Sketch of the experimental setup. The two SIS mixers (M1 and M2) of the balanced circuit are indicated with the -symbol and the crossed rectangle is the hybrid coupler (compare also with Fig. 3). The combination of the separately amplified IF signals of the balanced mixer ( and ) is done at room temperature (295 K) with a commercial IF hybrid coupler. Two separate blackbody loads, load 1 (LO port) and load 2 (signal port), are used at a temperature of either 77 K or 295 K. Their noise power is used to measure and near-carrier LO noise power simultaneously. The measurement is performed with a bias voltage sweep of either or while the other mixer bias voltage is kept constant within the first photon-assisted tunneling step of the SIS IV characteristic. The combined IF power is read-out at the output of the IF hybrid coupler. In the synthesizer driven LO a switch is used to choose one of the two power pre-amplifiers (PA1 or PA2) for the measurement. In this experiment load 1 is used as a 77 K termination of the LO port in order to determine the LO noise and is removed in an astronomical receiver. Windows, IR blocking filters and beamsplitter each have a frequency dependent transmissivity which is determined in Sec. II B.
In (a) for and in (b) for the effective load temperatures are calculated for frequencies in the measurement bandwidth. The thin solid lines in (a) and (b) show the effective load temperature referred to mixer input port A1 and the thick solid lines show referred to mixer input port A2. Equivalent circuit diagrams for the window-IR filter-beamsplitter and window-IR filter cascade in front of mixer input ports A1 and A2 are shown in (c) and (d).
Top: Balanced mixer circuit diagram showing the two mixer input ports A1 and A2, the hybrid coupler (crossed rectangle) and the two SIS mixers M1 and M2 having gains and . and are power coupling factors and is the phase error of a possibly not ideal hybrid coupler. Bottom: Detail photograph of the mixer chip focussing on the RF part of the circuit.
Schematic representation of the balanced SIS mixer IF output power trace for three cases (a)-(c). One of the two SIS mixers on the balanced mixer chip is constantly biased with a negative voltage within the first photon-assisted tunneling step. A voltage sweep is applied to the other mixer. Power coming from load 1 (LO port) is measured in the output () whereas power coming from load 2 (signal port) is measured in the output () of the IF output power trace. Thick solid lines show measurements where load 2 has a temperature of 295 K whereas thin solid lines show measurements with a load temperature of 77 K. Load 1 has a constant temperature of 77 K. For each case, it is indicated whether the noise rejection value (NR, Eq. (6)) is infinite or takes a finite value. (a) Ideal balanced mixer. Temperatures are equivalent values through the relation , where is the receiver gain, T is the input temperature of either 295 or 77 K and B is the IF bandwidth. The effect of near-carrier LO noise with equivalent noise temperature is to increase the power above the corresponding input noise power from load 1. (b) shows the effect of unequal mixer gain on the balanced mixer's IF output power for and or the effect of an asymmetry in the hybrid coupler for equal mixer gain and . The dashed line indicates the ideal case for the output in which the distance between the dotted lines in the output approaches zero. (c) Most likely situation during the experiment. In order to achieve the best balanced mixer performance, the difference between the two traces in the output has to be minimized while maximizing the difference of the two traces in the output (arrows) which is achieved by adjusting the phase shifter in the IF path (Fig. 1).
Total IF output power of the balanced SIS mixer as a function of measured over the IF frequency range 4–8 GHz with a power meter. Mixer M2 is constantly biased at a negative voltage on the first photon-assisted tunneling step. Therefore, for a positive voltage on the first photon-assisted tunneling step, the mixer works as balanced mixer. For a negative on the first photon-assisted tunneling step LO noise can be measured. For all other bias voltages , a shift between the traces for hot/cold calibration loads is observed indicated by the double arrow in (a). This is omitted in Fig. 4. Temperature values in the output of the figure indicate the effective temperature of load 2 as seen from mixer input port A2 (signal port). The output measures the total temperature as seen from mixer input port A1 (LO port), i.e., the effective temperature of load 1 together with a possible temperature contribution from the LO caused by near-carrier noise power. Arrows point to the bias region in which the temperature is measured. The traces in (a) and (b) were measured using the synthesizer driven LO whereas the traces in (c) and (d) were measured using the Gunn driven LO.
Equivalent noise temperature (◼) and noise rejection NR (●) as a function of LO frequency . Datapoints labeled with the abbreviation SLO were measured using the synthesizer LO whereas datapoints labeled with GLO belong to measurements which used the Gunn LO.
Circuit diagram showing the measurement principle used to study the impact of LO noise power on a single-ended mixer. For an ideal hybrid coupler, half of the signals received by the mixer input ports A1 and A2 are detected by M2, whereas half of the signals are absorbed by mixer M1. This is the case for . M1 is operated in the normal conducting state with resistance . Because of symmetry, the same principle applies to M2 being operated in the normal conducting state where now M1 is used as mixer.
(a) of the balanced mixer for using a synthesizer LO. (b) Rectangles show the bias voltages of the two mixers M1 and M2 during the measurement of the trace shown in (a). The figure shows two LO pumped SIS IV curves which are almost equal demonstrating the optimal device performance for this frequency. In (c) a single-ended measurement with mixer M2 at is presented which uses the synthesizer LO and during which . The receiver noise temperature is corrected for the effective load temperatures with respect to both mixer input ports A1 and A2. (d) shows the result of a single-ended measurement using mixer M2 where again and for . Instead of using the synthesizer driven LO we employed a Gunn driven LO for the measurement which results in a similar performance like observed in (a).
Summary of the parameters used to determine the transmissivity and emissivity of the windows, IR blocking filters and the beamsplitter in the measurement bandwidth 380–520 GHz. The beamsplitter is rotated by with respect to the LO beam axis. d is the thickness and is the complex refractive index. The uncertainty (not shown in the table) of the values below is considered in the calculation of the uncertainty of the effective temperatures, summarized in Table II.
Summary of the LO noise measurements using a balanced SIS mixer. The column summarizing the values for includes a second number in brackets which is the value . The equivalent LO noise temperature, , is listed in a separate column. At 465 GHz, the Gunn driven LO has very little output power resulting in an unusual large value compared to the result using a similar frequency for the synthesizer driven LO.
Article metrics loading...
Full text loading...