[Halld-pid] Time-of-Flight Meeting Minutes, March 28, 2018

[Mark Ito]

Friends,

Please find the minutes below and at

https://halldweb.jlab.org/wiki/index.php/GlueX_TOF_Meeting,_March_28,_2018#Minutes 
.

   -- Mark

_______________


    GlueX TOF Meeting, March 28, 2018, Minutes

Present:

  * *FSU: * Sasha Ostrovidov
  * *JLab: * Thomas Britton, Eugene Chudakov, Mark Ito (chair), Simon
    Taylor, Beni Zihlmann

There is a recording of this meeting <https://bluejeans.com/s/zCEdu/> on 
the BlueJeans site. Use your JLab credentials.


      Changing the Hit Factory

Beni has changed the TOF hit reconstruction code to use the FADC 
integral for the time-walk correction in the case that the amplitude is 
set to zero. Recall that this can happen when the voltage is above 
threshold at the beginning of the FADC read-out window. The ratio used 
is a hard-wired 0.2 (amplitude/integral). He analyzed a file both before 
and after this change and sees an increase in TOF efficiency. See 
logbook entry 3547610 <https://logbooks.jlab.org/entry/3547610> for 
plots that demonstrate the improvement. It can be greater than 10% in 
certain regions of the detector.

The timing resolution appears to be a bit worse with the change, but a 
quantitative assessment is still in progress. Part of this may come from 
an apparent shift in the timing calibration constants. This is not 
expected and not understood.


      Review of minutes from the March 14 meeting

We went over the minutes 
<https://halldweb.jlab.org/wiki/index.php/GlueX_TOF_Meeting,_March_14,_2018#Minutes>. 



        High Rates and Missed Pulses

There was further discussion of the presentation Beni made at the last 
meeting on losses at high rates.

  * There were "missing channels" where FADC read-out was completely
    apparently absent from the event for channels were a pulse was
    expected. This was a bug. Those pulses are indeed there in the data.
  * The FADC read-out thresholds have been raised for the TOF from 110
    counts to 160. This should improve efficiency. The level was chosen
    to be a bit below that of the discriminatory threshold (thus on the
    TDC side, note that the signal split is 50-50).
  * The feature of the firmware where "peak=0" is reported is allegedly
    fundamental in the algorithm and not easy to change.
  * We need to explore schemes where full waveform read-out is done,
    either selectively, or for every event, because of lost information
    at high rates.
  * Beni found out that there is an intrinsic dead-time in the
    discriminator of 20 to 25 ns, independent of input pulse width. This
    was chosen to match the dead-time of the F1-TDC. Since we are using
    CAEN TDCs in the TOF, this limits our double pulse separation
    capability. As a result, a discriminator width of 20 ns is used
    universally in GlueX.


        Amplifier-in-the-Base Tests

We need to study how to deal with the significant baseline shift 
incurred by the AC coupling of the amplifier-in-the-base scheme.


      PMT testing

In order to look the apparently poor quantum efficiency he was seeing 
for some PMTs with an LED light source, Sasha set up a test with an 
Americium source. With this set all PMTs give roughly similar and 
acceptable efficiency. The cause of the problem with the LED set-up is 
still under investigation.

Beni asked about gains. The new tubes have gains in the range of 4 to 
6×10^6 .


      Action Item Recap

  * Add items to the list of tests to perform with the remaining run
    time during Spring 2018:
     1. Test the effect of the raised FADC read-out threshold.
     2. Look with a scope at the discriminator input for the
        amplifier-in-the-base channel.
  * Initiate process of shipping tested PMTs back to JLab from FSU.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/halld-pid/attachments/20180401/ef2ba40d/attachment.html>