FPGA SPECTROMETER COMMISSIONING

1
FPGA SPECTROMETER COMMISSIONING

• Spectrometer description
• How it will be used
• Where we are at.
• What needs to be done
• More infohttp//www.naic.edu/phil/hardware/pdev

2
Description.. what we bought

• 20 spectrometer boxes purchased from jeff mock
  (jeff who??)
• 1 box can process 1 beam of alfa.
• 14 boxes allows for 2 simultaneous alfa
  experiments.
• 2 boxes for aeronomy, planetary radar
• 16 fileservers (each with 4 Tbytes of disc).
• IF mixer chassis IF to base band (I,Q) (dana)

3
Description.. 1 box

• 2 paths each handling 170 MHz Bandwidth
• A/D.. 12 bit 170 Mhz I Q samplers
• Digital mixer.
• Digital low pass filter (decimation 2..1024)
• Optional time domain data 4,8,16 bits
• Polyphase filter bank..16 to 8192 channels.
• I, polA/B, full stokes power computation.
• Sum spectra. Output 32,16,8 bit spectral data.
• Box output I/O rate 80 Mbytes/sec.

4
How it will be used

• Alfa covers 1225-1525 MHz x 7 beams
• Galfacts galaxy full stokes
• 65Mb/sec 7 450 Mb/sec.
• Palfa search Upgrade 100 to 300 MHz BW.
• 32Mb/sec7 225 Mb/sec.
• Aeronomy IF sampling, time domain data (Tamara)
• 32,20,10,.. bandwidths. 32 MHz Bw gets plasma
  line in 1 band
• Sampling synced to transmitter pulses and echo
  return.
• Planetary Radar IF sampling, time domain data.
  (Tamara)
• 40,20,10,.. Bandwidths with 2,4,8,16 bit
  samples.
• Fir filters for CW, sync filters for ranging.
• Digital doppler correction by interpolating,
  dropping samples.
• Single Pixel observations
• Use 7300 MHz bandwidth to cover up tp 2 GHz bw.

5
Where we are at.

• Alfa
• Hardware
• 14 spectrometer boxes with 7 mixer chassis
  installed.
• 14 file servers installed, 7 in use.
• Software
• Acquires data, writes to disc in pdev or fits
  format.
• Aeron
• Test data taken in 32Mhz bw Mode
• Planetary radar
• 20 MHz bw data taken on saturns rings.

6
What needs to be done.

• Hardware/FPGA programming
• Resolve radar rfi saturation in PFB butterfly
  stages. If this is not solved, the lower 170 MHz
  will be unusable.
• Take a closer look at birdies generated
  internally to the boxes.
• Debug sync filters for planetary radar (Tamara)
• Debug digital doppler (Tamara).
• Single pixel.. Need to widen the downstairs IF/LO
  to handle 2 GHz bw.

7
What needs to be done (software)

• Integrate running of spectrometer with GUI.
• This will take the most time.
• Fill in all fits header keyword data.
• Speed up startup configuration of device.
• Test winking cal, blanking.
• Automate optimum level setting and configuration
  for all possible configs.
• Pulsar users want psrfits format. This is
  different from the fits files we are generating.
• Analysis software. Continue developing IDL
  routines to verify the device operation.
• Work on system/program documentation (see web
  site).

8
When .

• Shooting for may 2008 for galfacts alfa
  observing. This will also enable any of the other
  alfa spectral line surveys to use the device.
• Pulsar fits file format will come after galfacts
  (since palfa is already taking data).
• Planetary radar/aeronomy needs FPGA programming
  and Tamara Hall is leaving.
• Till she leaves, radar/aeronomy work gets
  priority.
• 2 GHz single pixel still in the design stages.