|
|
| Line 28: |
Line 28: |
| | - Bill explained us his ideas about the wire stringing. He will use the strongback that is at IUCF. Lubomir will take care of bringing the strongback to JLab. The idea is to use tape pads to hold the wires at the right positions before gluing and soldering them. The wire positions will adjusted by hands (or with some tool) using the camera. We discuused once more different options for the tension measurements. | | - Bill explained us his ideas about the wire stringing. He will use the strongback that is at IUCF. Lubomir will take care of bringing the strongback to JLab. The idea is to use tape pads to hold the wires at the right positions before gluing and soldering them. The wire positions will adjusted by hands (or with some tool) using the camera. We discuused once more different options for the tension measurements. |
| | | | |
| − | - the templates for the | + | - the templates for the rohacell arcs will be ready next Wednesday. Two students will start cuting the arcs, supervised by Mark. Bill said the most important is to have the I.D. cut correctly. The arcs will be stored in a box and then they will start gluing the rings. |
| | + | |
| | + | == Electronics == |
| | + | |
| | + | - Fernando: CEM was awarded to produce the PCBs. After getting the first article (two of each) on 7/19, the PCBs will be tested by Fernando within two weeks before starting the mass production. In parallel, the first boards will be given to an ouside company for stuffing, they will be tested again and then we will contine with the stuffing of all the boards. |
| | + | |
| | + | - Lubomir looked with the differential probe at the output (with 18m cable) of the new anode and cathode cards. The dynamic range of the anode cards is about 2/3 of the range of the cathode cards; according to Fernando this is normal and comes from the difference in the gains. This explains the cutoff in the fADC spectra discussed before. |
| | | | |
| − | - Bill started working on the wire stringing fixtures. We agree that this work has highest priority right now. Bill discussed the way the wires will be pushed down locally, close to the soldering pads. His idea is to use a diffraction plate with grooves: 50 grooves/mm with about 10 microns depth. The grooves will keep the wires at the same spacing while pushing them down.
| |
| | | | |
| | == Cathode redesign == | | == Cathode redesign == |
| | | | |
| − | - Roger prepared the drawings according to the modifications from the last week: increased O.D. with flaps to improve the grounding. Except for few minor modifications (extra flaps in the areas without cards) the design is ready to go, hopefully next week. Roger will send the design to the vendor. We discussed what our requirements will be. Bill suggested to ask for prices: 1) for 60 cathode foils and also 2) if we use the whole material for about 120 foils. We will have 4 foils in the first article and then 4-6 weeks time to check them and start the rest. We discussed how to transport and store the foils. Bill suggested to have them stored in rolls. Mark had suggestions about the material to separate the foils in the rolls. | + | - Roger is ready with the design, except for a few cosmetic changes. He will send the gerber files to the company by tomorrow and expects estimations from the company for 70 cathodes by the middle of the next week. At the same time he will start the PR paperwork, so that he can submit it as soon as he gets the quotation from the company. |
| | | | |
| | == Full-scale prototype test results == | | == Full-scale prototype test results == |
| | | | |
| − | - Lubomir showed a [http://www.jlab.org/Hall-D/detector/fdc/tests_full_scale_prototype/DTvsMeanMaxPlusGarfieldOxygen0.1pc.pdf plot]: the mean signal vs drift time. One can see that the signal drops with the drift time. Normally this shouldn't happen: the electrons drift and, close to the sense wire, they are multiplied, so that the gain doesn't depend on the drift time. A possible explanation: the electrons while drifting can be attached to electronegative gas molecules. To check this, Garfield simulations were done adding small percentages of oxygen contamination in the gas mixture. The red dots in the plot represent the Garfield results with 0.1% oxygen contamination. The good agreement suggest that probably we have some oxygen in the chamber. Without oxygen, the Garfield simulations show that the signals do not change with the drift time, except for very small drift times. | + | - Lubomir showed several Beni's plot, linked above. |
| − | | + | |
| − | - We want to measure the oxygen contamination in the full scale prototype to make sure we understand the problem. Then, the plan is to measure the oxygen from the new design mechanical prototype, to be sure there's no oxygen there. If the problem really is in the oxygen contamination, then without it, we can expect to reach the plateau at about 75-100V ( a factor of ~2 in the gain) lower voltage. This will make also the dynamic range of the signals narrower. To test this we can use the new oxygen sensor, but before that we will try with the old one. We have to pump the gas because of the high impedance of the sensors, but the problem is how to regulate the flow through the sensor so that it doesn't disturb the flow through the chamber. We will go with the Beni's scheme, linked above, in which instead of the rotameter in front of the sensor we will use the current mass flow controller. The latter being used now in the gas system will be replaced by our new mass flow controller. The plan is to install the new gas equipment in a separate rack so that it can go altogether eventually to the shed in the hall. Brian Cross gave us a small pump that can be used in the above scheme.
| + | |
| − | | + | |
| − | - Beni prepared several plots, linked above, showing the drift time resolution. Two layers were used with the wires parallel. In one case the drift time in the two layers were compared but for tracks within a plane perpendicular to the layers and parallel to the wires. In the second case Beni included one of the layers in the track reconstruction and look at the predicted hit in the other layer. In both cases he got a resolution of about 200 microns which is encouraging.
| + | |
| | | | |
| | --> | | --> |
