Results of testbeam with Marebo chip
The 'Marebo' chip has been measured in the H8 test beam at Cern in may 1998.
They are 50 x 432 um^2 pixels bonded on a 300 um thick Si detector.
- Data taken at 0 degree, with vertical pixels
- Analysis program of Francesco Ragusa (adapted to Marebo)
- Cuts applied
- Fiducial volume: rows 2 to 62 (incl.) and columns 1 to 10 !
- The expected pixel is not dead (from dead list)
- Some bad columns are removed (bit missing).
- Col 1 and 3 for Marebo 1
- Col 4 for Marebo 2
- Col 4 for Marebo 3
- Because of resolution and problem with sensor at the edge
of the pixel (along long side), events are selected in a box
of +/- 20 um (rphi) and +/- 180 um (z)
- Probability of track Prbx > 0.02 and Prbx > 0.02
- Nb of hits between 0 and 2 (incl.)
- Reject events where the hit is lost due to the priority scheme. This is done
if the same pixels gives a later hit carrying TOT information. The event is
rejected.
Compared with FEA-FEB analyses, cuts 1, 5, 6 are standard.
Cut 2 is needed to reject those bad pixels with process failure (dead pixels).
Cut 3. These columns show efficiency problem. Col 4 had clearly bit 1 missing.
Cut 4. Minor cut to improve efficiency at edge of pixel
Cut 7: This gains between 1 and 2.5 % of efficiency depending on the chip. This may
be related to the beam intensity i.e. the chip activity. No late bit was used in the
Marebo chip.
MTh
