From tigwiki

Jump to: navigation, search

Use this place to add to your list of things you would like to see changed/added/removed/modified about TIGRESS. Please put your initials next to it. Use first-level bullet points to "title" a problem, then second-level to explain it and to keep track of actions taken. Please put a date when you entered it. I (GH) will be monitoring this page and trying to keep up to date on things.


[edit] DAQ

  • Remove "Reinitialize" from start menu -- GH, 2009-08-16
    • The "Reinitialize" option is waaaay too tempting for people like me (GH). It cost us 2 hours of beam time on this day. Take it out.
  • Modifications to event builder -- GH, 2009-08-16
    • Current event building on the front-end readout computer (lxdaq04) is a pain. The event buffer is hardcoded, so you either run out of buffers for gamma calibration events or you wait forever for your first event in a real RIB run. Either way event reconstruction is not guaranteed, and you would have to run an offline reassembly anyway. This cannot possibly work for multiple front ends.
  • New ODB with more than 2 detector types
    • As discussed by GH/CJP. -- CJP, 2009-09-18

[edit] LN2

[edit] TIG-64

  • ADC speed operational at 50 MHz - GH, 2009-09-18
    • Currently it is only reliable at 25 MHz. Needs to go to 50 MHz. Required by RAEA.
  • DC offset range needs to span full input range -- GH, 2009-09-18
    • Currently only spans about 50% of range. Op-amp cannot provide enough current for offset to all channels. Needs a component replacement by JPM.
  • Poor resolution of energy evaluation. Needs to be good. -- GH, 2009-09-18
  • Check if M=i^2 is a suitable set of values for the M parameter in energy evaluation -- GH, 2009-09-18
    • If not we would need to implement dividers on the pole-zero correction, at sampling speed. There arent' enough spare gates & nets on the current FPGAs to implement this for all 32 channels per mezzanine. Alternatives include making them TIG-32s.
  • Make zero-suppression mode work for more than 0th channel. -- GH, 2009-09-18
    • When zero suppression mode is enabled, only the 0th channel on each mezzanine works. 1-31 don't.
  • Channel-by-channel polarity? -- GH, 2009-09-18
    • CJP sez all channels need to have one single offset and one single unique polarity.
    • But JPM sez some of the channel inputs are reversed.
  • Map TIG-64 channels to match connector -- GH, 2009-09-18

[edit] TIG-10

  • TIG-10 reset -- GH, 2009-09-18
    • Doesn't reset correctly, leaves them in random and probably not usable mode that blocks everything else.
    • It's a little to get right, sez CJP.
  • Update TIG-10 firmware after everything else is done -- GH, 2009-09-18
  • Status word from each card -- CJP, 2009-09-18
    • To probe a collector or digitizer card to see if it's the one blocking everyone else
  • Read data out of master TIG-C
    • Otherwise we can't do the proper dead-time when there is pre-scaling at the master level.
  • Read firmware revision from card
    • Probably best to encode it in date form

[edit] Operations

  • Add Faraday-cup readout to Epics archive (CD 2009-09-17)

[edit] SHARC

  • Bambino cables - GLW 2009-08-20
    • The bambino cables don't plug into the preamps securely. The connectors are too long for the clamps on the preamp inputs to secure the cable. But maybe this will be solved during the move to Tig64s?
  • Target mechasim - GLW 2009-08-20
    • We couldn't plug in one of the CD detectors where we wanted to because of the size of the target mechanism. If another CD goes in, we will need all feedthroughs on the flange and so we need a different size target mechanism.
  • Barrel coverage around 90˚ - GLW 2009-08-20
    • Currently, the barrel has good coverage around 90˚, but not towards the upstream CD. Looking at the individual silicon detectors, the PCB surrounding the wafer is asymmetric. Currently, the 'fat' edge of the PCB is towards upstream, so we are missing out on some of the important proton data. The thin edge is at 90˚, but there isn't so much vital data in this region. I propose that we try to turn around the silicons, depending on if the brackets will allow it.
  • Preamp back strip->DAQ cabling - GLW 2009-08-20
    • The feedthroughs on the SHARC flange currently swap round the back signals (of upstream and downstream detectors) relative to the front signals. There was a lot of confusion about this (c.f. ELog entry 3895) because the connectors can't just go in the other way. If we end up changing the SCSi to 16pin ribbon connectors and still use the same feedthrough boards, we should maybe thinking about making some cables for the front strips (the right way around) and some for the back strips (ordered the wrong way around). Whatever happens though, I think it needs careful thought.
  • Different flange? (CD, 2009-08-26)
    • Should we have another flange with a larger exit tube?
  • Cross talk (Origin to be determined and evaluated): (CD, 2009-08-26)
    • Inside cable update?
    • Outside cable update? -> HV-board and connector?
    • Feedthrough update?
  • Numbers stamped on brackets (CD, 2009-08-26)
  • Rubber feet on detector boxes (CD, 2009-08-26)
  • Glue ribbon cables to connectors (CD, 2009-08-26)
  • Flange-mounting holes (CD, 2009-08-26)
    • Re-visit diameter
    • Screws should be mounted the other way around to ease access.
  • Evaluate procedures (CD, 2009-08-26)
    • Beam tuning
    • Target mounting
    • Cabling
    • Calibration (alpha and gamma)
Personal tools