def measurement(reading,channels,fpga,probe,g0,g1,g2,g3,generator,bw,LO,RF_power,fsteps):
print("LO frequency must be "+str(LO)+"[GHz] (manual set)")
print( "Current RF power is set to: "+ str(RF_power) + "dBm")
generator.write("power "+repr(RF_power)+"dbm\r\n")
generator.write("Output on\r\n")
for channel_number in range(0,channels,fsteps):
print("##########################################################")
print(" Current channel = "+str(channel_number)+" ")
print("##########################################################")
# A calculation to see what the current frequency should be (in MHz) and setting the signal generator to the middle of a spectral channel
freq = bw/float(channels)*(channel_number)
freq = max(0.01,freq + LO*1000) # LO addition
print( "Current frequency is set to: "+ str(freq) + "MHz")
generator.write("freq "+repr(freq)+"mhz\r\n")
time.sleep(0.2)
# Read the data from the ROACH
spectrum_z1_a, spectrum_z1_c, spectrum_z0_a, spectrum_z0_c = calibrate_inputs.get_data(fpga, channels)
# Constructing the voltage vectors V. We are only interested in the channel where the tone is, the rest can be discarded.
V = np.array([spectrum_z1_a[channel_number], spectrum_z1_c[channel_number], spectrum_z0_a[channel_number], spectrum_z0_c[channel_number]])
# Computing the power of the spectra for plotting
power_spectrum_z1_a, power_spectrum_z1_c, power_spectrum_z0_a, power_spectrum_z0_c = calibrate_functions.channel_power(spectrum_z1_a, spectrum_z1_c, spectrum_z0_a, spectrum_z0_c, channels)
print("\nVoltage vector V = "+str(V)+"\n")
# Check if the channel with the maximum power correspond to the input frequency
V_power = 10*np.log10((1+abs(V)**2))
print("Powers on channel: ")
print(V_power)
print("")
# Constructing the matrix M = <VV*> where V is the vector composed of the channels
M = calibrate_functions.compute_m(V)
print("The cross-spectrum matrix M =")
print M
# Computing the magnitudes of M
magnitude = calibrate_functions.compute_mag(M)
print('\nMagnitude is equal to: '+str(magnitude)+'\n')
# Computing the phase of M
rel_phase = calibrate_functions.compute_relative_phase(M)
print('Relative phase is equal to:\n')
print rel_phase
# According to the probe settings the maximum is either user-defined or chosen by the maximum correlation
index = int(probe)
# Normalising the matrix and changing to the definite phases
if index == 0:
phase[0] = rel_phase[index,0]
phase[1] = rel_phase[index,1]
phase[2] = rel_phase[index,2]
phase[3] = rel_phase[index,3]
elif index == 1:
phase[0] = rel_phase[index,0]
phase[1] = rel_phase[index,1]
phase[2] = rel_phase[index,2]
phase[3] = rel_phase[index,3]
elif index == 2:
phase[0] = rel_phase[index,0]
phase[1] = rel_phase[index,1]
phase[2] = rel_phase[index,2]
phase[3] = rel_phase[index,3]
elif index == 3:
phase[0] = rel_phase[index,0]
phase[1] = rel_phase[index,1]
phase[2] = rel_phase[index,2]
phase[3] = rel_phase[index,3]
else:
print('No valid index for the phase')
# # Normalising the matrix and changing to the definite phases
# if index == 0:
# phase[0] = 0
# phase[1] = -rel_phase[index,1]
# phase[2] = -rel_phase[index,2]
# phase[3] = -rel_phase[index,3]
# elif index == 1:
# phase[0] = -rel_phase[index,0]
# phase[1] = 0
# phase[2] = -rel_phase[index,2]
# phase[3] = -rel_phase[index,3]
# elif index == 2:
# phase[0] = -rel_phase[index,0]
# phase[1] = -rel_phase[index,1]
# phase[2] = 0
# phase[3] = -rel_phase[index,3]
# elif index == 3:
# phase[0] = -rel_phase[index,0]
# phase[1] = -rel_phase[index,1]
# phase[2] = -rel_phase[index,2]
# phase[3] = 0
# else:
# print('No valid index for the phase')
# Writing the phases and magnitudes of the channels into a larger array. This is used to plot the increase of the phase and magnitude over the measurement.
for i in range(4):
total_magnitude[i][reading][channel_number] = magnitude[i]
total_phase[i][reading][channel_number] = phase[i]
print('\nNormalisation has been done, the results are: \nMagnitude: '+str(magnitude)+'\nPhase: '+str(phase)+'\n')
# See if the data is consistent (a.k.a. sanity check)
calibrate_functions.consistency_magnitude(M)
consistent_phase = calibrate_functions.consistency_phase(M)
# Plot the spectra
calibrate_plot.plot_calibration(channels,channel_number, power_spectrum_z1_a, power_spectrum_z1_c, power_spectrum_z0_a, power_spectrum_z0_c,magnitude,phase,total_magnitude, total_phase,g0,g1,g2,g3)
# Saving the coefficients:
for i in range(0,4):
G[i,reading,channel_number] = magnitude[i]*cos(phase[i])+1j*magnitude[i]*sin(phase[i])
# Saving the relevant plots
#g2.hardcopy('normalized_magnitude.ps', enhanced =1, color=1)
#g3.hardcopy('normalized_phase.ps', enhanced =1, color=1)
raw_input('press enter to continue')
return G
def measurement(channels, fpga,g0, g1, g2, g3, generator, LO, RF_power, fsteps, gain_matrix_g, s0):
print("LO frequency must be "+str(LO)+"[GHz] (manual set)")
print( "Current RF power is set to: "+ str(RF_power) + "dBm")
generator.write("power "+repr(RF_power)+"dbm\r\n")
generator.write("Output on\r\n")
bw = 600.0
s0_2 = s0*s0
for channel_number in range(0, channels, fsteps):
print("##########################################################")
print(" Current channel = "+str(channel_number))
print("##########################################################")
# A calculation to see what the current frequency should be (in MHz) and setting the signal generator to the middle of a spectral channel
freq = (bw/channels)*(channel_number)
freq = max(0.01, freq + LO*1000) # LO addition
print( "Current frequency is set to: "+ str(freq) + "MHz")
generator.write("freq "+str(freq)+"mhz\r\n")
#time.sleep(0.05)
spectrum_z1_a, spectrum_z1_c, spectrum_z0_a, spectrum_z0_c = calibrate_inputs.get_data(fpga, channels)
# Constructing the voltage vectors V. We are only interested in the channel where the tone is, the rest can be discarded.
V = numpy.array([spectrum_z1_a[channel_number], spectrum_z1_c[channel_number], spectrum_z0_a[channel_number], spectrum_z0_c[channel_number]])
gain_matrix_h = numpy.linalg.pinv( gain_matrix_g[:][:][channel_number])
# obtain the estimated values of given the partialy
# calibrated matrix
s_prima = gain_matrix_h*V
# convert from linear polarization to circular
to_circular_matrix = numpy.matrix('1 -1j; 1 1j')
s_circular = to_circular_matrix*s_prima
# get the square module of each polarization component
s_left_abs_2 = s_circular[0]*s_circular[0].conjugate()
s_rigth_abs_2 = s_circular[1]*s_circular[1].conjugate()
# get the value of eps given equation 22a of the paper
# compact orthomode transducers using digital polarizatio synthesis
sin_eps = (s_left_abs_2 + s_rigth_abs_2)/(2*s0_2) - 1.0
eps = numpy.arcsin(sin_eps)
# values that are needed for leater
cos_eps = numpy.cos(eps)
sin_2_eps = 2*sin_eps*cos_eps
# calculate phi given equation 22b
sin_phi = (s_left_abs_2 - s_rigth_abs_2)/(s0_2*(sin_2_eps + 2*cos_eps))
s_cross_product = s_circular[0]*s_circular[1]
cos_phi = 2*(numpy.imag(s_cross_product))/(s0_2*(sin_2_eps + 2*cos_eps))
# the rotations matrix given equaton 24
rotation_matrix_eps = numpy.matrix('1 %f;0 %f'.format(sin_eps,cos_eps))
rotation_matrix_phi = numpy.matrix('1 0;0 {:.2f}'.format(cos_phi + 1j*sin_phi))
gain_matrix_g[:][:][channel_number] = gain_matrix_g[:][:][channel_number] * rotation_matrix_phi * rotation_matrix_eps
return gain_matrix_g
