-
Measurement setup: An HP 85106D vector network analyzer (VNA) with
WR-10 waveguide extensions was used for all measurements, calibrated from
68 to 118 GHz. The OMT main arm port (the angled one) was connected via
a 1" waveguide twist to VNA Port 2. The OMT side arm port was directly
connected to VNA Port 1. For some of the measurements (dated 26 Oct 2001)
the twist was not calibrated out; to view the reflect response of
the twist alone (terminated with a short), refer to [3].
The common (square waveguide) port is terminated with either a special
test load, or with a short. All measurements were taken at ambient (room)
temperature.
-
OMT 401-02, 401-3 Measurements: Assembly and measurement of these
pre-production units were done October 2001 at the Academia Sinica Institute
of Astronomy and Astrophysics (ASIAA), in Taipei, Taiwan by Dr. Ming-Tang
Chen. Two sets of data were taken, dated 20 Oct 2001 and 26 Oct 2001, and
are referenced in the following sections as [1] and [2], respectively.
The later set includes measurements of 401-03 with the polarizing pin in
each side arm closest to the common port removed.
-
Insertion loss: For insertion loss measurement of both arms, the
common port was terminated with a short. Reflection losses |S11| and |S22|
then correspond (to a close approximation) to twice the one-way loss through
the OMT side arm port and main arm port (plus external waveguide twist,
if not calibrated out), respectively.
The side arm loss is taken from |S11| measurements for OMT 401-02
[1,], OMT
401-03 [1,2],
and OMT 401-03 w/missing pins [2].
In all plots it is consistently around 0.3 dB over most of the band, increasing
to 0.5 dB at the lower band edge. This is around twice the expected loss;
this discrepancy was traced to the type of plating used (99.0% hard gold)
which is considerably more lossy than soft (99.9%) gold used on the prototype.
The OMTs will be replated and remeasured in the near future.
The main arm loss (without 1" w.g. twist) is taken from |S11| measurements
for OMT 401-02 [1]
and OMT 401-03 [1].
The later set of measurements for OMT 401-02 [2],
OMT 401-03 [2]
and OMT 401-03 w/missing pins [2]
include the loss of the twist adapter. The main arm loss is slightly higher
(~0.4 dB) than that of the sidearm, and increases to 0.55 dB at the lower
band edge. Again, this is higher than expected, mainly due to the gold
plating as with the side arm. However, the theoretical loss of both the
main arm and side arm are equal, which is clearly not the case here. One
possible explanation is that slight deformation or misalignment of the
septum can affect the main arm loss, and also introduce the ripple seen
in the plots.
-
Port Match: For return loss measurement of both arms, the common
port was terminated with the test load. The return losses |S11| and |S22|
then correspond the OMT side arm and main arm port match, respectively.
A high-precision test load termination (RLoss > 40 dB) is required to accurately
measure the match.
The main arm and side arm return loss data for OMT 401-02 [2],
OMT 401-03 [2]
and OMT 401-03 w/missing pins [2]
shows the ports are well matched, particularly the main arm. Note that
removing a polarizing pin introduces a high-frequency resonant dip in the
sidearm match, while causing an slight overall degradation in match on
the main arm. This is consistent with similar measurements at Ka-band [4].
-
Isolation: For measurement of isolation between main and side arm
ports, the common port was terminated with the test load, and |S21| or
|S12| taken. The isolation data is shown for OMT 401-02 [2],
OMT 401-03 [2]
and OMT 401-03 w/missing pins [2],
and is around 40 dB, the value expected from simulations.