Go to the Indico Home Page Oldest event Older event Up to category Newer event Newest event Printable version
  • Compact style
  • Indico style
  • Indico style - inline minutes
  • Indico style - numbered
  • Indico style - numbered + minutes
  • Indico Weeks View
    • Switch to the management area of this event
    Login

    TB: TDC replacement for the LHC LDM and for ISOLDE applications

    → Europe/Zurich
    864/2-B14 - SALLE J.B.ADAMS (CERN)

    864/2-B14 - SALLE J.B.ADAMS

    CERN

    25
    Show room on map
    Hide

    The TDC would be used in the LHC BSRL (or LDM as in Longitudinal Density Monitor) and for the ISOLDE applications and PUMA.

    Currently the modulke used is the Agilent U1051A / Acqiris TC890 , which cannot be acquired anymore. Recently one channel failed.

    3 modules operational (2 for the BSRL and one for ISOLDE) with 1 spare.

    BSRL requirements:

    • 50ps or better resolution
    • 100us window
    • max 15ns dead time
    • hit rate at least 2M hits per second 

     

    ISOLDE & PUMA TDC requirements:

    • 1ns resolution
    • 500us window (30us for PUMA)
    • 15ns dead time can be handled but ideally 1n

     

    For ISOLDE the proposal is to take over the TDC of the users/experiments as well. Few questions about why and if this can be accomodated with a fixed FESA class.

    • Clear specifications and an official request should be obtained. <= Action Miguel/Federico
    • How to limit the access to a HW we maintain to the users that need to reconfigure but might damage iit

     

    The resolution requested is very small, but the jitter from the reference has not been measured yet. What about dispersion in the fibre? <= Action on Stefano

     

    Current TDC (VME based) specs:

    • 6 channels
    • 50 ps resolution
    • Recording range up to 10 ms
    • Dead time <15 ns
    • Edge comparator (pos/neg)
    • ±1.5 V threshold, ±2.5 V max
    • 10 ns post-start dead time

     

    Alternative 1: TimeHarp 260 PICO (PCI-e 1 lane)

    • 25 ps resolution
    • Single channel (dual channel available)
    • 0 to -1.2 V input, -1.5 V max, Constant-Fraction Discriminator
    • 40∙106 Hits per second sustained (dead-time limited). 8M records storage
    • <25 ns dead time → longer than current <15 ns
    • Dead time can be reduced using 2 channels (extra slide)
    • Double pulse dead time essentially 0 (>double pulse still 25 ns)

    The Alternative 1 dead time is too long

     

    Alternative 2: MultiHarp 150 (USB 3.0)

    • 4/8/16 channels + sync
    • Spare channels for BSRL
    • Can separate ISOLDE into two zones for better signal quality
    • 5 ps resolution: fits neatly into 25/50 ps
    • Binning at time of acquisition to reduce histogram data size
    • Level trigger, bipolar
    • 80∙106 hits per second sustained. 1.5∙109 for 2k hits peak. 134M records storage
    • White Rabbit: future-proofing/second source of timing
    • Negligible dead time (650 ps)

    White rabbit connectivity would not actually add much as it is not BST via WR 

    Concerns about the driver and API: it is not given to be recompiled but it might be not needed. This should be verified <= action on Miguel

     

    Alternative 3: PicoHarp 330 (USB 3.0)

    • 1/2/3/4 channels + sync
    • 1 ps resolution
    • Everything else has much more uncertainty or jitter
    • 1000x too fine for ISOLDE
    • Choice of level trigger or CFD
    • 80∙106 hits per second sustained. 1.5∙109 for 1k hits peak. 256M records storage
    • Negligible dead time (680 ps)
    • Most expensive (especially for ISOLDE: 5 units)

    This is clearly over specced and too expensive

     

    Alternative 4: xHPTDC8 (PCI-e)

    • 8 channels (1 used as sync)
    • 13.02083̅3 ps bin size
    • Does not fit evenly into LHC timing
    • Level trigger, highest range of input signals (±5V)
    • Good for ISOLDE where users might plug unsuitable signals
    • 5ns dead time
    • 30∙106 hits per second sustained (all channels). 12∙106 per channel. Variable records storage
    • Can parallelize multiple cards: only one channel is lost to trigger
    • Uncommon AC termination: hard to set thresholds, especially for longer pulses

    The AC termination is felt as a main limit for flexibility.

     

    Final comments

    Among the presented alternatives only the  MultiHarp 150 is totally viable, with the caviat of the driver/interface.

    Other options should be examined and excluded before pushing for the  MultiHarp 150: BE-CEM FMC TDC, commercial FMC TDC. <= Action on Miguel.

    PUMA specifications are not clear yet, but the development is urgent. <= TB end of Nov beg of Dec action on TB organisers
    With the current timeline it is assumed to have it by the end of the year. Miguel to communicate them that it is not possible and to have an updated timeline. <= Action on Miguel.
    Before TB a document of spec from PUMA to be commented and discussed <= Action on Miguel.
    For the test of the line can be used a Agilent spare or a spare channel of the ISOLDE one

     

    There are minutes attached to this event. Show them.