to be understood to support the case for increasing the pixel area to a 5th and 6th layer. The aim is to place orders with a number of vendors: IZM, Selex, Advacam/Cea-Leti, HPK and see if the meet required production rates and quality. It would be better to have more vendors, around 6 so that we have a choice rather than being limited to using the current 4.
    Daniella pointed out that RTI have worked on bump bonding for CMS and should be considered. Vitaliy Fadyev is the currently the contact for them. Daniella will contact Vitality regarding RTI.
    RAL have been considered as small site that could work on the inner layer given their limited resources. However, John reported that the following discussions, he had been given the go ahead to look at developing 300mm capability.
    We need to consider module production, hybrid assembly and final testing.
    The programme of bump bonding needs to be complete by end of the year so that decisions can be made on the layout. The aim from the UK is make and test 50 assemblies with CEA-Leti & Advacam, and also a similar number with RAL. Craig will organise a meeting with Craig, John L, Richard and Paul to discus the programme for this.
  3. Update on RAL BB, John Lipp
    See John's slides
    John gave an update on the studies to understand the quality of the bumps. In the "old" daisy chain samples that included Au in the UBM, intermetallics were observed, confirming that removing the Au from the UBM was correct. The results from the module that did not have Au in the UBM showed no evidence of intermetallics. The UBM and inter-bump interface appear to be good. The remaining possible problems are that the parallelism of the FE-I4 and sensor may not be good enough or there may have been Cr contamination from resputtering. If it is the resputtering, this has now been fixed. Next step is to make more modules and evaluate them.
  4. Multi-module testing, Kenny
    See Kenny' slides
    Kenny showed the progress with the multimodule testing system. After some teething problems with connections the system is now setup to test up to 16 chips = 4 quads. This will start with modules on PCBs but will move to testing modules with flex hybrids on the half disk. The serial powering system is up and runnning with SCTDAQ and an interlock system.
  5. Hybrid mounting, Paul
    2 quads ready for sending to Glasgow, will be sent up together with the assembly jigs.
    Glasgow will send 6 FR4 populated hybrids to Liverpool.
    Richard reported that hybrids from Flexible technologies had cracks. This appears to be due to using a flexible solder resist rather than polyamide.
    There was a brief discussion on using rice paper as a replacement for the gluing stencil. Ilya will be asked to report on this at the next meeting.
    Kirk mentioned his experience with using a rubber stamp method for glue dispensing. He will report on this at the next meeting .
  6. AOB
    Paul reported that Sussane and Wuppertal cofirmed that we will get 40 untested PPSP chips. This will be in August/September.

Paul Dervan noted: This was the glue/paint that failed at about ~40C: http://uk.rs-online.com/web/p/conductive-adhesives/1863593/

16th April 2015 https://indico.cern.ch/event/388245/
  1. Hybrid to module assembly and electrical testing of modules with flex hybrids: Ilya/Paul
  2. Report on In bump bonding at RAL: John Lipp
  3. Summary from DAQ meeting: Craig
  4. Input to RD53 sensor discussion: All
  5. AUW talks: All
5th Mar 2015 https://indico.cern.ch/event/368657/ 11:15am
This site is powered by the TWiki collaboration platform Powered by PerlCopyright © 2008-2020 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback