Attendees:
LBNL: E. Hjort, R. Snellings, D. Olson, N. Xu, F. Wang, P. Jacobs, H.G.
Ritter, I. Sakrejda, S. Vaniachine, Q. Li
BNL: D. Hardtke, H. Wieman. (I forgot to ask who were present so I may
miss other names here.)

Geometry transformations:
-------------------------
Dave H. discussed geometry transformations among Pad, LocalSector,
Local, and Global coordinates. He illustrated the transformations and
database access with a simple example. (The geometry constants are
stored in database StTpcDb.) One question from Peter J. that needs
attention is the issue of physically invalid location after a
transformation, e.g. Local->Pad.

Dave proposes a wrapper for the transformation functions (in C++) so
they can be used in TPC pams Fortran codes.

Outstanding question is where to put the transformation functions:
StTpcDb class or some other utility class. No final word yet. Another
outstanding issue is that the electronics Db is not available yet.
Should be ready soon.

Gain study:
-------------
Eric H. presented his gain study. He looked at cosmic data tracks that
have the lowest signal/noise ratio, ie. those tracks with long drift
distance (>1.5m) and close to zero cross and dip angle. He made
corrections due to that (1) the tracks are not MIP (20-30% correction),
and (2) the cross and dip angles are not exactly 0 (a couple of percent
correction). He suppressed noise (20<time bucket<400) and used
(offline?) Asic setting 1,3,3,1.

Most probable ADC values are 30 (inner) and 22.5 (outer). The noise
level is 1.25. The gain constants obtained are 3558 (inner) and 1310
(outer). Voltage used were 1169 V (inner) and 1389 V (outer).

The ADC distribution of inner sector is broader than of outer sector. It
may be explained by the large difference in the pad length (1.15 cm
inner vs 1.95 cm outer, checked by Iwona after meeting).

Grand Challenge:
----------------------
Doug O. discussed GC status. Sasha V. has implemented ~500 tags for
testing the GC mechanism. It was made clear that GC would access TagDb
and return a list of events satisfying a user-query (these events are
retrieved from HPSS to disk).


Minutes by Fuqiang W.



Question from Roy Bossingham:
 - Could you add the magnetic field for the gain study?
 - The effective pad length is probably closer to the pad pitch
   than the length of the copper pad per se; clearly, the shorter
   pads will have wider dE/dx fluctuations, though I don't have
   hard numbers at hand.  Does it work out quantitatively?

REply from Fuqiang:
I forgot to mention the field in Eric's study in the meeting minutes sent out
yesterday. It is full field 0.5 T.

Eric's answer:
It's on one of my slides, which Iwona is putting on the web with the
minutes.  (it was 0.5 Tesla)

>
>  - The effective pad length is probably closer to the pad pitch
>    than the length of the copper pad per se; clearly, the shorter
>    pads will have wider dE/dx fluctuations, though I don't have
>    hard numbers at hand.  Does it work out quantitatively?

From Fig 1.10 or Eq. 1.38 in Blum & Rolandi I wouldn't have expected as much
difference in the shapes as I measured.  Maybe this is due to marginal
statistics on the inner sector or because it's not a perfect comparison (Ar
vs. P10, gamma is not 1000 for our data, etc.).  I haven't tried to be more
quantitative - do you have a method in mind, or perhaps another reference?


Roy's remarks:

for the purpose of smoothing ionization statistics, transverse diffusion
and coupling of the anode wires to adjacent pad rows also increase
the effective pad length.

From the data, one might check how much of the inner sector width
is due to variation along the pad row; as you recall (and probably
were the one who measured), some increase in the gain was seen toward
the middle of the anode wires.