Re: SiPM vs Planacon

[Elton Smith <elton@jlab.org>]

Hi George,

Since my trip to Paris got delayed... yeah I know a bummer.... I had
nothing better to do but to go over the document you sent. Might as well
make some comments in advance of the collaboration meeting.

Segmentation and Floor term:

Any uncorrected difference in gains between channels will certainly
contribute to the floor term in the resolution.  The contribution to
sigmaE/E should behave roughly as sigmaG/G, where G is the average gain
per channel, and sigmag~sigmaG/sqrt(n), where g is the gain of an
individual channel, and n is the number of channels that contribute to the
measurement. For discussion sake I take n~5 (is this roughly right?), so
sqrt(n)~2.2. Assuming we want to keep the floor term to less than 1%, this
implies that sigmag/G <~2.2%. Any non-uniformity in the light collection
should have the same effect. But it seems to me that one can correct for
this by including the correct weighting factors when computing the energy.
Am I not correct? In other words, the non-uniformity does not have an
intrinsic addition to the resolution. Or am I missing something?

Added note: Your comment later about the planacons and light collection
through the light guides is certainly valid in this regard if individual
gains cannot be adjusted. This is a question for Carl: Can one adjust the
gains of the 16 individual segments of the planacon? As far as I remember
there is a single HV input, so what is the gain uniformity?

SiPM option:

General comment: From the data I have seen so far of the SiPM tests, my
main concerns are gain and efficiency, not necessariy DR. However, Carl
may have some more data for us soon.

I)

Gain: Accordint to Carl (in coversations with SensL), the gain of the
SiPMs is actually 10^5, with 10^6 only after an amplification stage, so
not much different than the planacon tubes.

Photoelectrons per MeV(p.3 second bullet): From your numbers, 16 p.e./MeV
energy desposited, but 20 MeV-> 2.4 MeV deposited, so I conclude that we
get 1.9 p.e./Mev incident energy. This seems off by a factor of 4 from the
number at the top of page 4 that gives 8p.e/MeV incident energy. What is
missing?

II)

One item missing from the list is the issue of heat dissipation inside the
magnet. This would be a similar problem for both SiPMs and planacons,
except that the planacons are outside the solenoid enclosure so bringing
the light out has the benefit of getting the electronics out of an
enclosed space. There were also questions about 15% humidity controls...
this also needs to be addressed.

DR: With all the concern about the dark rate, it appears that the rates
are actually acceptable for the calorimeter. It would be good to specify a
maximum dark rate at a temperature above the due point (or where one can
reasonably cool). If the SiPMs meet this spec we are finished, no need to
do much better.


Planacon options

I

Linearity: Should this be 300 microA?

II

Nonuniformity: This is a very valid point, both for the light guides as
well as gain control of each channel. (See my first question/comments)

Conclusions:

While I am very hopeful that the SiPMs will meet their expected
performance, so far I have not seen the data to validate our optimism.
Perhaps you and Zisis have some data from SensL that corroborate your
assessment.


Enough for now. Cheers, Elton.



Elton Smith
Jefferson Lab MS 12H5
12000 Jefferson Ave
Suite # 16
Newport News, VA 23606
elton@jlab.org
(757) 269-7625

On Wed, 14 Mar 2007, George Lolos wrote:

> Hi Elton and Elke:
>
> Here's the critical review of the two BCAL read-out options.  I believe
> that I have addressed most - if not all the issues - but I consider my
> job done.  No more work out of me until the Collaboration Meeting :-) !
>
> Cheers,
>
> George
>