From GlueXWiki
Jump to: navigation, search

September 30, 2010 FDC meeting

Tentative Agenda

  1. Electronics
    • Update (Fernando)
    • Discussions of the FDC electronics (cont.) [FDC requirements]
  2. Engineering (Bill and David)
    • Wire stringing: fixtures, movement system [[1]], [[2]]
    • Discussions of the stringing procedure (cont.)
    • Other
  3. Full-scale prototype
  4. Other


Participants: Eugene, Fernando, Bill, David, Chris, Simon, Beni, Glenn, Gerard (on the phone), Roger, Mark, Casey, and Lubomir.


- Fernando: the last PCBs were shipped and will have them by Monday and then all will be sent for assembly (the contract is in place already). Chris checked all the boards visually, and three of them for continuity: only minor cosmetic problems. Gerard implemented the auto-load feature on fADC125. Bryan Moffit will put the new firmware on the module so that we can use it for the next tests. Gerard fixed the third fADC125 and soon will have it ready to be sent to CMU. No news concerning the cathode foils.

- Lubomir updated the document about the FDC electronics requirements, linked above, where you can find more details of the discussions listed below.

- Pre-amps in discriminator mode: GlueX-doc-1364 states: "the discriminator has been optimize for operation between 2-5fC and has a range of 20fC". While testing the discriminator cards with the full-scale prototype the lowest possible threshold was 10fC. With more chambers connected and more noise we may need to increase the threshold up to 40-60fC. Fernando: the above numbers correspond to the high gain mode, in case of a low gain these numbers have to be multiply by ~4.

- Quadratic vs linear interpolation in fADC125: if using fADC125 for drift time measurements, the quadratic interpolation gives a little advantage for 40ns peaking time. If comparing the timing from the strips and wires (this helps to separate the real from the fake hits) the fADC resolution is of the order of 1ns. In this case the quadratic interpolation gives ~25% better resolution than the linear.

- fADC vs discriminator for the wires: Eugene asked if we have decided. According to the simulations the discriminator gives much better position resolution for distances smaller than 3-2mm where good timing resolution is needed. For distances bigger than 4mm fADC gives ~20% better resolution. Gerard: realistically, one has to use also 1ns timing resolution for the discriminator; Lubomir will implement this in the Garfield simulations. Gerard wanted to have the raw and convoluted signals from the Garfield simulations. Lubomir will send him the program to generate the signals. Nevertheless, the biggest advantage of the discriminator card is that it is close to the detector and one can lower the threshold, which will improve the resolution. At the same time, one can use the cathode information from the FADC to reconstruct the drift time in the regions where it gives better precision.

- FDC gas mixture: Eugene asked about the status. As far as the electronics is concerned one should assume 8x10^4 gas gain that we got with 40/60 Ar/CO2 and 2225V/-500V voltage. Both wire and strip resolutions improve with increasing signal to noise ratio and the signals with 40/60 were sufficiently high to get the design resolution. The 4-5mm (distance to the wire) region is an exception, where long tails in the drift time distribution significantly deteriorate the resolution. This is because of the strong drift time vs electric field dependence. The problem can be solved by using 90/10 mixture but then we can't reach the same gas gain because of the limited quenching properties of CO2.


- Bill and a person from the machine shop will visit "Vision machining" to set up the wire production.

- Wire stringing table: The holes in the granite table are ready. Bill glued pins into the holes. David is working on the movement system. Two options: buying a $2k controller that works with new computers/software or using the existing JLab controllers with the existing software. First will try the second option; Beni will help with the controller and the documentation.

- Preparation for wire stringing: the holes in the granite tables are being drilled now and all the initial parts to be attached to the table are ready. David has been working on the movement system and the encoder. Two options to connect the stepper motor to a new computer: one would require a $2K adapter that will work with new computers and software, the other is to use existing JLab controller with existing software. We will try first the second option; Beni will help with the controller and documentation.

- Bill made a 3D model of the wire stringing system, linked above, and explained the procedure. Wire by wire first sense then field wires: stringing, pressing with screws, taping down to the outer side of the board, measuring/adjusting the tension. After stringing all the wires: applying epoxy, measuring/adjusting the position, soldering, and cleaning. The improvement proposed by Bill is to tape the individual wires to make sure they touch the board (something we discussed last time).

Full-scale prototype

- Beni continued testing different gas mixtures (90/10, 85/15 and now 80/20) and showed some plots above. Pictures from the scope show after-pulses that are typical for these gas mixtures with high Ar content at high enough HV. Because of the after-pulses the ADC saturates. Fernando: where's the plateau? The end of the plateau is where the after-pulses start. The beginning has to be find by lower the HV, Beni covered so far ~150V for some of the mixtures. To reduce the after-pulses Beni wants to use the pre-mixed bottle Ar/CH4 95/5% and to mix it with CO2. Methane is flammable above 5% in air. Using methane in the hall has to be discussed, but first Lubomir will simulate it with Garfield to estimate the advantages. It is know that organic molecule cause polymerization on the wires at high rates and gains. The other halls (A and C) have been using ethane as quencher (mxied with alcohol to prevent polymerization) for a long time.