CALICE MAPS Meeting, RAL, 16/01/09 ================================== Present: Rebecca Coath, Jamie Crooks, Paul Dauncey, Owen Miller, Marcel Stanitzki, Renato Turchetta, Mike Tyndel, Nigel Watson, Gary Zhang Minutes: Paul Minutes of previous meeting: No comments. Sensor 1.1: Jamie showed some slides on measurements with sensor V1.1; see usual web page. The "very low threshold" means around 2mV, which would be ~5TU assuming 0.4mV/TU. The feedback oscillation rate is around 3MHz. If we want to make a corrected version 1.2, then it would be good to have confidence that we won't find further errors. The gain of 1.1 could be checked with a variable signal (i.e. using the lasers) and so allow the pedestal to be determined independently of the square-top threshold scan. Similarly, uniformity could be checked with a wide-spot laser illumination on the back of the sensor. The feedback capacitance could be removed with changes to the M2 mask. The row address bug requires changes to the CS mask. The configuration load problem (requiring an additional external 1.65V supply) is another design error and could be fixed; Jamie will look into which mask(s) would need to be changed. The power-ground shorts are still thought to be due to a foundry processing error (but they have not admitted this, although they continue to withhold the process data). The cost of the redesign would be 11k for the CS mask, 5k for the M2 mask, 10k for handling and 4k for 3 more wafers (we have 9 on hold and the minimum order is for 12). Other masks to fix the configuration load error would push this up further. Including VAT, the total is 35k. Neither CALICE nor CfI have this level of funds in this FY. It might be possible to submit this as RAL infrastructure (as calibration sensors) and Mike will check. Otherwise, it would need to wait until the next FY and be paid from SPIDER funds. However, this was not part of the proposal costing and would also delay testing by three months. For comparison, a full four-seat shuttle run would cost 56k. Paul showed some slides on threshold scan measurements; see slides on the usual web page. He has not yet looked at the few pixels which have an RMS of ~6TU to see if they look Gaussian, i.e. do not show feedback effects, but will do so. The pickup is probably reduced in 1.1, and it seems around 300 pixels can be enabled before it occurs. Jamie commented that he suspects the reset buffers may be causing (at least some of) the remaining pickup. These share the monostable power supply. Separating the power supplies of these two could be difficult as they are implemented as an integrated unit. [Note added after meeting: The pixels with an RMS of ~6TU are not Gaussian but look flat-top like all other pixels. They are all along the edges of the sensor, at y=0 or 167.] Marcel showed some slides on test pixel noise measurements; see slides on usual web page. The "voltage range" here refers to the oscilloscope range selected. He now believes the actual noise is around 2.2mV, rather than the 3.6mV quoted in the previous meeting. The peak seen in the fast Fourier transform (FFT) is at 350MHz, which was thought to be so fast that it would have to originate from some external source. The calibration and comparison of the test and bulk pixels is still not understood. The two test pixels (old and new design) still indicate the Fe55 peak is around 200mV, so giving 1mV ~ 8e-. However, this would mean the noise would be only ~18e-, which is smaller than expected by around a factor of two. To get consistent noise values would require the Fe55 peak to be around 100mV. Marcel will try taking runs with several thresholds to check the shape above threshold is consistent and independent of the threshold; the issue will be to normalise the different runs. He will also do an Fe55 run with the bulk pixels on sensor 1.1, so as to cross-check that the response is around the same in terms of TU as for sensor 1.0. For this, he will always mask the upper half of the sensors to remove any ambiguity from the row address error. Conferences and talks: Marcel gave a seminar at Birmingham two days ago and slides have been added to the general CALICE-UK web page. Marcel reported the SiD LoI is already at 160 pages, although the target is only 100 pages. He will attend the SiD meeting in SLAC in Mar, using LCUK travel funds, and the AFCA meeting in Tsukuba in Apr, which is where the LoIs will be submitted to IDAG. Paul and Anne-Marie have just got a stand-alone GEANT4 program going and so Paul is now starting the pi0 study for SiD. Nigel and Owen are looking at photon resolutions for the ILD LoI. Paul and Nigel were approached by a CALICE collaborator, Roman Poeschl, who is helping to organise a calorimetry summer school in Beijing; see email linked from the usual web page. Roman would like someone to give three 45 minute lectures on DECALs at the school; these are scheduled for Sun 26 Apr. Paul asked for anyone interested to contact him asap. Both Nigel and Marcel may be unable to go to this meeting. There are no other conferences coming up before the end of the FY. The usual calls for abstracts for the summer conferences (IEEE, etc.) will come soon and people should respond to these as normal, on the assumption that either SPIDER or their institutes will fund the travel. The total travel budget for CALICE in this FY was 49k. Next meeting: This will be at 9-11am on Fri 6 Feb, before the general SPIDER meeting which starts at 11am.