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

Suggestion for routing the Bcal and Fcal cables




Dear collaborators,

During preparations for the system review we had some discussions about
cabling since how the cables are connected to the electronics can affect
the options for creating energy sums for the trigger. The engineering
staff is also starting to route cables and establish cable trays, so it is
not to early to consider their layout. Below are two suggestions for
discussion about how to map the Bcal and Fcal cables to the crates. These
cabling schemes are motivated by considerations for the trigger. Generally
a phi-symmetric trigger is desirable for our detector.

Comments/suggestions are welcome.

Cabling for the Bcal.
--------------------

General: For the trigger we want to cable opposite (left/right) ends into
the same fADC module so that sums will approximately compensate for
attenuation.  Adjacent modules should be cabled together to create local
sums of energy deposition.

The crate sums are delivered to the system processors and therefore are
created naturally the trigger now, without the use of additional data
lines. Therefore, it makes sense to cable the crates in a way that might
be useful for the trigger.

One cal wedge: Outer 4 x 2 (left/rigth) = 8 inputs
Inner: 24 x 2 (left/right) = 48 inputs
-> One wedge fills 3.5 fADC (16 channel) modules.
-> Two wedges fill 7 fADCs
-> Four wedges fill 1 crate (14 fADCs). This leaves 2 empty slots, which
might be sensible from the heat dissipation stand-point, but also
associates a single crate with 1/12 of the Bcal (30 degress in phi).

With this scheme we need a total of 12 crates for the Bcal (6 north, 6
south). Note that 10.5 crates would be used if all slots were cabled
completely (16 fADCs per crate). However, the even number of 12 splits
evenly between north and south which would be required by the present
concept for the location of racks.

I do not know if there is already a convention for numbering wedges or
crates, but I assume it makes sense to tie them to the coordinate system,
were y is up and x is pointing north, and phi clockwise relative to x.

north crates   phi (deg)
1              0-30
2              30-60
3              60-90
10             270-300
11             300-330
12             330-360

south crates   phi (deg)
4              90-120
5              120-150
6              150-180
7              180-210
8              210-240
9              240-270

In order to ensure coverage across the boundary between crates, the
trigger would need to create overlapping crate sums. These would include
1+2, 2+3, 3+4, 4+5, 5+6, 6+7, 7+8, 8+9, 9+10, 10+11, 11+12, 12+1. This
granularity corresponds to an azimuthal angle range of 60 degrees. For
example [I'm guessing here] a cosmic-ray trigger could be formed in the
SSP using opposing regions such as 2+3 * 8+9 or 3+4 * 9+10 or 4+5 * 10+11.
This concept should be checked by the trigger experts.

Cabling for the Fcal.
--------------------

General: For the purpose of making coincidences between the Fcal and the
FTOF, one might consider cabling the Fcal in either vertical or horizontal
columns (rows). Background rates, however, are a very strong function of
the distance from the beamline, and so it makes more sense to combine Fcal
blocks within concentric rings into the same crate. This also keeps the
azimuthal symmetry of the detector which I believe is a very desirable
feature of the trigger.

A rough division of rings (within two radii) is given below, where each
block is 4x4=16cm2. Each crate is assumed to house 16 fADCs (256 blocks)
For 11 crates this corresponds to 2816 blocks (nominal number is 2800).
The last crate corresponds approximately to a single ring of blocks at the
outer layer of the Fcal.

crate   area    radius(cm)
         (cm2)
1	4096	36.1081486
2	8192	51.06463346
3	12288	62.54114794
4	16384	72.21629719
5	20480	80.7402748
6	24576	88.44653962
7	28672	95.53318149
8	32768	102.1292669
9	36864	108.3244458
10	40960	114.1839917
11	45056	119.7571808

At the trigger level, different trigger thresholds or selection criteria
could be easily applied at the crate level. These thresholds should be
applied to everlapping sums to ensure that showers that straddle the
boundaries have the same efficiency in the trigger as showers in the
middle of a sum. This scheme with rings, requires the least number of
elements in the sum which is two. [Note that cabling in for example
quardrants will require four elements into the sum, since a shower could
in principle hit at the corner of for regions].

Energy sums which can be used for selection with a threshold would be the
sum of two crates, i.e. 1, 1+2, 2+3, 3+4, 4+5, 5+6, 6+7, 7+8, 8+9, 9+10,
10+11, 11+12, 12. This granularity provides a reasonable granularity even
for the simplest trigger scheme which is presently proposed.

This scheme requires cables around an entire Fcal ring routed to a single
crate. In particular this means that cabling would not be partitioned
north and south which might be the most elegant cabling solution.

-----------------

Elton Smith
Jefferson Lab MS 12H5
12000 Jefferson Ave
Suite # 16
Newport News, VA 23606
elton@jlab.org
(757) 269-7625
(757) 269-6331 fax