for GAVRT at Caltech Glenn Jones Aug. 03, 2008 2008 CASPER - - PowerPoint PPT Presentation

▶
for gavrt at caltech
SMART_READER_LITE
LIVE PREVIEW

for GAVRT at Caltech Glenn Jones Aug. 03, 2008 2008 CASPER - - PowerPoint PPT Presentation

Oct 14, 2022 28 likes •463 views

CASPER Development for GAVRT at Caltech Glenn Jones Aug. 03, 2008 2008 CASPER Workshop Acknowledgements Xilinx Generous FPGA and software donations Sandy Weinreb & Hamdi Mani Feed measurement data Useful stuff first! The

slide-1
SLIDE 1

CASPER Development for GAVRT at Caltech

Glenn Jones

  • Aug. 03, 2008

2008 CASPER Workshop

slide-2
SLIDE 2

Acknowledgements

 Xilinx – Generous FPGA and software

donations

 Sandy Weinreb & Hamdi Mani – Feed

measurement data

slide-3
SLIDE 3

Useful stuff first! The Simulink scope is terrible!

slide-4
SLIDE 4

gtkWave from Simulink for CASPER

slide-5
SLIDE 5

gtkWave from Simulink for CASPER

slide-6
SLIDE 6

Sick of ‘od -x’ for interpreting snapshot data?

slide-7
SLIDE 7

gtkWave Snap – View SnapBRAM data

slide-8
SLIDE 8

gtkWave Snap

gtkgen(wave) $ gtkwave temp.vcd

slide-9
SLIDE 9

Vector accumulator

slide-10
SLIDE 10

Vector Accumulator

slide-11
SLIDE 11

Utility blocks

slide-12
SLIDE 12
slide-13
SLIDE 13

Goldstone Apple Valley Radio Telescope

 34 m telescope in

southern California to be used by K-12 students to take data for astronomers.

 Currently being

equipped with a novel ultra-wide-band radiometer designed at Caltech.

slide-14
SLIDE 14

Wide band quad-ridge feeds

 0.5-2 GHz  Uncooled feed  LNAs cooled to 50K  4-14 GHz  Feed and LNAs cooled

to 15 K

slide-15
SLIDE 15

Receiver Noise Temperature

Noise and Gain of Polarization X GAVRT HFF Front-End, DSS13 Pad, July 17, 2008

10 20 30 40 50 60 70 80 90 100 2 4 6 8 10 12 14 16 18 20 GHz Noise, K

  • 24
  • 21
  • 18
  • 15
  • 12
  • 9
  • 6
  • 3

3 6 Gain, dB

Noise Gain

LFF X Pol Trcv and Gain 20 40 60 80 100 120 140 160 180 200 0.5 1 1.5 2 2.5 3 3.5 4 Frequency, GHz Trcvr, K

  • 20.0
  • 17.0
  • 14.0
  • 11.0
  • 8.0
  • 5.0
  • 2.0

1.0 4.0 7.0 10.0 Gain, dB

Noise Gain

slide-16
SLIDE 16

Crossover performance

Nois e of HFF and L FF, X and Y Pol, from 0.5 to 5 GHz

20 40 60 80 100 120 140 160 180 200 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 Frequency, GHZ Nois e, K

L FF Pol Y HFF Pol Y L FF Pol X HFF Pol X

slide-17
SLIDE 17

Front end box

 4 dual polarization

pairs of receivers

 Input from 0.5 to 18

GHz

 Select bandwidth from

100 MHz / 500 MHz / 1 GHz / 2 GHz

 Downconvert to

baseband.

slide-18
SLIDE 18

Basic Receiver Architecture

2 GHz BPF @ 22 GHz 1 GHz LPF 1 GHz LPF

22-40 GHz Tunable LO 22 GHz Fixed LO I I Q

The system consists of eight such receivers, arranged as four dual-polarization pairs.

LNA LNA

2-14 GHz Feed 0.5-4 GHz Feed

slide-19
SLIDE 19

Receiver modes

 The I and Q outputs can optionally be combined in a

hybrid to form upper and lower sidebands. Additional filter

  • ptions are also available.

 Currently only 8 of the 16 possible outputs are routed to

the digital back-end. This will be upgraded in the future.

IF Filter Bandwidth I/Q or U/L Selection Baseband Bandwidth Processed Bandwidth Image Rejection Comments 2000 I/Q 10 -1000 2000 25 dB Needs 1 GHz I/Q spectrometer 2000 U/L 100-1000 1800 15 dB Needs 1 GHz spectrometer 2000 U/L 500-1000 1000 20 dB For 500 MHz spectrometer 400 U/L 120-520 400 50 dB For WVSR 400 U/L 270-370 100 50 dB For VSR

DSS 28 Bandwidth Selections (MHz)

Each IF converter provides the following bandwidths with center frequency from 1 to 15 GHz

slide-20
SLIDE 20

The digital backend

 8x ADCs, 8x iBOBs  16x XAUI links to BEE2  2x 10 GbE links to Procurve switch  ~20x 1 GbE to small cluster

slide-21
SLIDE 21

Basic signal processing infrastructure

iADC Input Clk PPS 2Gsps

iBOB

ADC0 ADC1 XAUI0 XAUI1 Clock Out 100 MbE iADC Input Clk PPS 2Gsps

iBOB

ADC0 ADC1 XAUI0 XAUI1 Clock Out 100 MbE XAUI0 XAUI1 XAUI2 XAUI3 100 MbE Clock in XAUI0 XAUI1

8 way splitter Pulse Distribution Amp Pol 1 Real 0-1 GHz Pol 1 Imag 0-1 GHz

16 Gbps 8 Gbps 8 Gbps 8 Gbps 8 Gbps 16 Gbps XAUI0 XAUI1 XAUI2 XAUI3 iADC Input Clk PPS 2Gsps

iBOB

ADC0 ADC1 XAUI0 XAUI1 Clock Out 100 MbE iADC Input Clk PPS 2Gsps

iBOB

ADC0 ADC1 XAUI0 XAUI1 Clock Out 100 MbE

Pol 2 Real 0-1 GHz

16 Gbps 8 Gbps 8 Gbps

Pol 2 Imag 0-1 GHz

8 Gbps 8 Gbps

Corner FPGA Corner FPGA

XAUI0 XAUI1 XAUI2 XAUI3 XAUI0 XAUI1 XAUI2 XAUI3

Corner FPGA Corner FPGA

BEE2

Sampling Clock PPS

2x 10 GbE 24x 1GbE Switch

Computer Cluster

iADC Input Clk PPS 2Gsps

iBOB

ADC0 ADC1 XAUI0 XAUI1 Clock Out 100 MbE iADC Input Clk PPS 2Gsps

iBOB

ADC0 ADC1 XAUI0 XAUI1 Clock Out 100 MbE

Clock Splitter PPS Pulse Distribution Amp

PPS

Pol 3 Real 0-1 GHz Pol 3 Imag 0-1 GHz

16 Gbps 8 Gbps 8 Gbps 8 Gbps 8 Gbps 16 Gbps iADC Input Clk PPS 2Gsps

iBOB

ADC0 ADC1 XAUI0 XAUI1 Clock Out 100 MbE iADC Input Clk PPS 2Gsps

iBOB

ADC0 ADC1 XAUI0 XAUI1 Clock Out 100 MbE

Pol 4 Real 0-1 GHz

16 Gbps 8 Gbps 8 Gbps

Pol 4 Imag 0-1 GHz

8 Gbps 8 Gbps
slide-22
SLIDE 22

Thesis goals

 Build a unique instrument designed to take

advantage of the wide bandwidth provided by the GAVRT telescope to measure the following:

 Detailed spectral characteristics of giant

pulses from the Crab and other pulsars

  • Extensive statistics of giant pulses vs. frequency

 Nanostructure in giant pulses  Dynamic spectra of pulsars with

unprecedented bandwidth

 RFI performance in light of modern

mitigation techniques

  • IQ imbalance correction
slide-23
SLIDE 23

“Earlier, we had noted the potential spectral similarity between giant pulses from pulsars and that of the Sparker. It would be useful to determine the road- band spectrum of giant pulses, say from 1-m to 10-cm wavelength. In short, we are advocating the study of giant pulses from pulsars as convenient plasma laboratories that may further

  • ur understanding of the fleeting

Sparkers.” - Sri Kulkarni “Giant Sparks at Cosmological Distances?”

What we want to look at: Crab Giant pulses vs. Frequency

From: Cordes et al. 2004 ApJ 612 375 Thesis Goal: Detailed spectra of giant pulses

slide-24
SLIDE 24

What we want to look at: Crab Giant pulses vs. Frequency

From: Cordes et al. 2004 ApJ 612 375 Hankins & Eilek, ApJ 670:693-701, Nov 2007

Thesis Goal: Nanostructure in giant pulses

slide-25
SLIDE 25

What we want to look at: Giant pulses vs. Frequency

Hankins & Eilek, “Radio Emission Signatures in the Crab Pulsar.” ApJ 670:693-701, Nov 2007

slide-26
SLIDE 26

Current limitations to giant pulse observations

 Multiple frequency observations have

generally required simultaneous

  • bservation with many telescopes  little

data available

 Ultra-high time resolution has been limited

by:

 Feed/receiver bandwidth  Dispersed pulse is longer than memory

buffer

 Lack of dedispersed trigger

  • More susceptible to RFI
  • SNR of dispersed pulse too low to trigger on
slide-27
SLIDE 27

GAVRT Transient Capture Mode

iADC Input Clk PPS 2Gsps

iBOB

ADC0 ADC1 XAUI0 XAUI1 Clock Out 100 MbE iADC Input Clk PPS 2Gsps

iBOB

ADC0 ADC1 XAUI0 XAUI1 Clock Out 100 MbE XAUI0 XAUI1 XAUI2 XAUI3 100 MbE Clock in XAUI0 XAUI1

4 way splitter 4 way splitter

Ping Pong Ping Pong Align and Merge Small PFB Detector Vector Accumulator

Pol 1 Real 0-1 GHz Pol 1 Imag 0-1 GHz

16 Gbps 8 Gbps 8 Gbps 8 Gbps 8 Gbps 16 Gbps XAUI0 XAUI1 XAUI2 XAUI3 Align and Merge iADC Input Clk PPS 2Gsps

iBOB

ADC0 ADC1 XAUI0 XAUI1 Clock Out 100 MbE iADC Input Clk PPS 2Gsps

iBOB

ADC0 ADC1 XAUI0 XAUI1 Clock Out 100 MbE Ping Pong Ping Pong

Pol 2 Real 0-1 GHz

16 Gbps 8 Gbps 8 Gbps 16 Gbps

Pol 2 Imag 0-1 GHz

8 Gbps 8 Gbps

Corner FPGA Corner FPGA Center FPGA

XAUI0 XAUI1 XAUI2 XAUI3 XAUI0 XAUI1 XAUI2 XAUI3

Corner FPGA Corner FPGA

BEE2

Sampling Clock PPS

Duplicate hardware for other receivers Incoherent Dedispersion and Trigger 2 GB DRAM Circular Buffer 2 GB DRAM Circular Buffer Small PFB Detector Vector Accumulator Incoherent Dedispersion and Trigger 2 GB DRAM Circular Buffer 2 GB DRAM Circular Buffer 2x 10 GbE 24x 1GbE Switch

Computer Cluster

Duplicate hardware for other receivers

RAM buffers are sufficient to store 1 second of raw voltages from 2 chan * 2 pol * 4 Gsps Raw data input rate: 16 Gbyte/s Max data output rate: 2 Gbyte/s

slide-28
SLIDE 28

Frequency domain IQ correction specialized for spectroscopy

q(t) FFT or PFB Filterbank (Real) i(t) FFT or PFB Filterbank (Real) f0 f1 fn

c0 c1 cn

Corrected Spectrum

Cn = j if i(t) and q(t) were in perfect quadrature Looks like it requires n complex multiplies and adds, but FFTs are pipelined, so only requires 4-8 plus RAM for cn  much more efficient than time domain for same level of image rejection

Thesis Goal: RFI Performance

slide-29
SLIDE 29

Front Back

slide-30
SLIDE 30

Incoherent Dedispersion

slide-31
SLIDE 31

Do you see the pulsar?

No Dedispersion

slide-32
SLIDE 32

BEE2 DRAM Circular Buffer

XAUI0 XAUI1 DRAM0 - 1 GB DRAM1 – 1 GB 10GbE

 2^k sub-buffers per DIMM, k = 0…8   500ms to ~2ms @ 2 Gsps

Up to 20 Gbps 16 Gbps for 2Gsps@8bits

slide-33
SLIDE 33

Goals

 Spectroscopy/Polarimetry  Currently (iBOB based):

  • 4096 ch spectrometer single pol
  • Dual 512 and Single 1024 ch fast dump

for pulsar

 Goal: spectrometer with “zoom” mode

  • Need to add enable to PFB-FIR block and

VACC block

 Hope to do RFI excision

slide-34
SLIDE 34

iADC Stability Tests Preliminary!

slide-35
SLIDE 35
  • 6dBm Anritsu, 3dB modulation
slide-36
SLIDE 36

+0dBm Anritsu, 3dB modulation +0dBm noise, 20dB atten before ADC

slide-37
SLIDE 37
  • 6dBm Anritsu, 3dB modulation
slide-38
SLIDE 38

9dB modulation, 0dBm Anritsu

slide-39
SLIDE 39

6dB modulation, 0dBm, +/-5% scale

slide-40
SLIDE 40

6dB modulation -10dBm (10dB below total noise)

slide-41
SLIDE 41

No Input

slide-42
SLIDE 42
  • 6dBm Anritsu, 3dB modulation no noise
slide-43
SLIDE 43

Goals

 Pulsar Observations  Transient (giant pulse) capture

  • Incoherent trigger