IPF 2011: Superconductivity and wideband telecommunication

In general, the cooling required to put even high-temperature superconductors into their zero-resistance state is expensive and inconvenient. To succeed, a commercial application of superconductivity should satisfy a requirement identified by one of the pioneers of superconductivity, Ivar Giaever: For whatever device performance you want to achieve, superconductivity has to be the only solution.

Giaever’s requirement was invoked by Oleg Mukhanov in his talk at the 2011 Industrial Physics Forum in Dallas, Texas. Mukhanov works for Hypres , a company founded in 1983 to develop and commercialize applications of superconductor electronics. Based in Elmsford, New York, Hypres markets several products. Among the most recent and most promising is a wideband telecommunications receiver.

The key to packing as much information as possible into a radio-frequency signal is to operate in a wide frequency band. With current semiconductor technology, this is very difficult at high radio frequencies, especially when it comes to extracting the signal. The receivers are expensive, complex, and inflexible, and they suffer a range of performance problems.

Those drawbacks arise, Mukhanov explained, because semiconductor circuitry can’t keep up with gigahertz signals. To cope with a signal that is both high frequency and wideband, a conventional receiver first puts the signal through a series of analog band-splitting and down-conversion steps. The signal’s digital content is then extracted and processed using semiconductor narrow-band analog-to-digital converters and processors. As a result, the signal’s bandwidth has to be divided into several channels, each of which must be processed by a separate receiver chain. The division adds noise and distortions to the signal, reducing its overall quality.

A typical semiconductor-based wideband receiver has several channels, each of which has its own low-band filter, local oscillator, low-pass filter, and analog-to-digital converter. The parallel architecture is inherently inflexible and cannot be reconfigured through programming commands. Dividing the bandwidth into a different set of channels would require installing new chains of processing hardware.

Hypres’s new, superconductor-based receiver is simpler, cheaper, and more flexible. At its heart is an integrated circuit that includes a niobium RSFQ (rapid single flux quantum) analog-to-digital converter and digital signal processor. At around a few picoseconds, the RSFQ circuit’s characteristic response time is short enough to process multi-gigahertz signals.

What’s more, the Hypres receiver needs only one low-cost analog wideband filter, not half a dozen or so higher-cost narrow-band ones; dispenses with local oscillators; and can handle a wideband signal without splitting it into channels. If the signal is reconfigured for different frequency band locations, the receiver can be preprogrammed. No new hardware is needed.

Interestingly, the Hypres RSFQ digital technology itself depends on another innovative technology: the compact 4-kelvin cryocoolers that cryocooler companies developed for various applications including NASA space missions.

Mukhanov told his audience that the US Navy recently placed an order for cryogenic-filter and low-noise-amplifier technology. Whatever qualms the navy has about using cryogenic technology are evidently allayed by the prospect of superior performance.

Charles Day