The waveguide mixer follows the design of Walker et al. (1992) and
consists of a corrugated feedhorn followed by a circular to full height
rectangular transition and a waveguide ``T'' section. The vertical part
of the ``T'' contains a sliding, rectangular backshort, often referred to
as the E-plane backshort. The SIS junction is located one-half wavelength
behind the E-plane backshort. The junction is formed on a quartz substrate
with dimensions 
mils (millinches). The full
height rectangular waveguide has dimensions of 37 
mils.
Behind the junction is a second sliding backshort referred to as the
junction backshort. One side of the junction substrate is connected
to ground. The ``hot'' side of the substrate is connected to an impedance
matching network via a 1.0 mil diameter gold wire. The wire is soldered to the
matching network and silver painted (by hand!) to the tip of the junction
substrate. The purpose of the matching network is to transform the I.F.
output impedance (160 
, if we are lucky) to 50 , which
is the input impedance of the first I.F. amplifier. For the mixer to
provide optimum operation, a magnetic field of about one quantum
flux unit must be maintained across the junction. The magnetic field
works to suppress the Cooper pairs tunneling across the junction, which
are a source of both noise and instability. High mu metal field concentrators
are embedded in the mixer block for this purpose. The required magnetic field
is generated by a superconducting electromagnet attached to the mixer
via the field concentrators. The intensity of the magnetic field can be
varied by adjusting the bias voltage and current across it. Typically,
a magnet current of 
35 mA is sufficient for stable operation
of the mixer.