[Halld-cal] Dark hits in BCAL

[David Lawrence]

Hi Will,

  I was chatting with Elton today about your presentation at the calorimetry meeting this morning.
I’m sorry I missed it.  From my discussion with Elton and from what we talked about at the 
Calibration meeting on Wednesday, it sounds like the simulation of the dark hits is simply
giving a much larger contribution that what we are seeing in the real detector. This likely due
to the conversion factors used to convert photo-electrons and energy into mV not being correctl. 
As was discussed, these factors need to be adjusted in order to bring the simulation into better
agreement with the real data. 

 The relevant factors in mcsmear/smear_bcal.cc are the following shown with their current values:

BCAL_mevPerPE  =  0.668338
mV_per_MeV     =  0.881599 

 The value of mV_per_MeV is inversely proportional to BCAL_mevPerPE (the latter being used
in the calculation of the former). The dark hit amplitude and the signal amplitude are both multipled
by mV_per_MeV. Only the dark hits though are proportional to BCAL_mevPerPE. What this 
means is than in order to adjust the relative amplitude between signal and dark hits, without
changing the signal size in mV, both of the above parameters have to be modified.For example,
to effectively reduce the dark hit amplitude by a factor of 10 relative to the signal, we divide 
both BCAL_mevPerPE  and mV_per_MeV by 10 in the code:

BCAL_mevPerPE /= 10.0    =>  0.0668338
   mV_per_MeV /= 10.0    =>  0.881599 

note that the value of mV_per_MeV doesn’t actually change since the value of BCAL_mevPerPE
is used to calculate it.

I’ve made a few plots to show a comparison of with and without this change. You can see for the 
second plot (the logY one) that the signal part of a cell energy (above about 75MeV) is not
affected much. There is significant reduction in the lower energy part though. The last plot shows
the reconstructed cluster energy. This indicates a drop of about 10% in the reconstructed
energy (simulation was all 1GeV photons). I think this is what you were alluding to when you
said simply removing the dark hits from the cells with signal caused issues since the reconstruction
has been tuned to put the peak at the right place when dark hits are present. Either those parameters
will need to be re-tuned, or the value of mV_per_MeV scaled by a smaller factor than BCAL_mevPerPE.

The scale factor of 10 was arbitrary and the actual factor should probably come from a comparison
to the real data. I think we have two handles by which to do that. The first is the number of hits (first
plot). We could reproduce that plot with “no-signal” events and adjust the scaling factor until it matches
what we see from data taken with the detector using a random trigger. The second would be to convert
GeV into Pulse Integral fADC units and look at the “dark hits pedestal” width. This would again need
to be done with no-signal data. Hopefully, the same scale factor will give agreement in both, though
we may need to tweak the threshold to get that.

I’m not sure how much time you have to work on this so we can discuss how best to proceed.

Regards,
-David
























-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://mailman.jlab.org/pipermail/halld-cal/attachments/20141024/0655f589/attachment-0001.html 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Nhit.png
Type: image/png
Size: 18831 bytes
Desc: not available
Url : https://mailman.jlab.org/pipermail/halld-cal/attachments/20141024/0655f589/attachment-0005.png 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Ehit.png
Type: image/png
Size: 18341 bytes
Desc: not available
Url : https://mailman.jlab.org/pipermail/halld-cal/attachments/20141024/0655f589/attachment-0006.png 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Ehit_zoomed.png
Type: image/png
Size: 21705 bytes
Desc: not available
Url : https://mailman.jlab.org/pipermail/halld-cal/attachments/20141024/0655f589/attachment-0007.png 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Ncluster.png
Type: image/png
Size: 16331 bytes
Desc: not available
Url : https://mailman.jlab.org/pipermail/halld-cal/attachments/20141024/0655f589/attachment-0008.png 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Ecluster.png
Type: image/png
Size: 18909 bytes
Desc: not available
Url : https://mailman.jlab.org/pipermail/halld-cal/attachments/20141024/0655f589/attachment-0009.png