Origin of 1/f PM and AM noise in bipolar junction transistor amplifiers

In this paper we report the results of extensive research on phase modulation (PM) and amplitude modulation (AM) noise in linear bipolar junction transistor (BJT) amplifiers. BJT amplifiers exhibit 1/f PM and AM noise about a carrier signal that is much larger than the amplifiers thermal noise at those frequencies in the absence of the carrier signal. Our work shows that the 1/f PM noise of a BJT based amplifier is accompanied by 1/f AM noise which can be higher, lower, or nearly equal, depending on the circuit implementation. The 1/f AM and PM noise in BJTs is primarily the result of 1/f fluctuations in transistor current, transistor capacitance, circuit supply voltages, circuit impedances, and circuit configuration. We discuss the theory and present experimental data in reference to common emitter amplifiers, but the analysis can be applied to other configurations as well. This study provides the functional dependence of 1/f AM and PM noise on transistor parameters, circuit parameters, and signal frequency, thereby laying the groundwork for a comprehensive theory of 1/f AM and PM noise in BJT amplifiers. We show that in many cases the 1/f PM and AM noise can be reduced below the thermal noise of the amplifier.     

Up-converted 1/f PM and AM noise in linear HBT amplifiers

In this paper we describe a technique to predict the 1/f phase modulation (PM) and 1/f amplitude modulation (AM) noise due to up-conversion of 1/f baseband current noise in microwave heterojunction bipolar transistor (HBT) amplifiers. We obtain an accurate model for the amplifier and find the expression for voltage gain in terms of DC bias, transistor parameters, and circuit components. Theoretical 1/f PM and AM noise sensitivities to 1/f baseband current noise are then found by applying the definitions of PM and AM noise to the gain expression of the amplifier. Measurements of PM and AM sensitivities at 500 MHz and 1 GHz were in good agreement with the values predicted by theory, verifying the validity of this technique. This method can be used to optimize amplifier design for low PM and AM noise. We show that the amplifier PM noise can be reduced by 9 dB by adjusting the value of the input coupling capacitor.     

Residual phase noise measurements of VHF, UHF, and microwavecomponents

The results of residual phase noise measurements on a number of VHF, UHF, and microwave amplifiers, both silicon (Si) bipolar junction transistor (BJT) and gallium arsenide (GaAs) field effect transistor (FET) based, electronic phase shifters, frequency dividers and multipliers, etc., which are commonly used in a wide variety of frequency source and synthesizer applications are presented. The measurement technique has also been used to evaluate feedback oscillator components, such as the loop and buffer amplifiers, which can play important roles in determining an oscillator's output phase noise spectrum (often in very subtle ways). While some information has previously been published related to component residual phase noise properties, it generally focused on the flicker noise levels of the devices under test, for carrier offset frequencies less than 10 kHz. The work reported herein makes use of an extremely low noise, 500 MHz surface acoustic wave resonator oscillator (SAWRO) test source for residual phase noise measurements, both close-to-and far-from-the-carrier. Using this SAWRO-based test source at 500 MHz, we have been able to achieve a measurement system phase noise floor of -184 dBc/Hz, or better, for carrier offset frequencies greater than 10 kHz, and a system flicker phase noise floor of -150 dBc/Hz, or better, at 1 Hz carrier offset. The paper discusses the results of detailed residual phase noise measurements performed on a number of components using this overall system configuration. Several interesting observations related to the residual phase noise properties of moderate to high power RF amplifiers, i.e., amplifiers with 1 dB gain compression points in the range of +20 to +33 dBm, are highlighted     

Guidelines for designing BJT amplifiers with low 1/f AM and PM noise

In this paper we discuss guidelines for designing linear bipolar junction transistor amplifiers with low 1/f amplitude modulation (AM) and phase modulation (PM) noise. These guidelines are derived from a new theory that relates AM and PM noise to transconductance fluctuations, junction capacitance fluctuations, and circuit architecture. We analyze the noise equations of each process for a common emitter (CE) amplifier and use the results to suggest amplifier designs that minimize the 1/f noise while providing other required attributes such as high gain. Although we use a CE amplifier as an example, the procedure applies to other configurations as well. Experimental noise results for several amplifier configurations are presented.     

Secondary standard for PM and AM noise at 5, 10, and 100 MHz

A practical implementation of a portable secondary standard for phase modulation (PM) and amplitude modulation (AM) noise at 5, 10, and 100 MHz is described. The accuracy of the standard for both PM and AM noise is +0.14 dB, and the temperature coefficient is less than 0.02 dB/K. The noise floor S_φ (10 kHz) of the standard for PM noise measurements is less than -190 dBC relative to 1 rad²/Hz at 5, 10, and 100 MHz. The noise floor for AM measurements depends on the configuration. A calibrated level of PM and AM noise of approximately -130±0.2 dB relative to 1 rad² /Hz (for Fourier frequencies from approximately 1 Hz to 10% of the carrier frequency) is used to evaluate the accuracy versus Fourier frequency. Similar PM/AM noise standards are under test at 10 GHz. This new standard can also be used as an alternative to the normal method of calibrating the conversion sensitivity of the PM/AM detector for PM/AM measurements. Some types of time-domain measurement equipment can also be calibrated     

Measurement of static and vibration-induced phase noise in UHF thin-film resonator (TFR) filters

Measurements of the static phase noise and vibration sensitivity of thin-film resonator (TFR) filters operating at 640 and 2110 MHz have been made. They show that the short-term frequency instability of the filters is small compared with that induced in the oscillator signal by the sustaining stage amplifier PM (phase modulation) noise. In-oscillator measurement of filter performance under vibration indicates that fractional frequency vibration sensitivities (δf ₀/f₀) are on the order of several parts in 10^-9/g. Because the percentage bandwidth and order (number of poles) of the filters was fairly constant, so was the product of the center frequency and group delay. Thus, the fractional frequency vibration sensitivity of the filters can be expressed alternatively as carrier signal phase sensitivity to vibration. The τ-ω₀ product for the filters that were tested was on the order of 300 rad, so that the equivalent phase sensitivity to vibration was approximately 1 grad/g     

Low phase noise operation of microwave oscillator circuits

In this paper, we describe a theoretical basis, leading to new results, on the general conditions to be fulfilled by oscillator circuits to achieve a very low phase noise. Three main conditions must be fulfilled by a transistor oscillator circuit to reach the minimum phase noise. The energy stored in the resonator must be maximum. Its transfer to the controlling voltage port of the transistor current source must be first maximized. A possible conversion noise at the transistor output port will be also minimized by maximizing the energy transferred to that port. The proposed method has been applied to an experimental oscillator set up with a PHEMT transistor. A state-of-the-art phase noise of -80 dBc/Hz at 100 Hz offset from carrier with a 1/f(3) slope has been measured at room temperature with a 9.2 GHz, oscillator. The application of these new results to free-running oscillator circuits with one-stage then multistage transistor amplifiers demonstrate clearly the validity of the design method. The efficiency of this design method and its ease of use represent a real breakthrough in the field of low noise transistor oscillator circuit design.     

Correlation between upper and lower sidebands

Experimental measurements supported by a simple model show that the upper and lower phase modulation (PM) noise sidebands are always equal and 100% correlated, independent of the noise power originating from multiplicative or additive processes. Similarly, we show that the upper and lower amplitude modulation (AM) noise sidebands are also equal and 100% correlated, independent of the noise power originating from multiplicative or additive processes, Moreover, the single sideband (SSB) PM noise is always equal to one-half the total PM noise. The same is true for the AM noise. Although the upper and lower PM or AM noise sidebands are equal and correlated for broadband additive noise, the phase between the AM and the PM sidebands varies randomly with time. These conclusions still hold even when the RF noise sidebands are not symmetric about the carrier.     

High spectral purity microwave oscillator: design using conventional air-dielectric cavity

We report exceptionally low PM noise levels from a microwave oscillator that uses a conventional air-dielectric cavity resonator as a frequency discriminator. Our approach is to increase the discriminator's intrinsic signal-to-noise ratio by use of a high-power carrier signal to interrogate an optimally coupled cavity, while the high-level of the carrier is suppressed before the phase detector. We developed and tested an accurate model of the expected PM noise that indicates, among other things, that a conventional air-dielectric resonator of moderate Q will exhibit less discriminator noise in this approach than do more esoteric and expensive dielectric resonators tuned to a high-order, high-Q mode and driven at the dielectric's optimum     

Optimization of an ultra low-phase noise sapphire--SiGe HBT oscillator using nonlinear CAD

In this paper, the electrical and noise performances of a 0.8 /spl mu/m silicon germanium (SiGe) transistor optimized for the design of low phase-noise circuits are described. A nonlinear model developed for the transistor and its use for the design of a low-phase noise C band sapphire resonator oscillator are also reported. The best measured phase noise (at ambient temperature) is -138 dBc/Hz at 1 kHz offset from a 4.85 GHz carrier frequency, with a loaded Q/sub L/ factor of 75,000.     

Merits of PM noise measurement over noise figure: a study at microwave frequencies

This paper primarily addresses the usefulness of phase-modulation (PM) noise measurements versus noise figure (NF) measurements in characterizing the merit of an amplifier. The residual broadband (white PM) noise is used as the basis for estimating the NF of an amplifier. We have observed experimentally that many amplifiers show an increase in the broadband noise of 1 to 5 dB as the signal level through the amplifier increases. This effect is linked to input power through the amplifier's nonlinear intermodulation distortion. Consequently, this effect is reduced as linearity is increased. We further conclude that, although NF is sometimes used as a selection criteria for an amplifier for low-level signal, NF yields no information about potentially important close-to-carrier 1/f noise of an amplifier nor broadband noise in the presence of a high-level signal, but a PM noise measurements does. We also have verified experimentally that the single-sideband PM noise floor of an amplifier due to thermal noise is -177 dBc/Hz, relative to a carrier input power of 0 dBm.     

Microwave regenerative frequency dividers with low phase noise

We demonstrate regenerative divide-by-two (halver) circuits with very low phase modulation (PM) noise at input frequencies of 18.4 GHz and 39.8 GHz. The PM noise of the 18.4 to 9.2 GHz divider pair was L(10 Hz)=-134 dB below the carrier in a 1 Hz bandwidth (dBc/Hz) and L(10 MHz)=-166 dBc/Hz, and the PM noise of the 39.8 GHz to 19.9 GHz divider pair was L(10 Hz)=-122 dBc/Hz and L(10 MHz)=-167 dBc/Hz.     

Phase-sensitive parametric amplifiers in silicon waveguides

The gain and noise of phase-sensitive amplifiers (PSAs) in silicon waveguides based on non-degenerate four-wave mixing (FWM) were investigated. Numerical results show that the PSA in a silicon waveguide offers higher gain and lower amplitude noise with special initial relative phase and suitable pump power. The impacts of nonlinear losses caused by two-photon absorption and free-carrier absorption are also presented by varying free carrier lifetime; short free carrier lifetime should be chosen for optimal performance. Through silicon PSAs, amplitude noise and phase noise of a signal can be suppressed simultaneously.     

Phase noise measurements of 10-MHz BVA quartz crystal resonators

In this paper, we review a new piece of equipment that allows one to characterize the phase noise of crystal resonators using a phase bridge system with carrier suppression. This equipment allows one to measure the inherent phase stability of quartz crystal resonators in a passive circuit without the noise usually associated with an active oscillator. We achieved a system noise floor of approximately -150 dBc/Hz at 1 Hz and -160 dBc/Hz, at 10 Hz. A SPICE characterization of the carrier suppression system is given. An investigation of the phase modulation (PM) noise in 10 MHz BVA, SC-cut quartz crystal resonator pairs is presented.     

A nonlinear dynamic model for performance analysis of large-signal amplifiers in communication systems

A new nonlinear dynamic model of large-signal amplifiers based on a Volterra-like integral series expansion is described. The new Volterra-like series is specially oriented to the modeling of nonlinear communication circuits, since it is expressed in terms of dynamic deviations of the complex modulation envelope of the input signal. The proposed model represents a generalization, to nonlinear systems with memory, of the widely-used amplitude/amplitude (AM/AM) and amplitude/phase (AM/PM) conversion characteristics, which are based on the assumption of a practically memoryless behavior. A measurement procedure for the experimental characterization of the proposed model is also outlined.     

AM noise impact on low level phase noise measurements

The influence of the source AM noise in microwave residual phase noise experiments is investigated. The noise floor degradation problem, caused by the parasitic detection of this type of noise by an imperfectly balanced mixer, is solved thanks to a refinement of the quadrature condition. The parasitic noise contribution attributable to the AM to PM (phase modulation) conversion occurring in the device under test is minimized through the development of a dedicated microwave source featuring an AM noise level as low as -170 dBc/Hz at 10 kHz offset from a 3.5 GHz carrier     

High performance photonic microwave frequency down-conversion using a dual-drive dual-parallel Mach-Zehnder modulator

Photonic microwave frequency down-conversion based on carrier suppression single sideband (CS-SSB) modulation via an integrated dual-drive dual-parallel Mach-Zehnder modulator (DP-MZM) is proposed. The MZM on the up path of the DP-MZM is used to generate SSB modulation signal, while the MZM on the bottom path of the DP-MZM is unmodulated. By adjusting the amplitude and phase of the unmodulated optical carrier, two optical carriers are cancelled out, which improves the performance of the system with reduced local oscillator (LO) power and large suppression of mixing spurious sidebands. The frequency down-conversion approach is theoretically analyzed and verified by simulation. Simulation results show that the power of frequency down-conversion signal is at least 39?dB higher than that of the mixing spurious sidebands. Besides, 9.5?dB gain, 2.8?dB noise figure (NF) and 1.9?dB spurious-free dynamic range (SFDR) improvements can be obtained compared with the previous OCS modulation frequency down-conversion scheme while the required LO power is 10?dB, 5?dB and 5?dB lower than that of the OCS modulation scheme, respectively.     

Extremely low thermal noise floor, high power oscillators usingsurface transverse wave devices

This paper presents state-of-the-art results on 1-GHz surface transverse wave (STW) oscillators running at extremely high loop power levels. The high-Q single-mode STW resonators used in these designs have an insertion loss of 3.6 dB, an unloaded Q of 8000, a residual PM noise of -142 dBc/Hz at a 1-Hz carrier offset, and operate at an incident power of up to +31 dBm in the loop. Other low-Q STW resonators and coupled resonator filters (CRF), with insertion losses in the 5-9 dB range, can conveniently handle power levels in excess of two Watts. These devices were incorporated into voltage controlled oscillators (VCO's) running from a 9.6-V dc source and provide an RF output power of +23 dBm at an RF/dc efficiency of 28%. Their tuning range was 750 kHz and the PM noise floor was -180 dBc/Hz. The oscillators, stabilized with the high-Q devices and using specially designed AB-class power amplifiers, delivered an output power of +29 dBm and exhibited a PM noise floor of -184 dBc/Hz and a 1-Hz phase noise level of -17 dBc/Hz. The 1-Hz phase noise level was improved to -33 dBc/Hz using a commercially available loop amplifier. In this case, the output power was +22 dBm. In all cases studied, the loop amplifier was found to be the factor limiting the close-to-carrier oscillator phase noise performance     

Extension of wavelength-modulation spectroscopy to large modulation depth for diode laser absorption measurements in high-pressure gases

Tunable diode laser absorption measurements at high pressures by use of wavelength-modulation spectroscopy (WMS) require large modulation depths for optimum detection of molecular absorption spectra blended by collisional broadening or dense spacing of the rovibrational transitions. Diode lasers have a large and nonlinear intensity modulation when the wavelength is modulated over a large range by injection-current tuning. In addition to this intensity modulation, other laser performance parameters are measured, including the phase shift between the frequency modulation and the intensity modulation. Following published theory, these parameters are incorporated into an improved model of the WMS signal. The influence of these nonideal laser effects is investigated by means of wavelength-scanned WMS measurements as a function of bath gas pressure on rovibrational transitions of water vapor near 1388 nm. Lock-in detection of the magnitude of the 2f signal is performed to remove the dependence on detection phase. We find good agreement between measurements and the improved model developed for the 2f component of the WMS signal. The effects of the nonideal performance parameters of commercial diode lasers are especially important away from the line center of discrete spectra, and these contributions become more pronounced for 2f signals with the large modulation depths needed for WMS at elevated pressures.     

Simultaneous Determination of Transistor Noise, Gain, and Scattering Parameters for Amplifier Design through Noise Figure Measurements Only

A method for the simultaneous determination of transistor noise and gain parameters through noise figure measurements has been presented recently. An improved version of the method is presented here which can also yield all the scattering parameters needed for designing amplifiers. By means of a proper (computer-aided) data-processing technique, s11, s22, |s12| , |S21|, and s12s21 are determined. As experimental verifications, the characterization of a GaAs MESFET versus frequency (4-8 GHz) is reported.