CALICE MAPS Meeting, RAL, 25/06/09 ================================== Present: Jamie Crooks, Paul Dauncey, Marcel Stanitzki, Nigel Watson, Gary Zhang Phone: John Wilson Minutes: Paul Minutes of previous meeting: No corrections. Paul contacted Chris Parkes about Diamond and he said very general things about it being easy to use and the staff being very helpful, but did not give much practical information yet. Paul hopes to catch him in person at the Birmingham meeting in a few weeks. Marcel will start using the CAEN power supply soon so as to understand how to use it well before the beam test. Sensor testing: Jamie has set up a table of the sensors in the SPIDER Confluence area, under "TPAC/ TPC PCB Register"; see link from the meeting page. The status is that the four have been trimmed, one is currently being trimmed, four are ready to be trimmed and four are being bonded. There are a further two hi-res which have been made. Of these, the hi-res and two of the non-hi-res are glued for laser use, the rest are usable in the beam test. We should aim for at least ten trimmed and fully working sensors for the beam test, including two hi-res ones. These should be potted before being taken out, which can be done at Imperial, and it would be useful if the potting can leave a gap in the centre of the sensor for Fe55 tests. Jamie is modifying the trim setting resistor R75, as discussed at the last meeting, from 100kohm to 180kohm. Paul confirmed this had also been done on sensor #29 which is at Imperial. Paul will take sensor #50 to Imperial today. Paul showed some slides on results of trimming the four sensors done so far; see the usual web page. Jamie suggested the extra noise might be related to the number of bits set in the trim and Paul will check this. Jamie will check what the jumper setting used means for the ground of the substrate, in case noise is being picked up through this route. Paul will also go back and check whether TPAC1.0 shows similar effects. John asked about whether the noise vs trim plots on slide 12 showed any subtle trends and Paul will make averaged (profile) plots to check this. It is also worth checking the skewness of the individual pixel threshold scans from the trim to look for sources of the asymmetry. Marcel showed some slides on the Fe55 results and understanding of the calibration; see usual web page. It is not clear what the meaning is of the last digit in the intensity table but Marcel will check. He sees a gain of 0.155-0.160uV/e- with a peak width of 25-35e-. The 18mu hi-res sensor gives a very odd distribution but this will be checked; Marcel should also check the grounding jumper for the substrate as the hi-res may not be able to deplete the epitaxial layer if it is left floating. Overall, Marcel finds that 2.3TU = 1mV, or 0.43mV = 1TU, which is consistent with Jamie's measurements from the meeting on 06/02/09, as shown on his slide 6. With 0.157uV/e-, then this means 1TU = 2.7e-. The average noise determined from a threshold scan is ~7TU so this would be ~19e-. Beam test preparations: A count of the available equipment was made: Imperial: 3 USB_DAQs, of which one is faulty 3 sets of thick TPAC cables 3 mini-USB cables Birmingham: 3 USB_DAQs, of which one may be faulty 1 set of thick TPAC cables 1 mini-USB cable RAL: 4 USB_DAQs 4 sets of thick TPAC cables 6 mini-USB cables Bristol: 1 USB_DAQ As only ten V2 USB_DAQs were made, then the Bristol one may be an older version, so Paul will check this. Marcel will order ten more mini-USB cables and Paul will organise the fabrication of more thick TPAC cables from the reel at Imperial. Marcel will also give the spare two of his mini_USB cables to Nigel. Paul has got estimates of prices for more USB_DAQs and adapter boards. There are components and PCBs for twenty USB_DAQs already at Imperial, which were purchased in the previous FY. In small quantities, these would be ~150 pounds each to assemble, or around 2 kpounds for all twenty. The adapter boards would have to be made from scratch; this would be roughly 100 pounds each for components, 100 pounds each for PCBs and 150 pounds each for assembly, making 350 pounds per adapter board total. Hence, e.g. four would cost ~1.5 kpounds. Nigel thinks the stack can hold at most six TPACs, assuming no tungsten but allowing for PMTs at the front and back. Hence, we should aim for six layers at CERN, which requires seven USB_DAQs plus at least one spare. We should be able to cover this while leaving at least one system at RAL for the two weeks of the beam test. This means there is no absolute necessity to make more USB_DAQs and adapter boards. Birmingham are making a board for discriminating the PMT outputs and also performing logic on the signals. It will output both the single hits and logic outputs in both TTL and LVDS format. The bottleneck currently is the 40MHz clock and bunch train start distribution. The only sensible way to do this is to change the USB_DAQ firmware. Rui now has succeeded in getting the firmware to compile but it fails to route. Paul will try to get him in closer contact with Matt to encourage this along. Assuming the ability to run multiple USB_DAQs soon, then the next step is to assemble a system of more than one TPAC. This will be done at Birmingham, using a NIM crate to discriminate the PMTs and with up to four TPACs. When this is running and the Birmingham interface board is working, then the system will be taken to RAL for use with the CAEN power supply and also to build up to the full six layers. It will then be shipped to CERN by Jaap and Rui from RAL. Sensor simulation: Gary reported that there is a new simulation farm at RAL which means the sensor simulation should run 3-4 times faster than previously. This should allow the standard 21 points to be done within one week. The simulation will help understand the hi-res Fe55 fake peak; with the 21 points, Paul can calibrate his simple diffusion model and then generate the ~3000 points required to estimate the overall spectrum. Next meeting: This will be at Birmingham on Thu 16 Jul, starting at 10.00. (In case this needs to be changed, the fallback is Tue 21 Jul). There is also a general SPIDER meeting next Wed 1 Jul at RAL.