Noise extended Technologies Phase Noise measurements using Fiber Optic Delay Lines With contributions from Guillaume De Giovanni www.noisext.com
Phase Noise measurements 2 phase noise measurement types: Absolute also called total phase noise or output phase noise Residual also called added phase noise 2 types of DUT: Sources: Oscillators, Synthesizers, frequency translation devices viewed as a frequency sources (their Local Oscillator) Others: Amplifiers, cables, frequency translation devices, all viewed as a 2 ports phase coherent device (output vs input) 2
Absolute Phase Noise demodulation Direct Phase detection: Compares a reference signal to the signal to test Traditional phase noise analyzers Requires a same or better phase noise frequency source with DC-coupled FM port Measures the phase noise sum of the Reference and the DUT FM discriminator Phase extraction: Measures Frequency variations (noise) Does not need any reference signal State of the Art dynamic range solution 3
Frequency to Phase Conversion PHASE FREQUENCY Limited performance very close to the carrier but excellent dynamic range at mid/high offsets.
Frequency discriminator Phase Noise measurement based on Frequency Noise Measurement Path 2 τ = Path 2 Path 1 in seconds Path 1 Noise voltage (fm) = Kφ 2πτΔf(fm) 5 sin(πτ fm) πτfm
Measurement Diagram Front-end DCNTS 6
DCNTS (receiver section) noise floor 0.0-10.0-20.0-30.0 Spectrum Type L(fm) dbc/hz DCNTS Ch A 3.8GHz Ch B 3.8GHz XCOR 3G8 Res PN with filter -40.0-50.0-60.0-70.0-80.0-90.0-100.0-110.0-120.0-130.0-140.0-150.0-160.0-170.0-180.0-190.0-200.0 10 100 1k 10k 100k 1M 10M 100M 7
Keys to success Lasers with low Rin and low noise external modulator Thermally stabilized fiber spools Ultra low noise amplifiers (after photodiodes) Uncorrelated channels, no cross-talk Use of Innovative Fiber Spools with orthogonal behavior Software controlled Phase shifters Optimized Signal Processing Flow 8
Fiber Delay Residual Phase Noise RF over fiber link own phase noise requires 2 links 70 80 Bras B Bras A 90 100 Sφ (db rad 2 /Hz) 110 120 130 140 150 160 10 1 10 2 10 3 10 4 10 5 Fréquence (Hz) Average L(fm) for 1 fiber link: 10 Hz -126 dbc/hz 100-136 1k -142 10k -146 100k -148 Joint work with Femto-ST, Besancon, France 9
Absolute Phase Noise floor Simulation L(fm) for 50 us (10 km) single fiber channel: 10 Hz -75 dbc/hz 100-107 1k -132 10k -152 100k N/A for 50 us Still with no cross-correlations! But 10,000 xcor = 20dB better Potential noise floor: -172 dbc/hz @10kHz offset 10 10
Absolute Phase Noise floor Measured (using a State of the Art source or 2 sources) L(fm) for 10 us (2 km) and 200 xcor, simulation is: 10 Hz -75 dbc/hz 100-105 1k -132 10k -155 Measurement matches Simulation within 2 db Note: Measurement is from an older prototype than actual measured data used in simulation and fiber link had a higher residual noise but matches Simulation on old data within 2 db. Courtesy of JPL & Cal Tech for NASA 11
Test Source at 10 GHz with 1x 20us DL 0.0 Spectrum Type L(fm) dbc/hz 20u 10G amp zx60 FM DCNTS -10.0 20u 10G amp zx60 PM -20.0-30.0-40.0-50.0 PHASE -60.0-70.0-80.0-90.0-100.0-110.0-120.0-130.0-140.0 FREQUENCY -150.0-160.0-170.0 SinX/X response -180.0-190.0 10 100 1k 10k 100k Joint work with Thales Research & Technology, Palaiseau, France 12 1M
OEO at 10 GHz with 2x 5us DL 0.0 Spectrum Type L(fm) dbc/hz DCNTS -10.0-20.0-30.0-40.0-50.0-60.0-70.0-80.0-90.0-100.0-110.0 OEO modes spurious -120.0-130.0-140.0-150.0-160.0-170.0-138dBc/Hz @ 10kHz -180.0-190.0 10 100 1k 10k 100k Joint work with Thales Research & Technology, Palaiseau, France 13 1M
Conclusion Fiber Optics can help measure Ultra-low Phase Noise State of the Art Phase Noise measurements can be done without the need of expensive reference sources The design of the Fiber Optics links is key in obtaining high performance We are looking for partners and customers to put such instrument on the market: -170dBc/Hz @10kHz offset for an X band signal 14
Thank you! Questions and Answers 15