A DC SQUID amplifier with a novel tuning circuit

Radio-frequency amplifiers based on dc SQUIDs

“…Hilbert and Clarke [11] found a noise temperature of about 4 K at a frequency of 100 MHz for a SQUID with G P around 19 dB operated at 4.2 K, and they obtained T N of about 1 K for a bath temperature of 1 K. These noise temperatures are somewhat above the predicted values. Takami et al [15] measured T N ∼ 0.7 K for G P ∼ 20 dB and T = 4.2 K at an operating frequency of 150 MHz, which is also slightly above the predicted value. Prokopenko et al [18] measured a noise temperature of 4 K (G P ∼ 20 dB) at 3.6 GHz; in subsequent work [19], they found T N ∼ 2 K (G P ∼ 12 dB) at 4 GHz, see figure 9.…”

“…The frequency-dependent impedance of the input circuit can be matched to a 50 source using appropriate matching networks. Takami et al [15] suggested a matching network consisting of two capacitors and an inductor (see figure 4) to step up the impedance of a 50 source to a higher impedance across the input circuit of the SQUID. By making the input circuit resonant, the rf current flowing in the input coil can be enhanced by the quality factor Q of the input circuit, at the expense of a reduction in the amplifier bandwidth.…”

“…To a first order, the input impedance of a SQUID amplifier can be considered to be that of its input circuit. In the case where the input circuit is a simple inductor [11,15,20,33], one expects the input impedance Z i to be approximately j ωL: provided the input coil has no losses, Z i is purely reactive. For the microstrip SQUID amplifier [22], one expects the input impedance Z i ∼ ∞ on resonance.…”

“…Note that larger transformation ratios will lower the bandwidth of the device substantially. Takami et al [15] and Prokopenko et al [18,19,46] used conventional LC networks to impedance match their devices to a 50 source. No data about the input reflection coefficient were given in these publications.…”

“…Mück and Clarke [53] also measured the harmonic generation of a SQUID amplifier (see figure 14) for different input fluxes and found the measured data in good agreement with calculations that assume a sinusoidal transfer function for the SQUID. Another way to characterize the power-dependent gain of an amplifier is through the so-called 'saturation temperature' [15,54]. Here it is assumed that the input signal is thermal (broadband) noise from a resistor at a certain temperature.…”