Modeling the variation of the low-frequency noise in polysiliconemitter bipolar junction transistors

The variation of the low-frequency noise in polysilicon emitter bipolar junction transistors (BJTs) was investigated as a function of emitter area (A_E). For individual BJTs with submicron-sized A _E, the low-frequency noise strongly deviated from a 1/f-dependence. The averaged noise varied as 1/f, with a magnitude proportional to A_E^-1, while the variation in the noise level was found to vary as A_E^-1.5. A new expression that takes into account this deviation is proposed for SPICE modeling of the low-frequency noise. The traps responsible for the noise were located at the thin SiO₂ interface between the polysilicon and monosilicon emitter. The traps' energy level, areal concentration and capture cross-section were estimated to 0.31 eV, 6×10⁸ cm^-2 and 2×10^-19 cm ², respectively     

Measurement and comparison of 1/f noise and g-r noise in siliconhomojunction and III-V heterojunction bipolar transistors

This paper experimentally determines and compares the 1/f noise and the g-r noise, as components of the base noise current spectral density, in Si homojunction and III-V heterojunction bipolar transistors (HBTs) in common-emitter configuration. The noise spectra for each of these devices are obtained as functions of the base bias current (I_B), and the 1/f noise has been found to depend on I_B as I_B^γ, where γ~1.8 for the silicon BJT's and InP/InGaAs HBT's with high current gains (β~50), and γ~1.1 for the AlGaAs/GaAs HBTs with low current gains (4<β<12). The nearly constant current gain and the near square-law and inverse-square emitter area dependence of 1/f noise in silicon devices are indicative of the dominant base bulk recombination nature of this noise. The 1/f noise in the InP based HBTs has been found to be lowest among all the devices we have tested, and its origin is suggested to be the base bulk recombination as in the Si devices. For the AlGaAs/GaAs HBTs, the low current gain and the near unity value of γ, arise most likely due to the combined effects of surface, bulk, and depletion region recombinations and the base-to-emitter injection. The dependence of the 1/f noise on the base current density in the devices tested in this work, and those tested by others are compared to find out which HBT's have achieved the lowest level of 1/f noise     

Source identification and control of 1/f noise in AlGaAs/GaAs HBTsby using an on-ledge Schottky diode

Our experimental results indicate that the 1/f noise spectra density of HBTs is strongly influenced by the emitter ledge potential. By biasing an on-ledge Schottky diode, we can externally control the ledge potential and alter HBT's 1/f noise spectra density and its dependence on the base current. When biasing the on-ledge diode at V_LBE, the HBT's 1/f noise spectra densities are found to increase dramatically (up to 30 dB) from the unbiased value and exhibit stronger dependence on I_B (from I_B^1.42 to I_B^1.84). When biasing the on-ledge diode at V_L>V_BE, the 1/f noise spectra density is found to decrease (about 10 dB) and exhibit virtually no dependence on I_B (from I_B^1.42 to I_B^0.18). These findings help us identify the sources of the 1/f noise and create a novel four-terminal HBT (with an extra ledge electrode) for extremely low 1/f noise RF transceiver applications     

Low-frequency noise in polysilicon emitter bipolar transistors

The low-frequency noise in polysilicon emitter bipolar transistors is investigated. Transistors with various geometries and various properties of the oxide layer at the monosilicon polysilicon interface are studied. The main 1/f noise source proved to be located in the oxide layer. This source causes both 1/f noise in the base current S_Ib and 1/f noise in the emitter series resistance S_re The magnitude of the 1/f noise source depends on the properties of the oxide layer. The 1/f noise is ascribed to barrier height fluctuations of the oxide layer resulting in transparency fluctuations for both minority and majority carriers in the emitter, giving rise to S_Ib and S _re respectively. It is also shown that a low transparency of the oxide layer also reduces the contribution of mobility fluctuations to S_Ib     

Transistor noise in SiGe HBT RF technology

This brief presents experimental and modeling results of device noise in SiGe HBT RF technology. By careful bandgap engineering, a simultaneous reduction of RF noise, 1/f noise, and phase noise has been achieved. At a given I_B, transistors with different base bandgap profiles show similar 1/f noise. At a given I_C, however, transistors with a higher β (and hence lower RF noise) show lower 1/f noise. Circuit analysis and simulation shows that the phase noise is reduced as well     

On 1/f noise characteristics for 0.1 eV HgCdTe photoconductors

Sensitivities on 0.1 eV HgCdTe photoconductors with new electrode configuration and in different sizes were measured at 77K and under 10¹⁴phcm^?2s^?1 photon background conditions. After data for responsivity, generation-recombination noise (g-r noise) and minority carrier lifetime were reproduced by solving a one dimensional diffusion equation on excess minority carrier, discussions on 1/f noise were made and the following characteristics were concluded: (1) 1/f noise does not originate near electrodes for bias current. (2) 1/f noise hardly depends on sensor size and temperature in the 77–95K range, while g-r noise does. (3) 1/f noise is proportional to bias electric field, i.e. current density. (4) 1/f noise does not depend on photon background. From characteristics (2) and (4), it was concluded that 1/f noise has nothing to do with g-r noise. Finally, a new empirical formula was proposed for 1/f noise.     

Low 1/f noise characteristics of AlGaAs/GaAs heterojunction bipolartransistor with electrically abrupt emitter-base junction

It is shown that the use of an electrically abrupt emitter-base junction considerably reduces the 1/f noise of self-aligned AlGaAs/GaAs heterojunction bipolar transistor (HBT). Although this device does not have depleted AlGaAs ledge passivation layer, the low-frequency noise spectra show a very low 1/f noise corner frequency of less than 10 kHz, which is much lower than previously reported value of about 100 kHz from conventional passivated or unpassivated AlGaAs/GaAs HBT's. Except for a residual generation-recombination (g-r) noise component, the noise power is comparable to that of Si BJT. It is also found that the low-frequency noise power of the AlGaAs/GaAs HBT is proportional to the extrinsic GaAs base surface recombination current square. Unlike the other HBT's reported, the noise sources associated with interface state and emitter-base (E-B) space charge region recombination are not significant for our device     

Burst and low-frequency generation-recombination noise in double-heterojunction bipolar transistors

The low-frequency noise in a double-heterojunction bipolar transistor (DHBT) consisted of burst noise and generation-recombination g-r noise. The current dependence of the base burst noise with floating collector was of the form IB³and the current dependence of the collector g-r noise with HF short circuited base was as IC^3/2. The centers involved in the noise generation had an activation energy of about 0.40 eV, with an indication of a second center of lower energy in the collector noise.     

Extrinsic base optimization for high-performance RF SiGeheterojunction bipolar transistors

A comprehensive study on the effect of extrinsic base optimization on the RF performance of an advanced SiGe HBT is presented. An optimized extrinsic poly base with its interface to the epi-base passivated by boron ions is demonstrated to enhance the f_max and the current gain almost two times and to reduce the low-frequency 1/f noise ten times and noise figure (NF) 0.5 dB, achieving f_max of 45 GHz, 1/f noise corner frequency of 700 Hz at I_B=1.0 μA, NF⩽1.0 dB at 900 MHz. Early voltage V_A of ⩾200 V is achieved, while maintaining a BV_CEO of ⩾8.0 V     

Excess noise in AlGaAs/GaAs heterojunction bipolar transistors andassociated TLM test structures

Noise measurements both on transmission line model (TLM) test structures and on associated HBT's are presented. Contact noise is proved to be negligible in the TLM's related to the base structure of transistors. A Hooge parameter for p⁺⁺ doped GaAs is extracted. Activation energies are calculated from results versus temperature. Considering the TLM related to the structure of the emitter, it is shown that the g-r levels observed originate from the AlGaAs layer. Noise measurements on HBT's also exhibit excess noise. A value of the cutoff frequency between the equivalent input current white noise and the 1/f component is given. The base current dependencies associated with different measurement configurations suggest the 1/f noise to come from the base or the emitter-base junction. The g-r components are studied as a function of temperature. Activation energies are deduced. Finally a comparison of the TLM and HBT noise results is presented. The presence of the complex DX center and of g-r levels in the base region are proposed as possible origins for the g-r noise in HBT's     

低偏置电流下大功率半导体激光器低频电噪声特性

测量了大功率InGaAsP/GaAs量子阱半导体激光器在五十分之一阈值电流下的电压低频噪声功率谱密度.实验结果显示,激光器的低频电噪声呈现1/f噪声,在不同的偏置电流范围内,1/f噪声幅度随电流的变化关系不同,整体上随偏置电流的增大而减小,实验中并未发现g-r噪声.结合低偏置电流时激光器动态电阻的大小,给出了1/f噪声的模型,分析了在低偏置电流下的1/f噪声主要来自有源区和漏电电阻,其幅度的大小及其随偏置电流的变化趋势与激光器的可靠性有密切的关系.     

The effects of geometrical scaling on the frequency response andnoise performance of SiGe HBTs

We examine the geometrical scaling issues in SiGe HBT technology. Width Scaling, length scaling, and stripe-number scaling are quantified from a radio frequency (RF) design perspective at 2 GHz. We conclude that a SiGe HBT with emitter area A_E=0.5×20×6 μm² is optimum for low noise applications at J_c=0.1 mA/μm² and f=2 GHz using the design methodology, which guarantees optimal noise and input impedance matching with the simplest matching network. Finally, the optimal device sizes at f=4 and 6 GHz for low noise applications are also obtained using the same method     

Low-frequency noise characteristics of UHV/CVD epitaxial Si- andSiGe-base bipolar transistors

We report the first measurements of low-frequency noise in high-performance, UHV/CVD epitaxial Si- and SiGe-base bipolar transistors. The magnitude of the noise power spectral density at fixed frequency for both Si and SiGe devices is comparable for similar bias, geometry, and doping conditions, indicating that the use of strained SiGe alloys does not degrade transistor noise performance. The best recorded values of noise corner frequency were 480 Hz and 373 Hz for the Si and SiGe transistors, respectively, for multi-stripe devices with an emitter area of 0.5×10.0×3 μm². A functional dependence of the noise power spectral density on base current for both device types of I_B^1.90 was observed, and noise measurements as a function of device geometry suggest that the contributing noise sources are uniformly distributed across the emitter of the transistors, not at the emitter periphery     

1/f noise characterization of base current and emitter interfacialoxide breakup in n-p-n polyemitter bipolar transistors

We propose using base 1/f noise to characterize the distribution of diffusion and tunneling components of base-current (I_b) at the emitter poly/monosilicon interface in n-p-n polyemitter transistors. A noise model is constructed to interpret the I_b 1/f noise (S _iEB) dependence on these combined currents. Measured I_b dependences of S_iEB increase progressively from I_b^1.2 to I_b^2.0 for transistors having emitter structures concomitant with increasing current gains and series emitter resistances ranging between 115-1800 and 7-33Ω, respectively. This is indicative of tunneling components in I_b^2.0 that increase with higher interfacial oxide continuity, and persist in epitaxially realigned emitters     

Dead space effect on the wavelength dependence of gain and noise inavalanche photodiodes

We extend the dead space model proposed by Hayat et al. in order to determine the wavelength-dependent multiplication mean gain 〈G(λ)〉 and excess noise factor F(λ) in the case of mixed electron and hole injection, as it is the case when photons are absorbed within the multiplication region. We compare the predictions of the model with measurements performed on a silicon ultraviolet-selective avalanche photodiode with submicron thick multiplication region. We show that the multiplication gain is constant in the visible and near-infrared part of the spectrum, and increases in the UV range by a factor of 1.8. Furthermore, the excess noise factor is minimal for UV radiation and increases rapidly for longer wavelengths. It appears that the extended dead space model is very adequate at predicting the gain and noise measurement results. In order to unambiguously determine the effect of the dead space, we compare the predictions of our model with those of McIntyre's local noise model. The latter qualitatively describes the wavelength dependence of the gain, but greatly overestimates the excess noise factor     

GaAs FET's with a flicker-noise corner below 1 MHz

GaAs MESFET's with significantly reduced low-frequency noise are demonstrated through application of an understanding that the dominant noise source is generation-recombination (g-r) noise from deep level traps in the gate and backside depletion layers. A 1/f noise spectrum measured from 100 Hz to 10 MHz is modeled as the sum of Lorentzian noise spectra from a few traps subject to the temperature distribution inherent in a GaAs MESFET. The noise associated with a single mid-bandgap trapping level does not appear as an ideal Lorentzian, but rather as 1/f over nearly a decade frequency range by virtue of a time constant that is a strong function of temperature ( exp[Ea/ kT]) and an estimated temperature distribution of 22°C across the active region. The major g-r trap was characterized as having an activation energy of 0.75 eV. By reducing the g-r noise, flicker noise was decreased by more than 15 dB compared to our conventional GaAs MESFET's and the noise corner was reduced to less than 1 MHz from a typical 40 MHz. This significant reduction was achieved by using MBE layers designed to have lower trap concentrations and high channel doping. These results are within 10 dB of the estimated 1/f noise limit due to the quantum mechanics of carrier scattering [5], [14].     

Low-frequency noise of InP/InGaAs heterojunction bipolartransistors

The equivalent base noise SI_b of InP/InGaAs heterojunction bipolar transistors (HBT's) with a circular pattern emitter is investigated experimentally at a low frequency ranging from 10-10⁵ Hz. The measured SI_b exhibits the 1/f dependence in an overall frequency range without any accompanying burst noise. Furthermore, SI_b varies as I_b^γ for the base current I_b and as d^-2 for the emitter diameter d, where the value of γ ranges from 1.62-1.72 depending on d of HBT's used. The 1/f noise model, which rigorously deals with the recombination current at the base surface I_bs as a function of I_b as well as of d is proposed. Applying our noise model to the dependence of SI_b on I_b, as well as on d, reveals that even though γ is less than two, the origin of SI_b is due to the recombination of electrons at the exposed base surface near the emitter edges. On the basis of theoretical considerations for the diffusion length of electrons and traps at the base surface, the Hooge parameter α_H for the noise due to the base surface recombination is deduced to be in the order of 10 ^-2 for the first time     

Low-frequency noise in TFSOI lateral n-p-n bipolar transistors

Low-frequency (1/f) noise is characterized as a function of base current density (J_B) on thin-film-silicon-on-insulator (TFSOI) lateral bipolar transistors. In the low injection region of operation, the noise power spectral density was proportional to J_B ^1.8 for J_B<0.4 μA/μm², which suggest that the noise in these devices is primarily dominated by a uniform distribution of noise sources across the emitter-base area. However in the high current region of operation (J_B>0.4 μm²), the noise bias dependence shifts to J_B ^1.2, indicating current crowding effects, alter the contribution of noise sources near the extrinsic base link region of the device. In addition to the expected 1/f noise and shot noise, we have observed a bias dependent generation-recombination (Gm) noise source in some of the devices. This G/R noise is correlated to random-telegraph-signal (RTS) noise resulting from single trapping centers, located at or near the spacer oxide and/or the Si to SIMOX interface, which modulate the emitter-base space charge region     

Effect of Schottky barrier alteration on the low-frequency noise ofInP-based HEMTs

For the first time the effect of increasing the Schottky barrier's Al content of InP-based InAlAs-InGaAs HEMTs from 48 to 60% on the low-frequency (LF) drain and gate current noise is investigated. It is shown that the LF gate current noise S_IG(f) for the 60% case decreases by almost three decades, while the LF drain current noise S _IDS(f) stays at the same level. From small coherence values, it can be concluded that drain and gate noise sources can be treated separately which facilitates the LF noise modeling of these HEMTs     

Numerical modeling of fluctuation phenomena in semiconductors and detailed noise study of mid-wave infrared HgCdTe-heterostructure devices

The aim of this paper is to present an effective numerical model of fluctuation phenomena in semiconductor structures with an arbitrarily defined doping profile and variable-band structure. The model enables the spectral intensity of the noise current to be calculated. It is known that the 1/f noise may result in fluctuations of the carrier mobility. It is not clear, however, why strong 1/f noise is observed in reverse-biased HgCdTe nonequilibrium photodiodes when saturation currents are usually very low. In the present paper, we try to answer this question. Although the nonequilibrium mode of operation leads to the reduction of the generation-recombination (g-r) noise, it increases the electric field as well as the band mobility of carriers and its fluctuations. The observed low-frequency noise is due to the fluctuations of current density caused by mobility fluctuations assisted by the electric field.