ADC5G Testing

This repository contains scripts and bitcodes for testing the ASIAA 5 GSps ADC (https://casper.berkeley.edu/wiki/ADC1x5000-8) on ROACH1 and ROACH2.

Requirements

• Python >= 2.7
• numpy >= 1.6.2 (only for the 'array' object)
• corr >= 0.6.9 (only for katcp_wrapper.FpgaClient)

Getting Started

Begin with connecting to your ROACH and programming it:

from corr import katcp_wrapper
roach = katcp_wrapper.FpgaClient(host, port)
roach.is_connected() # should return True
roach.progdev(bitcode)
roach.est_brd_clk()

The last line should return the approximate clock rate of the FPGA fabric; this should be close to adc_clk/8 where adc_clk is the frequency of the sine wave you provide the ADC.

Calibrating data-to-clk

Glitches may occur if the rising-edge of the iSerDes capture clock (which is generated by a PLL locked to the externally provided ADC clock) does not occur within the data "eye", which can occur due to routing delays or component delays. Luckily the e2v EV8AQ160 provides a ramp test-vector which can be used to detect these glitches. The phase-relation of the input/capture clocks can then be adjusted until these glitches disappear. The 'adc5g' package provides a function for doing so:

from adc5g import *
set_test_mode(roach, 0)
set_test_mode(roach, 1)
sync_adc(roach)
opt0, glitches0 = calibrate_mmcm_phase(roach, 0, ['scope_raw_0_snap',])
opt1, glitches1 = calibrate_mmcm_phase(roach, 1, ['scope_raw_1_snap',])
unset_test_mode(roach, 0)
unset_test_mode(roach, 1)

The calibration steps through all 56 phases and attempts to find the optimal one, i.e. the one with zero glitches furthest away from the peak. The function returns the optimal phase it found and an array containing the number of glitches detected in the ramp at each phase step.

Taking a Snapshot

If you simply wish to take a quick snapshot you can use the following command with snap_name being the name of the snapshot block in your design:

import numpy
raw = numpy.array(adc5g.get_snapshot(roach, snap_name))

If you are using the digicom bof files provided in the 'boffiles' directory then snap_name="raw_%z"%z where z is the ZDOK number of the ADC you wish to capture data from, either 0 or 1.

Running the Test Suite

If you want to run the full test suite from a remote machine over TCP/IP using tcpborphserver do the following from the root directory of this repository:

python2.7 test_adc5g.py -v -r roach_name 

where roach_name is the IP or hostname of the ROACH on the network. By default this will test only the ADC in ZDOK 0 of the ROACH; to test the one in ZDOK 1 add -z 1. If you'd like to see a list of options you can use --help.