[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [GlueX] factors in bcal readout decision




Hi Richard

Thanks for posting the table of design considerations. I have merged your
table with the one that Alex sent me and posted a bare-bones draft of a
document that can summarize our evaluation on the portal (GlueX-doc-795).
It will be updated as we work toward the workshop on April 23-24. If there
are any comments/suggestions please send them to me.

From now on I will be posting discussion of this topic to halld-cal for
any interested participants.

A note on collection of light into smaller SiPM arrays: From a cost-point,
one of our main problems is the increase of a factor of 4 in the number of
light sensors for the bcal compared to the cost book. Therefore, we should
be considering how to make larger Winston Cones to match the current
1.2cm^2 SiPM arrays to reduce the number of channels. To make the obvious
point: if we have 4200 channels, all costs per channel add up to lots of
$$. This includes cables, low voltage supplies, ADCS, TDCs,
discriminators, etc.

Cheers, Elton.



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

On Tue, 3 Apr 2007, Richard Jones wrote:

> Colleagues,
>
> Please find attached my spreadsheet derived from our discussion of the factors driving
> the choice of readout scheme for the BCal.  The items in red are highlighted because
> they are potential performance drivers.
>  1. In the case of the planacon, the relevant scale is the sector of a device devoted
>     to one readout channel, eg. a 2x2 pixel section of a 8x8 anode device.  The global
>     variation of 1:1.5 is so large that local variations must be looked at closely. It
>     can be addressed by adding additional diffusion to the light path to scramble the
>     map from the BCal to the cathode surface.
>  2. In the case of the Sensl module the uniformity is probably very good (modulo
>     thermal gradients), and the dark rate is the primary concern.  If the photon yield
>     is really as large as 8 avalanching pixels per MeV then I really believe that
>     100MHz is no problem. We should be able to confirm this quickly using the test
>     beam data and the measured SiPM pde.
>  3. If photon statistics are as superabundant as we believe then there is zero
>     motivation to cool these devices below 20C.  Rather, we can save major costs by
>     reducing the 4x4 pad readout device with a 2x2 device.  How much cost reduction
>     does this entail?  This possibility might by itself tip the balance decisively in
>     favor of the SiPM readout.
> Richard Jones
>
>
>