rsa_digital_demod.py

"""
VISA: RSA Digital Demod
Author: Morgan Allison
Date created: 1/17
Date edited: 2/17
This program sets up RSA5k/SignalVu-PC/SignalVu remotely to
acquire and demodulate a 1 MHz QPSK signal with RRC filter 
and alpha of 0.3.
Windows 7 64-bit, TekVISA 4.0.4
Python 3.6.0 64-bit (Anaconda 4.3.0)
NumPy 1.11.2, MatPlotLib 2.0.0, PyVISA 1.8
To get PyVISA: pip install pyvisa
Download Anaconda: http://continuum.io/downloads
Anaconda includes NumPy and MatPlotLib
Download SignalVu-PC programmer manual: http://www.tek.com/node/1828803
Download RSA5100B programmer manual: 
http://www.tek.com/spectrum-analyzer/inst5000-manual-7
Tested on RSA306B, RSA5126B and MSO73304DX with SignalVu
"""

import visa
import numpy as np
import matplotlib.pyplot as plt

"""#################SEARCH/CONNECT#################"""
rm = visa.ResourceManager()
inst = rm.open_resource('TCPIP::192.168.1.11::INSTR')
# inst = rm.open_resource('GPIB8::1::INSTR')
inst.timeout = 25000
inst.encoding = 'latin_1'
inst.write_termination = None
inst.read_termination = '\n'
instId = inst.query('*idn?')
print(instId)
if 'MSO' in instId or 'DPO' in instId:
    # IMPORTANT: make sure SignalVu is already running if you're using a scope.
    # The application:activate command gives focus to SignalVu.
    # *OPC? does not respond to application:activate, so there's no good 
    # way to synchronize this command
    print('Configuring SignalVu on instrument.')
    inst.write('application:activate \"SignalVu Vector Signal Analysis Software\"')

    # Undocumented command to disconnect SignalVu from sample rate control
    inst.write('sense:signalvu:acquisition:control:sample:rate 0')

# preset, clear buffer, and stop acquisition
inst.write('system:preset')
inst.write('*cls')
inst.write('abort')

"""#################CONFIGURE INSTRUMENT#################"""
freq = 1e9
span = 40e6
refLevel = 0

# set up spectrum acquisition parameters
inst.write('spectrum:frequency:center {}'.format(freq))
inst.write('spectrum:frequency:span {}'.format(span))
inst.write('input:rlevel {}'.format(refLevel))

# open new displays
inst.write('display:ddemod:measview:new conste')  # constellation
# inst.write('display:ddemod:measview:new stable') # symbol table
inst.write('display:ddemod:measview:new evm')  # EVM vs Time

# turn off trigger and disable continuous capture (enable single shot mode)
inst.write('trigger:status off')
inst.write('initiate:continuous off')

# configure digital demodulation (QPSK, 1 MSym/s, RRC/RC filters, alpha 0.3)
symRate = 10e6
alpha = 0.3

inst.write('sense:ddemod:modulation:type qpsk')
inst.write('sense:ddemod:srate {}'.format(symRate))
inst.write('sense:ddemod:filter:measurement off')
inst.write('sense:ddemod:filter:reference rcosine')
inst.write('sense:ddemod:filter:alpha {}'.format(alpha))
inst.write('sense:ddemod:symbol:points one')
inst.write('sense:ddemod:analysis:length 20000')
# print(inst.query('sense:acquisition:samples?'))

# start acquisition
inst.write('initiate:immediate')
inst.timeout = 100000
# wait for acquisition to finish
inst.query('*opc?')

# query results from the constellation display (details in programmer manual)
results = inst.query('fetch:conste:results?')

# remove terminating whitespace (.rstrip()), split the string result (.split())
# into an array of 3 values, and convert those values to floating point numbers
results = results.rstrip().split(',')
evm = [float(value) for value in results]

# print out the results
# see Python's format spec docs for more details: https://goo.gl/YmjGzV
print('EVM (RMS): {0[0]:2.3f}%, EVM (peak): {0[1]:2.3f}%, Symbol: {0[2]:<4.0f}'
      .format(evm))

# get EVM vs time data
evmVsTime = inst.query_binary_values('fetch:evm:trace?')

# determine points per symbol to create a plottable x axis
ptsPerSymbol = inst.query('sense:ddemod:symbol:points?')
if ptsPerSymbol == 'ONE\n':
    ptsPerSymbol = 1.0
elif ptsPerSymbol == 'TWO\n':
    ptsPerSymbol = 2.0
elif ptsPerSymbol == 'FOUR\n':
    ptsPerSymbol = 4.0
elif ptsPerSymbol == 'EIGHT\n':
    ptsPerSymbol = 8.0
else:
    ptsPerSymbol = 4.0  # default

symbol = np.array(range(len(evmVsTime))) / ptsPerSymbol

plt.plot(symbol, evmVsTime)
plt.title('EVM vs Symbol #')
plt.xlabel('Symbol')
plt.ylabel('EVM (%)')
plt.show()

inst.close()