4.3 Experimental Device Performance
79
f respectively. In the 0
◦ and 90
◦ cases, the orientation angle of the pulses remains
close to the target values (represented by the dashed lines) and the ellipticity remains
less than 4
◦ between 0.6 and 4.5 THz, demonstrating that the generated pulses are
linearly polarised over the entire broadband frequency range. In the 45
◦ case, there is
more variation in both the ellipticity and the orientation angle, particularly at higher
frequencies, however the polarisation state has clearly been rotated in between the
0
◦ and 90
◦ cases.
The peak THz electric field strength obtained from electro-optic sampling was
3.4 kVm
−1 , using Eq.2.5. This is in agreement with the average power of 1 µW
measured by a calibrated pyroelectric detector. Comparable emission strengths were
obtained from reference devices with one large pixel, which have the same layout as
the device used in Sect. 3.2.1. This demonstrates that these multi-pixel devices have a
comparable efficiency to the standard interdigitated PCE design, which suggests they
do not suffer from any significant destructive interference of the radiation emitted
from separate pixels.
4.3.2 Generating Arbitrary Linear polarisation States via
Electrical Control
Having proved the principle of rotating the polarisation state of the emitted THz
pulses by varying the relative bias voltage applied to each set of contacts, the characteristics of the emitted THz radiation were investigated as the orientation angle of
the polarisation state was varied over a 360
◦ range. In order to rotate the polarisation
state, the bias voltages applied to the horizontally and vertically emitting pixels were
varied according to
V H = V max cos(ψ T ),
(4.6)
V V = V max sin(ψ T ),
(4.7)
where V max = 10 V and ψ T is the target orientation angle. The bias voltages applied
to produce each ψ T are shown in the inset of Fig. 4.5a. The target polarisation angle
was varied in 22.5
◦ steps and polarisation-resolved time-domain traces taken at each
step.
The maximum amplitude, E max =
E 2
x + E 2
y , at the time-domain peak of the
THz pulse at each target angle is reported in Fig. 4.5a, normalised to the mean value.
The variation in amplitude is less than ±15% over the 360
◦ range, demonstrating
that the device maintains a fairly consistent THz emission strength at all angles.
Fig. 4.5b shows the orientation angle of the emitted THz pulses, averaged over their
0.3−5.0 THz bandwidth, versus the target orientation angle. The measured orientation angle remains close to the target angle for all points, with the ideal case
represented by the dashed line, demonstrating that the orientation angle of the emit-
79
f respectively. In the 0
◦ and 90
◦ cases, the orientation angle of the pulses remains
close to the target values (represented by the dashed lines) and the ellipticity remains
less than 4
◦ between 0.6 and 4.5 THz, demonstrating that the generated pulses are
linearly polarised over the entire broadband frequency range. In the 45
◦ case, there is
more variation in both the ellipticity and the orientation angle, particularly at higher
frequencies, however the polarisation state has clearly been rotated in between the
0
◦ and 90
◦ cases.
The peak THz electric field strength obtained from electro-optic sampling was
3.4 kVm
−1 , using Eq.2.5. This is in agreement with the average power of 1 µW
measured by a calibrated pyroelectric detector. Comparable emission strengths were
obtained from reference devices with one large pixel, which have the same layout as
the device used in Sect. 3.2.1. This demonstrates that these multi-pixel devices have a
comparable efficiency to the standard interdigitated PCE design, which suggests they
do not suffer from any significant destructive interference of the radiation emitted
from separate pixels.
4.3.2 Generating Arbitrary Linear polarisation States via
Electrical Control
Having proved the principle of rotating the polarisation state of the emitted THz
pulses by varying the relative bias voltage applied to each set of contacts, the characteristics of the emitted THz radiation were investigated as the orientation angle of
the polarisation state was varied over a 360
◦ range. In order to rotate the polarisation
state, the bias voltages applied to the horizontally and vertically emitting pixels were
varied according to
V H = V max cos(ψ T ),
(4.6)
V V = V max sin(ψ T ),
(4.7)
where V max = 10 V and ψ T is the target orientation angle. The bias voltages applied
to produce each ψ T are shown in the inset of Fig. 4.5a. The target polarisation angle
was varied in 22.5
◦ steps and polarisation-resolved time-domain traces taken at each
step.
The maximum amplitude, E max =
E 2
x + E 2
y , at the time-domain peak of the
THz pulse at each target angle is reported in Fig. 4.5a, normalised to the mean value.
The variation in amplitude is less than ±15% over the 360
◦ range, demonstrating
that the device maintains a fairly consistent THz emission strength at all angles.
Fig. 4.5b shows the orientation angle of the emitted THz pulses, averaged over their
0.3−5.0 THz bandwidth, versus the target orientation angle. The measured orientation angle remains close to the target angle for all points, with the ideal case
represented by the dashed line, demonstrating that the orientation angle of the emit-