---+ Cryo board rev3 issues

  • Memory issues:
    • Pull down DDR CKE pin if we want to use a preloading strategy
    • Move CKE from a 47-ohm terminating resistor to a 4.7k one (100 mW!)
    • Add terminating resistors for SDRAM DQS, DM pins
    • Pointing hand Do not place (DNP) address terminating resistors (RT5, RT7). This closely matches the DfMUX schematic; it works; and it saves a lot of power.
  • Power sequencing and integrity issues:
    • The heater switches should be pulled LOW to avoid on-during-float behaviour
    • WDI pin on MAX6734 should have a pull-up pin (otherwise nothing drives it while the FPGA is being programmed, and we depend on it to wander predictably.)
    • VAuxP isn't connected to the backplane (should it be jumpered there?)
  • Front panel issues:
    • Evaluate RJ45 placement (they aren't even; does it matter?)
    • Evaluate SD card placement (will it be accessible with a front panel?)
  • Layout issues:
    • The serial number jumpers (J12-J19) shouldn't have traces on the top layer. It's too easy to cut the traces when snipping the jumpers.
  • Component loading issues:
    • Pointing hand J29, J31 should be marked DNP (these are redundant programming headers for the SPI flash. We don't need them.)
    • Pointing hand MAC and FPGA DONE LEDs (D15-D21) were loaded backwards.
    • Pointing hand Load R100 with a jumper; don't load R102. This enables one of the power-supply regulators by default.
  • WinBond now has 16MB (128Mbit) SPI flash; it may be desirable to use it instead of the W25Q64s for bitstream contents.
    • Why? There's room for <2 Spartan-6 LX150 bitstreams on the current part; we hope to cram 2 on there, plus bootloaders, via compression.
    • The bitstream compression is crude, and it would be terrible to run out of space as the bitstream grows.
    • They are pin compatible (but Insight doesn't support them yet?) It should be simple to replace one of these by hand to evaluate.

  • Filter changes
    • Measurement Channel - Sense Channel R18 & R19 changed from 100k to 1k C43&C44 changed from 6.8nF to 220nF (X7R)
    • Measurement Channel - Drive Channel C13 & C14 changed from 6.8nF to 2.2nF, R5 & R10 changed from 100k to 1k, R8 changed from 100k to 2.49k, C20 changed from 6.8nF to 220nF X7R
    • Heater Channels - Heater Driver R12 changed from 100k to 0 Ohms, R15 remains 100k, changed R54 from 100k to 2.49k, changed C60 from 6.8nF to 220nF X7R, changed C28 from 6.8nF to 470pF


This topic: CryoElectronics > DigitalFMux > CryoElectronicsDesignFiles > CryoElecRev3_Issues Topic revision: r17 - 2010-12-16 - GraemeSmecher
© 2020 Winterland Cosmology Lab, McGill University, Montréal, Québec, Canada