Links related to the MVD pedestal problem.

Many links are from John's electronic links page.

The problem is shown here, in SangSu's amu cell plot These are real data. The mean value is just due to the way the ADC operates. The correlator is on. The AMU cells are filled every beam clock, and the pre- and post-samples are taken from cell n and cell n+2, where n=1-64.

I - AMU/ADC control lines:

Dick suspects the AMU/ADC operation. Therefore I collected some info on them:
From Timing and Control Requirements for a 32-Channel AMU-ADC... 

          ADC    AMU write    w/correlator
    ----------------------------------------------------------------
        raddr4       -        raddr4
        raddr5       -        raddr5
        amprst       -        amprst
        rena         -        rena         read enable
        az           -        az           auto-zero
        cmprst       -        cmprst       
        dbrst        -        dbrst
        ics          -        ics          internal comparator switch
        clken        -        clken
        fsc          -        fsc
        load         -        load
        addr         -         -
        addr         -         -
        addr         -         -
        addr3        -         -
        addr4        -         -
          -         wen        -           write enable 
          -         waddr0     -
          -         waddr1     -
          -         waddr2     -
          -         waddr3     -
          -         waddr4     -
          -         waddr5     -
          -         raddr0    raddr0
          -           -       raddr1
          -           -       raddr2
          -           -       raddr3
          -           -       conn      (dis)connect the difference amp
        ----------------------------------------------------------
These are some measurements taken on the PC-MCM:

Some key control signals for the ADCs

Scope traces of key output signals from AMU/ADC chips

II - Feedback

Pat suspects feedback, possibly from the MCM surface mounts to the Kapton cable. Note the pads don't have a Kapton cable.

Here are noise measurements from John




Here are notes and questions from yesterday and today, some of which have answers in the links above... 

     test                          what it checks
   ---------------------------------------------------------------------
   all channels disabled           test feedback from MCM to kapton cable
   compare IB to OT packets        test feedback, shortest to longest kapton
   correlator off                  is trouble at lower/upper end of range?
   amu/adc settings                setup OK?  
   reload MCM before calib         tests if slow drift is due to losing a 
                                   digital clock tick, getting out of phase
   for a given packet, how many different AMU patterns are there? 4 or infinite?

   show same packet, different calibs
 
   can we see this on the bench?
   
   amu spy line available for bench testing?

   get the timing diagram  of all control lines to AMU/ADC

   old pictures of this on the bench to confirm that this is new 

   temp data bench vs now on the MCM

   what were conditions for SangSu's  amu picture
     - real data - 
     - corr on/off
     - AMU cell gap for pre/post sample (1 or 2) 
     - offset (440 ch) from where?
     - clocks to the AMU (2 pecl+ttl)

missing : got a dcim in 1008/ 510 cheezy etc does not work right now.