CALICE MAPS Meeting, RAL, 17/05/07 ================================== Present: Jamie Crooks, Paul Dauncey, Anne-Marie Magnan, Yoshi Mikami, Owen Miller, Vladimir Rajovic, Marcel Stanitzki, Konstantin Stefanov, Renato Turchetta, Mike Tyndel, Giulio Villani, Nigel Watson, John Wilson Minutes: Paul Minutes of the previous meeting: No corrections. Paul has not put the "Chronopixels" talk link on the website yet; Marcel will send it to him. Sensor design: Jamie reported that the sensor is now expected at the end of June, which is earlier than we thought. However, he is not sure if this date is for completion of fabrication or for us to receive it so there is an uncertainty of around a week. He should get a firmer date from the foundry soon. He also confirmed there will be no metallisation of the substrate so there should be no issue with shining the laser on the back. Before submission, Jamie had time to complete the pre-sample test structure but not the pre-shaper one. Much of the circuit is common to both so this allows a large amount of the design to be tested anyway. He is currently going through and clearing up the documentation. There are updates of the review docs needed and also he will redo much of the simulation with the exact final design, including parasitic capacitors. Renato reported that the foundry had measured the deep p-well properties and the results were in line with their expectations. This means they are more confident about the process and so have recommended we no longer need to do a high dose deep p-well wafer split. This allows us to make more of the "usable" wafers and so the 12 wafers will now be split as o 6 wafers: 12mu epitaxial, deep p-well (the "usable" sensors) o 2 wafers: 12mu epitaxial, no deep p-well o 2 wafers: 5mu epitaxial, deep p-well o 2 wafers: 5mu epitaxial, no deep p-well Each wafer will have 20 sensors. Hence, the 6 wafers should give over 100 sensors if there is a reasonable yield. Marcel raised the issue of cost, as SiD are quoting a price of $2/cm2 from Hamamatsu for silicon diode pad wafers in large quantities. This was thought to be a nonsensical figure as there is no way to know the cost of a production in 10 years. The financial case for MAPS is that is would be sourced as a multi-vendor contract which then requires the foundries to complete for the business. Tying into one supplier would give no control over cost or schedule. Sensor simulation: Giulio has not yet started the simulation which is needed for the LCWS physics studies. The PC farm has been upgraded from 2 to 18GBytes virtual memory and this could be pushed futher. However, he has not simplified the n-well layout to speed up the simulation and so each run will take 4-5 days, with 10 runs being able to be done in parallel. This will make finishing all 21 points in time for LCWS almost impossible but he should do as much as possible in the available time. Sensor tests: There was a schematics review of the sensor PCB in the morning. The design looks to be in good shape and most comments were pretty minor. The biggest issue is the 220-pin connector(s) for the cables to the USB_DAQ board. The next step for the sensor PCB is the layout review on Tue 26 Jun, starting at 13.00. Vladimir suggested that he could bring this forward, but it might be better to wait until he has solved the technical issues of the PCB layout before rearranging the date. Also, Jamie should have a firmer date on the sensor return in a short time. Obviously, it would be good to speed up the schedule but we can see how things develop over the next week or so. Matt Noy was unable to come to the meeting so there was no news on the USB_DAQ. Paul has arranged to use an existing 100uC strontium-90 source at Imperial for the sensor source tests. The PDR estimate was that we needed a 3uC source to get a sufficient rate so the Imperial source should be usable. Strontium-90 is a beta emitter and the system will need a scintillator "trigger" (in fact, time tag as the system will run in ILC bunch timing mode) on the other side of the sensor so as to measure the bunch crossing of the hit. This scintillator still needs to be found. For the cosmics system, John is also looking for scintillators. He wants thin scintillators so as to define the solid angle well. They are considering light guides to avoid putting PMTs too close into the sensors. It might be worth looking into using SiPMs which we could get from the AHCAL group. Giulio reported that the LabView-based software to drive the laser is working. He needs some hardware to do the laser calibration and this will be made by June. For the use with the CALICE DAQ system, the laser and stage both need to be controlled from the DAQ via a socket, rather than LabView, and this layer of software has not been started yet. This was estimated to be around a week's work. Physics simulation: Paul went through the analyses required for LCWS; see slides on the usual web page. The subpixel cell size was fixed at 5x5mu2 and so the simulation production has now started. There are four major analysis topics (as listed in the previous meeting) which still look possible to do. These were divided up as follows: o Simulation parameter dependences: Anne-Marie o Understanding source of resolution: Yoshi o Linearity and resolution vs energy: Anne-Marie o PFA comparison with diode pads: Marcel Results are needed by the end of next week and should be circulated by email before then for discussion. Yoshi showed some results on the latest Mokka version 06-03-p02; see slides on usual web page. He sees the problem with small cells is possibly reduced compared to 06-00 but has not been cured completely. It is not due to the rangeCut parameter. He also sees a significant tail in the diode pads simulation for high energies, which is not present for MAPS. This is hard to understand if the tail is due to a physical loss of energy such as leakage; this needs more study. Anne-Marie showed results using the same version of Mokka; see slides on usual web page. She has produced the resolution vs threshold plots with more detail on the various modelling steps. Her dead cells are done with the 5/47 ratio corresponding to the sensor being fabricated. Marcel showed results on various topics; see slides on usual web page. He has been running a MAPS sensor from a different project and showed some results using an Fe-55 source. This is a gamma source and so gives no timing information, unlike the planned Sr-90 source we will use. Hence, the background is probably harder to reduce. He has also tried to look for multiple hits in a pixel by fitting the Landau shapes for more than one MIP peak. The fit results vary in the right direction but it is difficult to calibrate the fit without some knowledge of the expected results. Finally, Marcel had a first try at the "no harm" PFA study of summing pixels over an area equivalent to the diode pads. This is only with SimCalorimeterHits so far and he has not included pixel clustering but obtains 35%/sqrt(E), compared with 30% for the standard option. Anne-Marie noted there was a GEANT4 "feedback" meeting the week following LCWS (the location was not known). If anyone was able to go, they could report on the small cell size problems; however, there was no one who was definitely available. Conferences: Marcel has set up a conferences web page, accessible from the main CALICE-UK web page or directly at http://hepwww.rl.ac.uk/calice/Conference-planer.html We have one talk at each of LCWS and EPS and have an abstract submitted to TWEPP and two to IEEE. Renato is talking to the Vertex07 organisers for a talk there also. Of these, the people likely to attend are LCWS - Anne-Marie, Konstantin, Marcel, Mike, Owen, Paul, Yoshi EPS - Marcel (?), Nigel (?) TWEPP - Giulio, Jamie (?) IEEE - Giulio, Jamie, Marcel Of the LCWS attendees, Anne-Marie was the only one who actively wanted to talk so she will do this one. For EPS, Nigel thought he could attend for a day to give the MAPS talk if no one else volunteers before the conference. Note, those people at LCWS should all try to come along to the informal discussion with SiD which is being set up by Marcel. Marcel is giving a talk on "UK Ideas for ILC Detectors" at the IoP half-day meeting in Oxford next Wed 23 May. He will need to include at least the DAQ and MAPS from the CALICE side. He is following a talk by Ties Behnke and so does not need to do the basics of detector requirements. He will give a practise on Tue 22 May at a time which he will circulate round beforehand. Next meeting: Tue 19 Jun at 13.00 in a PPD meeting room. The week following will be the sensor PCB layout review on Tue 26 Jun at 13.00.