Detector Naming Scheme

By Dino Bazzacco
In the following I will describe the proposed numbering/labeling scheme.
I would like to generalize a bit the EUROGAM one as we have now a larger number of detectors of different type (we should not forget that there will be also quite a number of ancillary detectors). My idea is that the name of a detector is an alphanumeric string which can be converted to a number representing the detector. The conversion detector number to name should also be possible. There should be a certain flexibility and shorter forms of names should be possible to allow faster typing. Names are case insensitive.

The name of a detector has 4 fields:

First field one character for the type of detector Second field a number with one or more digits to specify which detector

Third field one alphabetic character to specify which individual element of a composite detector. This field is optional and defaults to: a. This field must be present in order to give explicitly the next.

Forth field an alphabetic string to specify detector type: This field is optional and has a default which depends on the character given in the first field. For gamma detectors the default is: Ge. 
Examples



For the conversion of the name string to a detector's channel we need some
system wide constants:

    Number of detectors of each type: MaxN
    Number of individual crystals for each type of composite detector: Mult

Furthermore we need to decide whether the channels are contiguous or whether
each detector type reserves values in a certain range. I suggest this last
possibility but the scheme is flexible and can be easily changed to have
contiguous values.

Considering now only gamma detectors we have:

MaxN(C)=15  Mult(C,Ge)=7  Mult(C,Bgo)=24
MaxN(Q)=26  Mult(Q,Ge)=4  Mult(Q,Bgo)=16
Maxn(T)=30  Mult(T,Ge)=1  Mult(T,Bgo)=10

To determine the channel of a detector we must still define in which range
to position them and this can be managed thorough an additional quantity
which gives the value for the first crystal of each type of detector.

If we call this quantity Offset we have:

det#=Offset(F1,F4)+(F2-1)*Mult(F1,F4)+(F3-1)

In order to determine Offset we can use a table or a kind of recursive
function like:

Offset(C,Ge)=   -1
Offset(Q,Ge)=   Offset(C,Ge)+MaxN(C,Ge)*Mult(C,Ge)
Offset(T,Ge)=   Offset(Q,Ge)+MaxN(Q,Ge)*Mult(Q,Ge)
Offset(C,Bgo)=  999
Offset(Q,Bgo)=  Offset(C,Bgo)+MaxN(C,Bgo)*Mult(C,Bgo)
Offset(T,Bgo)=  Offset(Q,Bgo)+MaxN(Q,Bgo)*Mult(Q,Bgo)

In this way the Germanium crystals will use 0...999 and the Bgo crystal
will use 1000....
If we replace 999 with Offset(T,Ge)+MaxN(T,Ge)*Mult(T,Ge) detector channels
will be contiguous.
Of course all this can be hidden into a single computer function and a function
for the reverse process Channel to name can also be easily written.

In practice there are some peculiarities which must be solved.
For example the Bgo shield of the clover detectors has 16 individually powered
PMT's but only 8 output signals as the PMT's are pairwise daisy chained.
In order to keep the numbering scheme one must account for this fact for
example imposing an equivalence of crystals a and b, c and d ..., when using
the name to identify the channel in the electronics or a spectrum.
The same problem exists for the Bgo shields of the clusters where we have
some pairwise coupled PMT's ...
All these complications can (or better should) be hidden to the user through an
appropriate computer routine (or wire file)

We will use this scheme also for labelling the cables in which case a fifth
field will be added in order to state its function (signal,Hv,Temperature...)