23-Jul-03 Hello, Last week there was a series of messages about, "ADF production planning". If Saclay's participation in the Run IIB L1 Cal Trig must end with the proto-type ADF card then I believe that we need to look at a full list of the work that needs to be accomplished to complete the ADF part of the Run IIB Cal Trig. The discussion last week focused only on ADF circuit board production. Should we start a list of all the ADF jobs that still need to be covered after Saclay's participation ends. The items that I can think of are listed below. Please correct the mistakes that I have made or add to the list. 1. ADF Crates and Power Supplies Saclay was providing the crates and power supplies for the ADF system. My understanding is that this is commercial equipment but that still leaves the responsibility for purchasing, testing, installing, and repair when necessary. 2. BLS to ADF Cables Saclay was providing the special cables to connect the current BLS Cables to the high density ADF cards. These are special cables made from 80 Ohm coax with different connectors at each end. There is at least the responsibility and work of manufacturing and testing and labeling and planning the routing of all of these cables. Once installed it will be hard to get at these cables. This is a non trivial amount of work. 3. ADF Design CAD Support Do any of the engineering groups here regularly use the same CAD tools as Saclay used for the design of the proto-type ADF card ? I hope so. If not we are stuck making a hard choice: Do the production, testing, and long term support of the ADF card without CAD tools that let you look at or modify the schematic or pcb designs. Have some one here purchase and ramp up on the CAD tools used by the Saclay group. Move the design to a CAD system that some engineering group here already uses. This could imply making the production run from a separate design entry than was used for the proto-type. 4. ADF Circuit Board Production Purchase parts including those needed for the long term support of the ADF card. Select an appropriate assembly vendor and provide the supervision and support that all assembly runs need. 5. ADF Testing Certainly we will need to test the ADF cards before they are installed in the actual Run IIB Cal Trig system. This testing will be complicated because the ADF is a high density analog and digital card. Many of the problems that could happen on an ADF card are much much easier to find in a specialized test than in the running production L1 Cal Trig system. Examples of the tests that one would like to make on these cards are: Digital Testing Verify the proper operation of: the VME Interface, of Configuring the FPGA's, of all inputs to the FPGA's (e.g. from the ADC's), verify all the outputs from the FPGA's (e.g. the serial digital links). Analog Testing to verify the proper operation of a high speed multi channel ADC system is not simple task. Things that you would like to check on individual ADF cards before installing them in the actual L1 Cal Trig system include: DC noise and stability, range of the Zero Energy Response DAC, DC linearity and gain, distortion check via FFT testing at 1/3 or so of the full sampling rate, gain at a couple of high frequency points to check the anti-alias filter, and channel to channel cross talk. Without explicitly making these tests many of the analog problems that could happen on an ADF card may not show up until the card has spent a lot of time in the actual L1 Cal Trig system. Setting up all of these tests and writing the software to exercise the ADF cards in these tests will be a lot of work. Hopefully the production run will go well and there will not be a lot of circuit board assembly problems discovered during the ADF card testing. To repair any ADF cards that fail during Run IIB, some of this test setup should be maintained for the duration of the Run. 6. ADF Production Card Re-Work The repair of production problems should be done at the same location that houses the test setup for the ADF circuit boards. This repair work will require the normal electronics debugging skills and surface mount re-work skills and tools. 7. ADF Firmware Although I hope that this firmware will be carefully planned and documented ahead of time the ADF firmware may need to evolve quite a bit as the system is used. There almost certainly will need to be special firmware for testing of the individual ADF cards. Likely separate special firmware will be needed for digital and analog testing. Because the strategy for filtering BLS trigger pickoff signals is so different at 132 nsec running vs 396 nsec running there may need to be different versions of filter firmware to optimally support the two running modes. The differences between 132 ns and 396 ns are big enough that if you knew ahead of time that the Tevatron was always going to run at 396 ns you might even build the ADF cards with a different FPGA part. Provide long term support for ADF Firmware. Maintain a version of the FPGA vendor's design tools to work with the FPGA parts that are used on the ADF card. 8. Constants Data Base Assuming that different channels will have the same firmware but different filter constants to compensate for BLS signal gain differences and BLS signal timing differences and BLS signal waveform differences - there will be a non-trivial amount of work and responsibility for collecting and maintaining all of this data. Work on a strategy to handle all of this could start now. 9. Other Components in the ADF System Are there other components in the ADF System (e.g. timing and control signal distribution cards) that need someone needs to take responsibility for ? 10. Raw BLS Signal Access Access to the raw BLS signals is a very important tool in debugging the problems that happen in the Calorimeter system. Looking at the BLS signals with a scope is the only way that one can see real electrical signals from the Calorimeter system. Looking at raw BLS signals is used to see: problems in the calorimeter itself, problems in the Cal pre-amps, problems in the precision readout part of the BLS card and problems in the trigger signal pickoff part of the BLS card. Because of the low density BLS cabling in the current Run I Cal Trig right now it is easy to access the raw BLS signals. We routinely look at raw BLS signals on a scope (even sometimes during Physics beam running). Access to these signals in the Run IIB system may/will be a lot more difficult. It is hard to imagine maintaining the Calorimeter system without being able to see the raw BLS signals. One solution is a passive card that you could plug into a slot instead of the ADF card. Such a "BLS signal access card" is probably a simple thing but it is just one more job to get done.