Multiquantum EPR bridge design and noise performance characterization

Robert Alan Strangeway, Marquette University

Abstract

In a multiquantum electron paramagnetic resonance (EPR) spectrometer, a biological sample is irradiated with two or more coherent microwave frequencies with a frequency spacing generally from 1 to 100 kHz. Multiquantum EPR (MQEPR) bridges were initially developed in 1990 utilizing a single sideband modulator. In this bridge, transmission resonators and reflection traps were used to reduce intermodulation sidebands, generated by the single sideband modulator, that are identical in frequency with the EPR signals. It was also difficult to use by non-microwave engineers; hence, the need for an alternative MQEPR bridge design was established. The first major problem solved in this work is the selection and design of a microwave bridge configuration which will function as a MQEPR bridge and meet the specific requirements that it generates two or three closely spaced coherent microwave (X-band) frequencies separated on the order of 1 kHz to 100 kHz, is straightforward to operate, generates 23 dBm per frequency with spurious content ${\le}{-}70$ dBc incident on the loop-gap resonator, is tunable over a 200 MHz bandwidth, and minimizes intermodulation product generation in the receiver portion. The double sideband/fixed filter configuration was selected. It generates multiple signals by using synthesizers and mixers to translate the microwave oscillator frequency, filters to isolate the desired mixer sideband, and one amplifier per bridge arm to eliminate intermodulation products in the incident irradiation. The performance of this configuration meets the specific requirements. The second major problem solved in this work is the identification and characterization of the sources of noise in the double sideband/fixed filter MQEPR bridge configuration. The sources of noise identified were receiver noise, the synthesizer noise floor, incoherent and coherent phase noise, and noise on receiver intermodulation products. The receiver sets the lower limit of the system noise floor at low incident power. At higher incident power, the incoherent phase noise of the synthesizers, $-$123 dBc/Hz at 100 kHz offset, sets the lower limit. The coherent phase noise of the microwave oscillator is suppressed to the first order.

This paper has been withdrawn.