Cosmics with high Max ADC, settings as pannel before. The baseline is now also for this signals stable. In addition there is now space to change the SAS-voltages on the readout-board, because both voltages away from 0 [V].

 Cosmics seen by test-setup with new FEE-Card resistors SV=1750 [V] SR=100 kOhm, BR = 180 kOhm SAS-VrefF=-2.97 [V] , SAS-VrefP =-2.78 [V]. With this settings we raise our power consumtion per Card from ~2 Watt to ~2.3 Watt an 15 % effect which can be done by the cooling system.

 Base line problem for larger signals settings as before.

 Cluster evolution in a padrow for the test-setup settings as below.

 Cosmic shower seen by the test-setup, settings as below.

 Cosmics seen by the test-setup SV 1750 [V], R=150 kOhm shaper R, SAS-VrefF=2.91 [V] SAS-VrefP = 0.0 [V].In the first two pictures the well separated MIP's in one padrow (only one Pad per cluster is shown). The third picture shows the merged cluster of the two MIP's on the second padrow, but still possible to deconvolute such cluster. The fourth picture shows a typical noise on one pad with this configuration.

 Typical noise for a randomly selected Pad with no Laser signal nearby.Our maximum noise is on the level of 2.5 [ch/bin] with an RMS/Pad <= 1 [ch]. This is still not optimal but we hope to improve this by shielding the 1 m long data cable to the Readout-Board.For the given setup and with an max. ADC signal of 120 [ch] for an MIP we would end with 48 < S/N < 60.

 We tried to minimize the undershoot, in addition we did an arithmetic mean for the ADC of to consecutive timebins to simulate our 200 ns/bin clocking of the FEE-Card.Also wee calculate back to an 8-bit ADC with an non-linear function to estimate our signal to noise ratio.

 We got first Laser-signals for our test setup using STAR-MicroDaq + TPC Readout-Board + FTPC FEE-Card.The steering voltages for the FEE-Card aren't optimized to minimize the undershoot.