Heidelberg weekly meeting

03.410 (Physikalisches Institut, Heidelberg)


Physikalisches Institut, Heidelberg


For Vidyo please use room extension 10455005 or use this link.

    • 10:00 10:45
      Round Table 45m

      Weekly reports of work progress

      Speaker: Annie Meneses Gonzalez (PhD Student)

      Present: André S. , Annie M. , Lars N. , Thomas R. (Vidyo), Christoph B. , Adrian H. , Luigi V., Heiko A.

      Minute-taker: Heiko (from memory)


      • LV: new vacuum chuck  available for the probe station, but the chip still moves when pressing the needles to the pads.
        Different solutions are discussed:
        perfectly fitting metal cut-out will cause stress on different sized sensors;
        the sensor can be limited on one side by a metal edge to stop the movement;
        HA: the chip movement is only an effect due to the tweaked setup (the needle card is tilted to only contact one edge with needle); this problem will not be present for the newly ordered needle card for Oxford with needle only on one edge.
      • AS: raises concerns about dust particles in the clean room, which might be even more relevant for systems using vaccum (eg. the new vaccum chuck).
        It has to be made sure that the chips and therefore also the tools stay clean as dust can heavily affect all further handling steps and even destroy the sensors/module (example BelleII)
        LV: reports of residual oil traces on the vacuum chuck after production which had to be cleaned with pressured air.
        HA: pressured air/nitrogen also available in the clean room
      • HA: the mu3especs for the MuPix10 chip are progressing well; a pull request for the newest version v0.3 is pending
      • HA: The Process Design Kit (PDK) for TSI is now available in the ASIC lab HD; the NDA was signed this week; this means it is no longer managed via Ivan
      • AH: progressing well on the alignment of the mimosa testbeam data for AtlasPix; the aligment with GBL is involving manual steps which require some experience; practicing continues
      • AM: see slides; reports on the newest findings concerning placement of guardring/ p-stop structures in between the pixels:
        1) p-stop and p-spray very important for increasing substrate resistivities, as they allow to electrically disconnect neighboring pixels (inter pixel capacitance)
        2) the p-spray should not contact the deep n-well directly. the larger the distance, the higher the simulated breakdown voltage; 4um seems to be a good enough value. It is a trade-off between breakdown and inter pixel capacitance.
        3) the metal metal distance between ground potential and the high voltage also effects the breakdown voltage. 5um are safe.
        4) the relative positioning of the metal layer and the p-spray is important to reduce local high field regions on the SI-SIO2 interface
        5) As a possible reason for the breakdown difference observed for AtlasPix2 between AMS and TSI, the SI-SIO2 interface charges are considered; the standard value is 10^11 charges/cm²; if the process has flaws this number might be higher; simulations for varying interface charge densities are currently on going.
      • CB: after software problems have been solved for the s-curve scan, first measurements have been obtained for MuPix7-AMS, recreating the previous results produced by Johannes Grimm with the "old" setup. The measurements of the TSI chip is the next step.
      • LN: expecting photo mask designs for the flexprint production still this week; masks will be reviewed in HD; the production maybe ready before the testbeam end of April
        AS: testbeam not so important at the beginning, first get it running ; do not sprint to testbeam endeavours as in the past.
      • LN: reminder to the discussion of the chip termiantion value, currently defined to be 85Ohm in the Mu3eSpec-Document.
      • AS: magnet production progressing well; pictures of produced parts available; small delay of delivery date (3 weeks) due to parts produced with sub-contractors
      • AS: announcement that the CMOS project for the L4 layer in the Atlas LS3 Upgrade is no longer considered an option by Atlas (recommendation by the review committee)
        Never the less the Heidelberg activitiy will continue and also the submission of AtlasPix3 will be done.


      Post-Meeting discussion on the Termination:

      • AS: value 85 Ohm triggered by Frank, coincides with the twisted pair impedance.
      • LN: chip is connected to the flex, here the impedance is not under control yet and also subject to process variations; impedance matching between all parts (twisted wire-->flex-->flex-->chip) should be discussed; LTU results required.
      • AS: chip spec will be dictated by the other systems.
      • LN: the current simulation tool (mentor graphics) creates heavily varying results depending on how you choose the volumes around the flex (glue,chip:silicon.metal layers ...)
      • AS: currently an additonal polyimide layer is introduced between the 2 flex cables to increase the impedance
      • LN: the new production will produce both types, with and without the additional layer
      • HA: if there is so much variation we need to know what the range is and can design the chip to sit in the middle of the band (matching)
      • AS: removing the additional layer and going overall to lower impendaces should be considered. twisted pair cable with lower impedance should be available.
      • AS: to make better predictions (and designs) a new impedance simulation tool might be necessary
      • LN: maybe SolidWorks(KIP)
      • AS: ask Thomas Mittelstaedt (MW CAD)
      • LN: old open source program used by Beat Meier might be available
        "ATLC – the Arbitrary Transmission Line Calculator" http://atlc.sourceforge.net
      • LN: does the coupling between flexprint layers need to be considered for the architecture of the LVDS drivers and receivers.
      • HA: post-pone this question until values for the coupling are known.
      • AS: Experts in the electrical workshop might have experience with this
    • 10:45 10:50
      AOB 5m