Minutes of the FT upgrade meeting held at RAL on 28/1/99 ======================================================== Present ------- Don Clarke, Tim Greenshaw, Rob Henderson, Derrick Hill, Graham Houghton, Gary Markey, John Morris, Girish Patel, Dave Sankey, Mike Smith, Terry Sloan. 1) Actions arising from previous meeting ======================================== Graham H -------- Graham talked about the visit to Daresbury made by people from SpemCo. He explained that orders have been placed for enough circular PCBs to equip 6 planar chambers. The delivery date for these is the end of March, two weeks later than originally foreseen due to the time required to obtain the materials. SpemCo intend to make the vias (the plated through holes connecting the electrodes on the two surfaces of the PCB) as follows; the PCB is etched; holes are drilled through at the positions of the vias; electroplating is used to deposit copper in the holes, but also on the electrodes. The problem with this procedure is that it increases the thickness of the electrodes from the initial 15 to 18 microns by whatever amount is necessary to form the vias; a thickness of 7 microns was suggested. This additional thickness presents more material to particles traversing the chamber. Graham was asked to enquire if the 7 microns can be significantly reduced without resulting in unreliable electrical contact between the two surfaces. If this is not the case, we will consider returning to connecting through the hole using wires soldered onto both surfaces. SpemCo suggested that they machine the PCBs after receipt of one of the Nomex panels, to enable them to check all dimensions etc. The procedure Spemco will adopt to minimise delays caused by the slow arrival of raw materials is to develop their production techniques using any available material, so that they are up to speed when the genuine stuff arrives. From that point production should take only two weeks. They outlined three possible checking procedures, visual inspection, camera and automatic optical inspection. We will make the decision as to which, if any, of these are necessary after having seen the PCBs produced during the above development phase. Action Graham H to contact SpemCo about necessary thickness of copper ------ in vias. Tim --- Tim said that Manfred had confirmed that 6cm was the largest necessary forward movement of the Forward Tracker following the upgrade. We decided to allow for a possible forward movement of 7cm to allow a small safety margin for the forward trigger. Questions were asked about the spring constant of the spring within the front right FT support foot. It was remarked that incorrect choice of this spring constant could result in a twist of the tracker as is observed by Rob in the data he has recently analysed. Tim also stated that the tolerances to which the Noryl was machined for the Radials were 50 microns...but couldn't remember whether this was ±50 or ±25 microns. Action Rob, Don and DESY engineers to study whether known slopes of ------ rails could possible account for Rob's observations. If answer to above is yes, Graham H to look at feasibility of modelling FT tank and support to confirm that observed rotations are of a size that might be explained by mechanical twist of FT resulting from support positions. Tim to inform Graham of Radial tolerances as soon as possible. Derrick ------- Derrick told us that Automatic Precision Turning Leicester Ltd. (APT) can produce our pins at a cost of £1.69 per pin, if we provide the materials. Derrick and Graham also described their visit to Medelec. Their impression of the firm was favourable. Medelec would produce the pins using three tubes drawn to different diameters which would then be ``swaged'' together. No-one understood the details of this procedure, but Medelec assure us that loads of 50g (0.5N), the maximum we are contemplating, are well within the limits which the pins they produce will tolerate without any slipping of the different tubes. We can test this with the 20 pins that have been promised in the next week or two. This solution obviates the need for any brass tubing. This was only used previously to simplify machining. Medelec also have a nice solution for the connectors to fit over the pins. They propose using a box section tube, with walls thin enough to be flexible, with dimensions which ensure a tight push fit. Derrick's samples appeared to work well. Don has not heard any more details from the German pin manufacturers as yet, but did have information on the DESY pin cleaning procedures. We have to examine carefully any pins we receive to check that the necessary standards of cleanliness have been attained! Don also had information on a Swiss supplier of hardened copper tube suitable for pin manufacture. We decided to try and use Medelec as our main supplier of pins and the connectors to go on those pins, but to pursue the APT option as a back-up solution. The Medelec pins will not have the stepped outer diameter of the pins used for the current planars. Don explained that one purpose of this stepped outer diameter is to ensure that there is a long path along the Noryl surface between the sense and grid wire pins which ensures there is no surface conduction path between the sense and grid pins. Using the constant outer diameter may make it advisable to drill a larger diameter hole around the sense wire position than currently proposed. Current Proposal for discussion | | | cell interior | cell interior |-- |-------- --------- ------- <--> 4mm depth <--------> 10 mm depth --------- ------- |-- |-------- | | Noryl | Noryl | Action Derrick to study Medelec pins. ------ Don and Graham H to organise tubing for APT so they can produce sample pins for study. Don and Graham S to finalise Noryl design w.r.t. above point. Girish ------ Girish reported that he has converted all Mike's 6 wire FTsim planar mod.s into 8 wire FTsim. He has also duplicated all the ``noise'' hit generation mechanisms, the unpacking, the calibration banks etc. so that he is almost in a position to produce level 4 monitoring for the 8 wire planars. Both the new FTsim and Ftrec are running at RAL. Action Graham S/H to provide Girish with drawings which allow positions of ------ detectors in FTsim to be checked. Girish to check positions. Graham to provide above pictures in electronic form for web. Mike to resurrect graphics and check genuinely have 8 wires everywhere we should have 8 wires etc. Girish to remove foremost Radial in FTsim. Girish and Rob to get above running at Lancaster so Rob can start segment finding work. L4 hists to John for study as soon as available. 2) Progress at Daresbury ======================== Graham H summarised the situation at Daresbury. The work that is essentially completed is the following: Circular PCBs ordered (see above). Nomex panel design completed. Design to Hexcell composites in Belgium. Order imminent, but question as to how many panels should be ordered. Three quotes for machining of Noryl blocks varying from £6000 to £19000, price partly dependent on tolerances (see above actions). Crimp pin design/order proceeding (see above). Noryl block design completed, but see above comments concerning design with Medelec's proposed pins. Design outstanding for the following (on the whole relatively simple) items: PCB side walls. PCB grid connector. FT feet. Dowels for Noryls blocks and for PCB/Nomex location. Connector for sense/grid pins (but see above comments on Medelec box section tube). Spacer collar for end plate (~2mm thick ring to ensure good fit of forward endplate to foremost U channel after removal of foremost Radial). Gas manifold. FT/CT interconnection rods. Blanking plates for feedthroughs. Graham asked that we think about the above and tell him of any omissions. Graham also reported that Gary had made good progress in the preparation of the clean room facilities. Air conditioning installed. Electrical work completed. New flooring in place. Painting done. Two surface tables in place. Wiring table ready. The following jigs and fixtures are needed: Wiring jig. Tensioning jig. Nylon blocks for supporting chamber side walls while chambers loaded and resin curing. (Note that load is large; the horizontal chamber is covered with 5cm by 5cm mild steel rods!) Pillar drill/press to simplify insertion of pins into Noryl (particularly if Medelec design used). There was some discussion of the wiring procedure. Following warnings from Don about the difficulty of locating small wire trimmings, Gary suggested leaving relatively long ``tails'' on the wires and attaching these to sellotape before trimming. Don stressed that the forward tracker should be electrically isolated from the rest of the H1 detector (i.e. feet must be insulating as must be connections between forward and central trackers). Action Tim to organise transport to Daresbury of equipment that might ------ be useful for tests. Tim and Graham to clarify budget situation (following large increase in price of chamber side walls and possible increase in price of Noryl machining) before order of Nomex panels. Number to be ordered dependent of money available. Don to send Graham sample gas connectors. All to check all outstanding mechanical work covered in above list. 3) Updated schedules ==================== Graham H showed the latest schedules, but these considered March 2000 to be the start date of the HERA upgrade, not May as is actually the case. Action Graham to provide schedules for start of upgrade in May 2000. ------ 4) Progress with trigger ======================== Don presented the current plans for the forward trigger. He described calculations of the expected minimum ionising particle efficiency this will achieve, made using optiCAD. A signal to noise ratio of about 10 to 1 looks achievable using 1cm scintillator, wave length shifter bars and HPD readout (gain 1000) at the rear of the forward tracker (FTi1). At the front of the forward tracker (FTi2), 2cm scinillator and photomultiplier readout will be used as in the FIT. The signal to noise ratio for a MIP should be better than 10 to 1; we have several such set-ups working within the H1 environment. Don estimates the total cost of the forward trigger to be about DM300000, (i.e. about £100000). 5) Finances =========== John described a new source of funding available within CLRC. This does not appear to be relevant for us unless we come up with a novel approach to (say) drift time determination using lasers, sources or something that is not an upgrade, but genuinely new, and has wider applicability than just the H1 forward tracker. 6) Late news ============ John was able to confirm that we are able to continue placing orders despite the general stop that has been applied within RAL. Don has now got the cards for the FADCs. 7) Next meetings ================ Thursday, 4th March, Daresbury. Tuesday, 6th April, Liverpool. Tuesday, 11th May, DESY.