CALICE MAPS Meeting, RAL, 10/11/08 ================================== Present: Jamie Crooks, Paul Dauncey, Marcel Stanitzki, Renato Turchetta, Mike Tyndel, John Wilson Phone: Owen Miller, Nigel Watson Minutes: Paul Minutes of previous meeting: No comments. Sensor 1.1: Jamie showed some slides on measurements with sensor V1.1; see the usual web page. Sensors from one of the wafers show a lot of mechanical wirebond failures, which might be due to the metal layer for the pads being too thin. The other wafers are better in this regard. Jamie sees configuration errors on sensor #36 only in region 0 while Paul and Matt find they are only systematic for sensor #37 in region 3 (although there is also a single intermittent bit error in region 1). In both cases, the faulty region reads back only zeros. These two are being swapped so Jamie and Paul/Matt can cross-check the other's results. Sensor #36 was the one used for an Fe55 scan over the weekend; results for this are not yet available. The second fabrication, of the other standard sensors and also the hi-res sensors, is stalled until the cause of the current failures is understood. This means we will not get the second batch until Jan at the earliest. Jamie has tried to measure the resistance of the test structure resistor. The main issue is that the 4Mohm resistance is so large that leakage currents through the input pad protection diodes mask the direct measurement. He has therefore measured full IV curves for the resistor and found the slope to give the resistance itself. For the one sensor measured so far, this gives 4.1Mohm for the old layout and 3.6Mohm for the new, although he thinks he needs to measure the test structures on several sensors to be sure of the results. Jamie will check how many masks would need to be changed to correct the row address bug. He thinks it would be one or two, which may then be feasible to correct. However, the CALICE budget is completely spent so the money would have to be found elsewhere if we wanted to correct the masks. Jamie also thinks it would confuse the issue with the foundry to say we have a design fault before the cause of the other problems with sensor 1.1 are found. The new and old shaper design in the test structures both show a diagonal slew towards a corner pixel in the 3x3 test structure array which is not understood. This looks similar to a lesser effect seen in the sensor 1.0 laser scans. Jamie will need more PCBs for mounting sensors at RAL soon. Marcel would also like another USB_DAQ board if there is one available. Paul will organise getting these to RAL. Conferences: There are three conference presentations (Nigel at CLIC08, Rebecca at IEEE and Marcel at IPRD08) which should be added to the conference web page. Anne-Marie will give the LCWS MAPS talk next week and needs to give a practice before then. This was set for 11am on Thu 13 Nov by phone. Jamie and Marcel will be writing proceedings for IEEE and IPRD08, respectively, which both need to avoid using diagrams and figures from the draft paper as well as from each others' proceedings. The IEEE deadline is the end of the week so Jamie will circulate a draft soon. [Note added after the meeting: Renato has been invited to give a talk at a workshop at Delft, see poster on the usual web page.] Sensor 1.0: Jamie also showed some results on the sensor 1.0 fine laser scan. He clearly sees an asymmetry for the four diodes. The effect looks the same for the two active test pixels on the sensor. The data for these two were taken in the same scan so no time-dependence of the laser intensity is likely to be the cause. It also cannot be easily explained by the laser focus due to the small range of motion. This may be similar to an effect seen by Paul in the Imperial laser scans, which was assumed to be due to imperfections on the sensor surface. Marcel showed some slides on the results of a long Fe55 run; see the usual web page. The pixels with a width set to zero are those for which the fit failed to converge sensibly. He sees a falloff of the gain for pixels close to the memory boundary between regions which is not easy to explain if the Fe55 peak is really due to photon conversions within the diodes. Hence, this result is a puzzle and needs further study. The afterglow is consistent with a saturation effect in the threshold comparator, which Jamie thought could be an issue some time ago. He estimated it would become significant at thresholds of around ~1000TU, rather than ~300TU, so the scale is not understood. Jamie thinks the effect is reduced by going to a higher threshold common mode so he suggested doing scans with a common mode of ~3000 rather than the usual 2048. [Note added after the meeting: Marcel sent round some more plots of the means and widths as a function of the x column position; see usual web page.] Beam test: We are still hoping to put sensor 1.1 in beam in early 2009. However, the CALICE travel budget is very tight and so we should try to confirm we really need the funding asap to secure it. If the current sensors are found to work well, then we should be able to go in Jan or Feb. If the current sensors are found to have a processing error and we have to wait for sensors from the second fabrication round, then the eariest we could feasibly go is Mar. If we get SPIDER funding, then Apr (i.e. the new FY) would be possible but if the DECAL part of SPIDER is not approved, then we would probably have to be sure of doing the beam test in the current FY. Mike thought we should investigate rolling the travel funds to the next FY but Paul doubted this would be allowed. Next meeting: This will be at 10am on Mon 8 Dec.