Universal electronic synthesis by microresonator-soliton photomixing (2505.08707v1)
Abstract: Access to electrical signals across the millimeter-wave (mmW) and terahertz (THz) bands offers breakthroughs for high-performance applications. Despite generations of revolutionary development, integrated electronics are challenging to operate beyond 100 GHz. Therefore, new technologies that generate wideband and tunable electronic signals would advance wireless communication, high-resolution imaging and scanning, spectroscopy, and network formation. Photonic approaches have been demonstrated for electronic signal generation, but at the cost of increased size and power consumption. Here, we describe a chip-scale, universal mmW frequency synthesizer, which uses integrated nonlinear photonics and high-speed photodetection to exploit the nearly limitless bandwidth of light. We use a photonic-integrated circuit to generate dual, microresonator-soliton frequency combs whose interferogram is fundamentally composed of harmonic signals spanning the mmW and THz bands. By phase coherence of the dual comb, we precisely stabilize and synthesize the interferogram to generate any output frequency from DC to >1000 GHz. Across this entire range, the synthesizer exhibits exceptional absolute fractional frequency accuracy and precision, characterized by an Allan deviation of 3*10-12 in 1 s measurements. We use a modified uni-traveling-carrier (MUTC) photodiode with an operating frequency range to 500 GHz to convert the interferogram to an electrical signal, generating continuously tunable tones across the entire mmW band. The synthesizer phase noise at a reference frequency of 150 GHz is -83 dBc/Hz at 100 kHz offset, which exceeds the intrinsic performance of state-of-the-art CMOS electronics. Our work harnesses the coherence, bandwidth, and integration of photonics to universally extend the frequency range of current, advanced-node CMOS microwave electronics to the mmW and THz bands.
Paper Prompts
Sign up for free to create and run prompts on this paper using GPT-5.
Top Community Prompts
Collections
Sign up for free to add this paper to one or more collections.