24September2017

Materials and Electronics Engineering

Review on Signal Detection of Low DC Current in Nanodevices and Various Sensors

Yantao Liang, Jiapin Chen, Franklin Li Duan, Dong Xu, Jun Wang, Hanling Yang, and Yafei Zhang*

Abstract
icon-htmlFull Text Html
icon-pdf-smPDF w/ Links
icon-citExport Citation
Figures

Key Laboratory for Thin Film and Microfabrication Technology of the Ministry of Education, Research Institute of Micro/Nanometer Science and Technology
Shanghai Jiao Tong University, Shanghai, 200240, China

Materials and Electronics Engineering 2014,1:1

Publication Date (Web): December 10, 2010 (Review)

DOI:10.11605/mee-1-1

*Corresponding author. E-mail:  This email address is being protected from spambots. You need JavaScript enabled to view it.

 

Abstract

 


Block diagram of pico-ampere parallel test system. The system consists of a V93K test head, pico-ampere cards, DC parametric cards (VI32), digital cards, pogo cables, pogo tower, wafer prober interface board (WPI BOARD) and a probe card (POGO TOWER).
        As the operating current of various new nanoelectronic devices and signals of various sensors are scaled down to nano-ampere range, technologies for ultralow current detection thus become critical for the development of nanodevices and related fields such as sensors. Low current characterization is roughly divided into two categories: direct methods and indirect methods. The direct approach uses the high precision current meter / equipments to detect the small signals which may reach the resolution of 10-15 A. This sort of equipment is usually expensive and not quite suitable for general large scale application. The indirect method makes use of various electronic circuitry and theory to measure the ultralow electric current and can reach the resolution up to 10-14A. In this paper, we mainly review and compare a few popular indirect methods, including: sampling resistors, feedback operational amplifiers, biochip method based on MOSFET operating in sub-threshold region, I-F transformation, chaos theory and parallel test system with pico-ampere measurement capabilities. Pros and cons of direct approaches are also compared both for direct and indirect categories.

 

Keywords

Nanoampere current detection, T network feedback, Biochip method, Parallel test system, nanodevices.

Full Text Html

Introduction 

      Weak signal by definition has two meanings: (1) the amplitude of the useful signal is very weak compared with the ambient noise. For example, when the SNR (Signal to Noise Ratio) of input signal is at the level of 10-1, 10-2, or as low as 10-4, then the useful signal is submerged in the noise and becomes too low to be detected; (2) the absolute value of the useful signal’s amplitude is extremely small, example of such is like low electric current of 10-9 A, 10-12 A, even 10-15 A. The goal of signal detection is then to suppress the noise to enhance the SNR so as to extract the useful signal from the noise. In generally, the way to do this is to convert the low current signals into voltage by corresponding sensors and then amplify the voltage to a measurable range. In this process, the noise (background noise of the sensor, the inherent noise of the measuring instrument, and the external interference, etc) is also amplified at the same time. Various methods then are developed based upon the study of the origin and physics of these noise sources in order to suppress these noises using various electronics theory and computer skills [1-7].

 1.1 History of small signal measurements 

      Signal detection technology was started as early as 1928, when Johnson found that the noise is generated by thermal agitation motion of electrons. Then, the discovery and development of transistors and integrated circuit (IC) have significantly boosted the technologies for signal detection. By using DC amplifiers made of transistors and ICs, the technology for ultralow current detection has comes into a new era. The first chopper modulated DC amplifier was invented in 1957, allowing the circuitry to achieve a sensitivity of 10-9 A for short-term measurement. The sensitivity was later improved to a level of 10-15 A by McCaslin who used low leakage current MOSFETs circuitry [8]

        There is a rapid progress in past decades for noise suppression to enhance the SNR since 1962. U.S. PARC developed the first coherent detection lock-in amplifier which increased the SNR to 103 in 1962. In 1968, Auger electrons were measured from the background of large numbers of secondary electrons. In the early 80’s, the signal which less than 1nV was allowed to obtain full-scale output and the amplification of the signal is close to 200 dB while SNR increased to 106. SNR got a rapid improvement [9] with the advent of VLSI (Very Large Integrated Circuits) and ASIC (application specific integrated circuits). By rough estimates, the measurement limit will improve an order of magnitude every 5 to 6 years. So, it has become possible to detect the weak electronic current which was regarded as immeasurable. 

1.2 Significance of small signal detection 

Weak signal detection or nano ampere measurement is important in both scientific research and engineering practice. Low current signal detection is found in various scientific fields such as biomedicine, chemistry, electrochemistry, astronomy, physics and magnetism, and to discover new natural laws. Weak currents need to be characterized accurately in the application of power-line monitoring equipment, photoelectric transformer, semiconductor device leakage, ionization rate of air in the upper atmosphere, beam and radiation monitoring, leakage current test of insulators, switch, relay and other electronic components; SEM beam current measurement; analysis of nanometer electronic devices’ characteristics; the detection of electrostatic, photocurrent and biological electronic current. 

There are also many cases we need to characterize low DC current for various high-techs and military reconnaissance. With the development of semiconductor technology, the era for nano semiconductor device has arrived. According to famous Moore's law, in the next few years integration density and computing power will continue increasing and cost per bit will continue decreasing. The size of nanoscale MOS device continues shrinking and this greatly enhanced the performance of integrated circuits as well. The negative effects of MOSFET shrinking are increasing difficulty to completely shut down the FET and significant increase in power consumption due to various leakage current. These negative effects grow dramatically with improved integration and the shrinking of microelectronic devices on the chip, therefore new nano-materials and new nano-electronic solid-state devices are urgently needed to replace the existing Si [10]. Nano-devices, due to its critical dimension close to the nanometer range (cross-sectional dimension of unipolar device channel, barrier/ potential well layer thickness of bipolar junction scale tunneling devices, nano-scale island-to-island devices, etc.), various quantum tunneling and quantum transport, the device on/off current fall into the low current range (10-9 A). For example, the drive current of a single walled CNTFET is around 10-8 A [11]. In military use, the total amount of ionization produced by the detector is directly related to the accumulated electronic charge in the output circuit, and the electronic current in this circuitry is very low (in the range of 10-8 A ~ 10-14 A). In health care, the electric current in the range of (10-6 A) is generally used to stimulate cellular energy to promote healing. 

Nano ampere current detection is an important part of STM (Scanning Tunneling Microscope) and nanoscale devices. In tunneling state, there is a negative exponential function between the tunnel current and the tunnel gap. The gap between the probe and sample is about 10 angstroms. And the tunnel current will change one order of magnitude if the gap changes 1 angstrom. The tunnel current signal is very weak, usually at the level of 10-9 A. We can calculate the corresponding changes in extremely weak gap by detecting the tunnel current. 

Entering the 21st century, the DC electric current of various nanodevices and sensors has also entered nano ampere range (<10-9 A). At the same time, the noise performance of nanodevices are also deteriorated by heat dissipation, parasitic capacitance and surface effects in the nanoscale geometry. These contribute the new challenges to the ultralow current detection. Developing new technologies for low current detection are crucial for nanodevices and sensors in related fields. The nano ampere low current characterization have been evolving into a new technology in the field of low signal detection. 

1.3 Methods for low current detection 

Nanoampere DC current detection techniques are roughly divided into two categories: direct and indirect. 

1.3.1 Direct approach

The direct approach is just to use high precision equipment to measure nano ampere DC current. Table 1 briefly compares the features and resolution of various equipments. 

Agilent Technologies is the global leader, the world's largest electronic test and measurement company. Its Agilent 4080 supports the measurement of ultralow electronic current up to a resolution of 10-15 A. Agilent B2900A series precision source or measure unit’s can reach resolution of 10-14 A. Agilent 3458A digital multimeter can reach the highest sensitivity up to 10-12 A, and the highest levels of precision can reach up to 8.5 digits of measurement resolution and 0.1 part per million transfer accuracy. 

Table 1. Comparison of direct method for low electric current detection

     Keithley Instruments, Inc. establishes a core technology based on very weak electronic signals measurement, and applied this technique to test the rapidly developmental industrial products. For example, Keithley 6485/E picoampere current meter’s measuring range is 2*10-14 A~2*10-2 A, and the limit resolution can reach 10-14 A. Every instrument in Keithley’s Series 2400 SourceMeter family gives us the capability to measure nanoampere current in one compact unit. Each one combines a programmable power source with a highly repeatable, 51/2-digit multimeter (DMM) in a half-rack-sized enclosure. Its current measuring range is 10-11 A~10 A; the precision is 0.035%, such as Keithley 2410, and Keithley 2400-C. 

FLUCK also offers high-quality electronic instrument for testing and fault detection to various industrial fields. For instance, FLUCK 8808A, a 51/2-digit multimeter, can measure weak electronic current at a limit resolution of 10-9 A without loading the circuit being measured, which is the only micro electronic current measurement instrument with low impedance I-V conversion technology. 

In domestic markets, designed and produced by MEACON according to actual demands of industrial field, the MK-1000 micro electronic current detector is mainly used to measure static electricity, photocurrent, biological electronic current, and can accurately measure the nanoampere current. The GT8232 nanoampere current meter’s limit resolution can reach 10-11 A, which is made by Xi'an Feiteng Instrument Company Limited. The 6000-02 DC current meter, designed by Beijing North Siyuan Electronics Technology Centre, can accurately measure DC nanoampere current signal at a limit resolution of 10-12 A. 

1.3.2 Indirect approach 

The above equipment is usually expensive and not suitable for applications at a large scale. One can use the indirect method to circumvent these difficulties. In this paper, various indirect methods detect nanoampere current or even lower are reviewed. Table 2 briefly compares the advantages, disadvantages and detection limit of these indirect methods. 

Table 2. Comparison of indirect methods for low DC currents mesurements

     These indirect methods are: sampling resistors, the feedback operational amplifiers, biochip method based on MOSFET operating in sub-threshold region, I-F transformation, chaos theory and a parallel test system. According to different characteristics of the signal and the noise, we should take corresponding approaches. In the following chapters we are going to discuss these indirect methods in details. 

Review of indirect methods for low current detections 

2.1 Sampling Resistor (SR) Method 

The principle of sampling resistor method is converting the current to voltage signal by inserting sampling resistor in the loop. Figure 1 shows the schematic, output voltage Vo is:

Vo=Vs*(R1+R2)/R2=Is*Rs*(R1+R2)/R2           (1)

As shown in equation (1), the sensitivity of output voltage is proportional to the sampling resistance. Because nanoampere current is very weak, the sampling resistance is relatively large in order to achieve high resolution and sensitivity. For example, the sampling resistance should normally choose mega ohm level or even higher for nanoampere current measurement. 

But the sampling resistance cannot infinitely increase because the thermal noise and distributed capacitance of the resistor are proportional to the resistance, drift and noise increase with resistance rising. Therefore, the amplifier’s sensitivity cannot continue to be improved with the increasing of resistance when the sampling resistance is large enough. Meanwhile, the instrument is susceptible to noise, zero drift and other factors working at the highest sensitivity [12]. As a result, the smaller the resistance, the better the temperature stability; the smaller the time constant of input circuit, the faster the response speed of the instrument. According to all reason above, the sampling resistance should be as small as possible in order to ensuring the accuracy.

Figure 1 Schematic of sampling resistance method

Figure 1 Schematic of sampling resistance method

 2.2 Feedback Operational Amplifiers Method 

The feedback operational amplifiers method is to realize I-V conversion by using operational amplifier with high input impedance and introducing feedback network, converting ultralow electronic current into measurable voltage. This method has some advantages such as simple circuit structure, a variety of forms and wide scope of application. The following mainly introduces the operational amplifier with single-resistor feedback (OPA SRF) and the operational amplifier with T-shaped feedback network (OPA TFN). And other types of operational amplifier feedback methods can be obtained by using these two forms appropriately. 

2.2.1 Operational Amplifier With Single-resistor Feedback (OPA SRF) 

Schematic is shown in Figure 2. Is is the electronic current to be measured, Rf is feedback resistance. Ideally, the OPA has infinite differential mode voltage gain, infinite input resistance, zero output resistance, and zero input bias current. So the bias current Ib of OPA has no effect on Is, the electronic current If flowing through the feedback resistor is approximately equal to Is. According to the principle of virtual short circuit and virtual open circuit, the output voltage Vo of OPA can be given as follow:

Vo= -Is*Rf                               (2)

        So we can calculate the electronic current Is by measuring the output voltage Vo. Even if the electronic current to be measured is very small, we can obtain observable output voltage as long as the feedback resistance is large enough.

Figure 2 Schematic of operational amplifier with single-resistor feedback

Figure 2 Schematic of operational amplifier with single-resistor feedback


     But in fact, the input impedance of OPA is not infinite, and the increase of resistance Rf is limited by the input impedance of OPA. Considering the impact of the bias current Ib on the Is, output voltage Vo can be given as follow:

Vo= -(Is-Ib)*Rf                           (3)

As shown in equation (3), on the one hand, the electronic current signal Is will be submerged in noise, and cannot be measured if Is is less than Ib, in other words, the electronic current to be measured is less than the input bias current of OPA. On the other hand, the measuring accuracy is higher if Is is far greater than Ib. 

However, the maximum electronic current can be measured by this method is limited by the maximum output current of OPA, and the minimum electronic current can be measured by this method is limited by the input bias current of OPA [13]. As shown in equation (3), in this method, the primary factor influencing the measurement sensitivity is the input bias current Ib of OPA, and the secondary factors are the zero drift and noise voltage. The noise coefficient of OPA has a decisive effect on noise characteristics of the whole detection circuit. Therefore, in order to improve measurement sensitivity and accuracy, the selected OPA should meet the following conditions: (1) High input resistance Ri; (2) Low input bias current Ib; (3) Low offset voltage, low drift, low noise. (4) High gain and common mode rejection ratio (CMRR) [14, 15]

For example, the current detection circuit which is used in FAIMS system, designed by Yan, has pre-amplification stage and post-stage amplification stage. Both of the stages use the operational amplifier with single-resistor feedback to amplify the input signal. And the high frequency interference from system is reduced by filter circuit and shielding circuit. Additional, LabView is used to process the measured current signals, and the measurement accuracy is better than 0.1 pA [16]

Disadvantages of the operational amplifier with single-resistor feedback method are: useful signal and noise are amplified simultaneously, and the output noise power also increases, so there is no effective improvement in the SNR [17]

2.2.2 Operational Amplifier With T-shaped Feedback Network (OPA TFN) 

However with the increasing of resistance, the thermal noise will increase, the stability and accuracy of resistor will decrease in the operational amplifier with single-resistor feedback method [18, 19]. Meanwhile, parasitic capacitance will increase, which is caused by high impedance and associated with the feedback resistance, resulting in increasing of the response time and affecting the response speed of the circuit. Therefore, the feedback resistance is not infinite, limiting the measurement of nanoampere DC current or even lower. The following introduces an operational amplifier with T-shaped feedback network. Schematic is shown in Figure 3, resistors R1, R2, R3 form a T-shaped resistor network for further amplification of nanoampere current, and the amplification factor is (1+R3/R2)*Rl. The resistors R1, R2, R3 provide high sensitivity for the whole circuit. The feedback compensation network consists of R1 and C1, R3 and C2 respectively, providing loop compensation and noise suppression, meanwhile, reducing the bandwidth and preventing the self excited oscillation [20, 21].

According to the principle of virtual short circuit and virtual open circuit, the output voltage Vo of OPA can be given as follow:

Vo=-Is*(R1+R3+R1*R3/R2)                 (4)

So we can calculate the electronic current Is by measuring the output voltage Vo. 

Figure 3 Schematic of operational amplifier with T-shaped feedback network

Figure 3 Schematic of operational amplifier with T-shaped feedback network


     Advantages of operational amplifier with T-shaped feedback network are as follows. First, using T-type feedback network, we can reduce feedback resistor without losing the impedance of OPA [22, 23], and improve the sensitivity and accuracy of the circuit [24]. Second, the feedback compensation network can provide loop compensation and noise suppression, and reduce the bandwidth and preventing self excited oscillation. However, the current can be effectively amplified depends on the overall performance of the OPA. This method has high detection accuracy for nanoampere current, but ultra weak current signal such as pico-ampere current cannot be accurately amplification generally. 

The operational amplifier with T-shaped feedback network method is widely used in the detection of nanoampere DC current flowing in nanoscale devices, because its circuit structure is simple and easy to build, and this method also has high sensitivity and accuracy. Using operational amplifier with T-shaped feedback network method, Huang designed an I-V converter, as shown in Figure 4 [24]. In order to achieve accurate measurement, OPA 128LM has been selected which has high input impedance, ultralow input bias current (±40 fA), and ultralow noise. The electronic current to be measured inputs from the Input terminal, is converted to the voltage Vo via I-V converter, ranging from -5V~0V, then is processed by follow-up detection device. Finally, accomplish the measurement of the nanoampere current.

Figure 4 Operational amplifier with T-shaped feedback network detection circuit 1. R1, R2, R3 form a T-shaped resistor network for further amplification of nanoampere current. The resistors R1, R2, R3 provide high sensitivity for the whole circuit.

Figure 4 Operational amplifier with T-shaped feedback network detection circuit 1. R1, R2, R3 form a T-shaped resistor network for further amplification of nanoampere current. The resistors R1, R2, R3 provide high sensitivity for the whole circuit.


     Yan Fuxing achieved the detection accuracy of 1 pA, using the operational amplifier with T-shaped feedback network method, by selecting a low noise resistor, capacitance, ultralow leakage current, high insulation boards to reduce the noise and by routing reasonable to reduce interference [25]

In additional, Wang achieved the detection accuracy of 0.1 pA, using above method, controlling by the follow-up AD converter and micro control unit (MCU) [26]. As another example, Wei accomplished the detection of ultralow electronic current as low as 10-14 ampere, using I-V converter with T-shaped feedback network as sampling circuit, selecting high stability input impedance OPA and feedback resistor, combining with filtering technology, as shown in Figure 5 [27].

Figure 5 Operational amplifier with T-shaped feedback network detection circuit 2. This circuit includes four stages circuit. The first stage is I-V conversion circuit with T-shaped feedback network. The second stage is voltage amplification circuit which is used to amplify the voltage signal output from first stage. The third stage and the last stage are lowpass filter and band rejection filter respectively, which are used for filtering the high-frequency noise and the low-frequency noise, enhancing the precision of this circuit.

Figure 5 Operational amplifier with T-shaped feedback network detection circuit 2. This circuit includes four stages circuit. The first stage is I-V conversion circuit with T-shaped feedback network. The second stage is voltage amplification circuit which is used to amplify the voltage signal output from first stage. The third stage and the last stage are lowpass filter and band rejection filter respectively, which are used for filtering the high-frequency noise and the low-frequency noise, enhancing the precision of this circuit.

2.3 Biochip Method Based On MOSFET Operating In Sub-threshold Region 

  Biosensors has entered the nanoscale or even smaller because of the booming in biosensors research and production, it is more difficult to detect the ultralow electronic current. So the primary task is amplifying the ultralow current to ideal range using ultralow current amplifier, increasing the demand for ultralow current amplifier, such as fields of low current analysis and measurement, clinical diagnosis and DNA probes. The essential requirements of the bio-signals amplifier are: (1) low noise; (2) high stability; (3) high gain accuracy. 

        It is well known that the sub-threshold current appears an exponential function of the gate voltage in MOSFET, sub-threshold circuits not only suffer from power fluctuations and process fluctuations between die to die seriously, but also changes significantly with temperature. Fortunately, it suffers little variations for transistors within the same die (assuming constant temperature distribution), which makes it possible to realize ultralow current detection circuits by integrating all the circuit modules on a single chip [28, 29]. In addition, compared to the voltage mode circuits, current mode circuits have the advantages of wide bandwidth, high dynamic range, fast speed, high accuracy, etc. Therefore, integrating current sensing amplifier and current generator of bio-sensing systems on a single biochip can eliminate the interference causing by process and temperature, increase the detection range, and improve the detection accuracy and detection speed. 

        Further, considering the noise problem, 1/f noise is the dominant one because of low frequency operation in the sub-threshold region. And the 1/f noise is mainly caused by the surface effects of the CMOS transistors. In sub-threshold region, carrier transport is mainly due to diffusion rather than drift as in the strong inversion regions. So, the carriers are mainly flows in the bulk region rather than the surface, thus reducing the 1/f noise. Meanwhile, 1/f noise in PMOS transistors is dominated by the carrier density fluctuation, 1/f noise in NMOS transistors is mainly due to mobility fluctuation. So, 1/f noise in PMOS is much less than that of NMOS [30, 31]. What’s more, traditional amplifiers need to preestablish DC bias, thus it will produce background noise, and cause interference on the detection of nanoampere current. 

  In conclusion, noise in biochip system can be greatly reduced by establishing DC bias direct using low input current, making all the transistors operate in sub-threshold region. 

        Zhang designed an ultralow current amplifier operating in sub-threshold region. The stability of the output current and precision of gain were improved by incorporating OPA and feedback loop [31]. The 1/f noise was reduced by choosing PMOS transistors operating in the sub-threshold region, and further reduced by increasing the size of dominant transistors and the application of the drain shifting technique. The electronic current can be amplified to processible range with this ultralow current amplifier, and then input to next circuit for testing. 

      Generally, the amplitude of electronic current to be measured is extremely small, the gain of single-stage OPA is finite. Since the electronic current cannot be amplified to desired range by single-stage OPA, two-stage or multi-stage amplifiers should be taken into consideration. Zhang designed a multi-stage current amplifier to achieve the detection for ultralow current which integrated all circuit modules on one chip [32]. Topology of this multi-stage amplifier is shown in Figure 6.

Figure 6 Topology of multi-stage amplifier.

Figure 6 Topology of multi-stage amplifier.

     In this circuit, the ultralow current (generally at the level of pA) was amplified and converted to a differential voltage signal by a differential trans-resistance amplifier composed of OPA0, OPA1, M0, and M1, and then further amplified by a trans-conductance amplifier composed of OTA0, OTA1, and OPA2 with differential input and signal output. Finally, it output as a current signal Io and flowed to the following stages. In order to reduce noise, all MOSFET transistors in this circuit were operating in the sub-threshold region, and mismatch and process variations were reduced through reasonable layout. In addition to the supply voltage, all the voltages were generated by the on-chip circuits, which avoided the extra noise and process variations. With only 1nA of DC quiescent current, this multi-stage amplifier could stably provide current gain of 57.6dB and 3dB bandwidth of 4.38MHz with only 22.8 fA of offset current and 66.9 fA/Hz1/2 of noise current [32].

Figure 7 Saw tooth oscillator for on chip measurement of sub-picoamp current.

Figure 7 Saw tooth oscillator for on chip measurement of sub-picoamp current.

     Another example, B. Linares-Barranco integrated the sawtooth oscillator, driving by current, on the chip to detect ultra-low current of fA, as shown in Figure 7 [20]. Ensuring the operational current of oscillator as small as the reverse current of the diffusion transistors, we can reasonably estimate for injecting current. Meanwhile, the capacitor voltage Vc at C1 was monitored by a high-speed analog buffer. The discharge slope K can be obtained, which is proportional to the discharge current Iin. As shown in equation (5):

K=△Vc/△t=-Iin/C1                      (5)

By comparing the measured discharge slope K with a known reference discharge slope at known current, we can deduce the current Iin

The above all kinds of biochip methods primely restrain the noise of the device and the noise interference from outside, improving the detection precision to fA. The DC current in nanoscale devices can be accurately measured. 

2.4 I-F Transformation 

        The basic principle of current to frequency transformation is: firstly, the input current is converted to DC voltage by I-V converter. Then the DC voltage is converted to frequency signal which is in proportion to the input current. Processing by scaler and computer, finally we can deduce the input current from frequency signal [33, 34]. I-F converter based on weak current detection can be divided into two types: feedback current amplification I-F transformation and feedback current integral I-F transformation. 

2.4.1 Feedback Current Amplification I-F Transformation 

        Block diagram of feedback current amplification I-F transformation is shown in Figure 8, I-V converter is the core of the whole detection structure, generally consists of OPA with feedback resistor. Since the nanoampere current is very weak, the output DC voltage of I-V converter circuit is finite, generally at the level of mV. It must be amplified by following voltage amplifier so that the output voltage can reach the desired magnitude. Moreover, the polarity of input nanoampere current signal may be positive or negative, so it needs to be processed by polarity selection circuit in order to ensure the polarity of output DC voltage is positive. And additional voltage limiter circuit is needed if subsequent circuits have requirements on output voltage amplitude [35].

Figure 8 Block diagram of feedback current amplification I-F transformation.

Figure 8 Block diagram of feedback current amplification I-F transformation.

 The work processes of feedback current amplification I-F converter are: the input current Is is converted to DC voltage which is in proportion to input current by I-V converter circuit. The DC voltage is amplified by voltage amplification module, and then will be converted to frequency signal which is in proportion to the amplified voltage using V-F converter circuit. Finally, the pulse frequency is recorded by counter or processed by computer. We can deduce the input current Is by measuring the frequency in proportion to it. But the stability of feedback current amplification I-F transformation is poor in nanoampere current detection, and this method is susceptible to noise and interfere from other external factors. 

2.4.2 Feedback Current Integral I-F Transformation 

      Block diagram of feedback current integral I-F transformation is shown in Figure 9 [36]. This circuit consists of integrator, amplitude discriminator, monostabillity circuit, pulse current source and subsequent frequency measurement circuit. Generally, Miller integrator is constituted by connecting high and stable integral capacitor to OPA feedback loop [37]. The capacitor is charged by input current, and the current is converted to voltage signal like sawtooth waveform, and then the voltage signal will be converted to frequency signal. Ultimately, the input current Is can be deduced from the frequency measured by frequency measurement circuit or scaler.

Figure 9 Block diagram of feedback current integral I-F transformation.

Figure 9 Block diagram of feedback current integral I-F transformation.

      Work processes are: the output voltage V1 of integrator gradually increases with the loading input current Is. The output voltage V2 of amplitude discriminator will change when V1 reaches the threshold voltage of amplitude discriminator. Then monostabillity circuit will be triggered and will generate the pulse voltage V3 which will be converted to the pulse current If by pulse current source, the If will be fed back to the integrator, and V1 will be decreased. Thereby the amplitude discriminator will return to its original state.But V1 will rise because of the presence of the input current Is, this circuit will work in repeated cycle like this, generating a series of periodic pulse by monostabillity circuit [36]. The larger the input current, the faster the capacitor integral and the higher the pulse frequency. On the contrary, the pulse frequency will decrease [34]. Meanwhile, the periodic pulse frequency will be recorded by frequency measurement circuit or processed by computer, which is in proportion to input current Is, then we can deduce the input current Is. 

        Owing to the use of integrating capacitor, temperature coefficient of feedback current integral I-F transformation is lower than that of high resistance. And the stability of feedback current integral I-F transformation is better than that of feedback current amplification I-F transformation. In addition, the integrator’s charge is the time integral of ultralow current, thus it has good inhibition of noise and interference from outside [37]

        Because the pulse frequency is in proportion to the input current when the integral capacitance and reference voltage are fixed. For example, an input current at the range of uA can be linearly converted to the frequency below 1GHz by an operational current to frequency converter designed by R. Yadav. At the same time, the range of current detection and corresponding frequency can be changed by modulating the integral capacitance [38]. In order to measure the current varying from 2.5 pA to 1 mA, Vigano selected the current to frequency converter (CFC) based on the balanced charge integrating techniques. Comparing with other switching techniques, this converter has the advantage that it is without dead times and with no loss of charges. Since the output frequency depends on the input current, additional AD converter is added to measure the output voltage of the integrator and to calculate partial counts in the threshold comparator card. This method decreases the response time, increases the dynamic range. Finally, through a lot of tests and continuous improvement, input current varying from 10-11A to 10-3A was linearly converted to the frequency varying from 10-1Hz to 107Hz, as shown in Figure 10 [39]. Moreover, I-F converter achieved conversion accuracy of less than five over one hundred thousand, based on ADuC847 MCU, designed by Yu Lei [40]. Smart multi-channel data acquisition system also realized the measurement of the weak electronic current, based on I-F transformation, designed by Wang Hao [41].

Figure 10 Relation of current and frequency. Input current varying from 10-11A to 10-3A was linearly converted to the frequency varying from 10-1Hz to 107Hz.

Figure 10 Relation of current and frequency. Input current varying from 10-11A to 10-3A was linearly converted to the frequency varying from 10-1Hz to 107Hz.

2.5 Chaos Theory 

The chaotic signal processing can trace its history back to 1963. E. N. Lorenz found that the small differences of initial value could lead to completely different outcomes when he was making numerical simulation for a doubtless third-order ordinary differential equation. He discovered the chaos as well as the high sensitivity on initial value of chaotic signals. After 1980s, the chaos theory has been built initially, across almost all natural science, like chaotic time series analysis and forecast theory. The following rapid development of research on theory and experiment provided a new opportunity for the applications of chaos. People actively explored the way to benefit mankind with chaos instead of avoiding and resisting chaotic phenomenon [42]. For example: in 1980, mathematician Mandelbrot drew the first image of Mandelbrot Set with computer. The Mandelbrot Set has been recognized as a symbol of chaos. Then Procaccia and Grassberger proposed the G-P algorithm for calculating the strange attractor fractal dimension of experiment system in 1983, which triggered a wave of research on calculating fractal dimension of time-sequence [43]

In 1984, Chaos, written by Chinese academician Bolin Hao, was published in Singapore. It further promoted the development of the chaos science. Wolf et al. subsequently proposed the numerical algorithm of time series of Lyapunov spectrum in 1985, which had become the most basic method to determine whether a time-sequence was a chaotic time-sequence or not. After 1990s, chaos theory and its applications in all fields had become a focus of science research, which caught multi-disciplinary researchers’ attention, and became a remarkable frontier topic and academic focus. For example, Carroll and Pecora of Naval Research Laboratory realized chaotic synchronization in the electronic circuit in 1990, which led to a new enthusiasm of the research on chaos digital communication. Then in 1992, the publishing of nonlinear science series greatly promoted the development of the theory and application research on chaos science in China , which written by Hao, Zheng and Wu [6]

As we know, chaos theory discusses the unity of certainty and randomness existing in human society and nature, as well as the unity of order and disorder. People consider that random excitation can only cause random response and certain excitation can only cause certain response. But chaos phenomena make people  surprised to find that certain excitation can also cause random response [44]. This phenomenon is a great impact for people’s traditional concept and it is time to recognize the previous conclusions. Furthermore, with the chaos point of view, it is possible to solve the problems which are difficult once [42]

Chaos measurement is a new signal processing method that differs from the existing various signal measuring methods, and it has high sensitivity. More importantly, inherent certainty of chaotic systems and non-equilibrium phase change of chaotic oscillator are sensitive to small signals and have a strong immunity to noise signals. All these characteristics determine the chaos measurement method can be used for parameter estimation problem of nanoampere signals under any zero-mean noise background. Weak signal detection technology based on chaos theory mainly includes the following forms:

1. Construct chaos measuring system with chaos theory. Because chaos system is overly sensitive to parameter perturbation, we regard measured signals as the parameter of chaos system, which means the signals become a periodic disturbance of chaos system. The system will show the cyclical and dynamic behaviors [45, 46].

2. Making use of prior knowledge that some background signals are chaos, reconstruct the phase space of the background signal according to the feature of received signals. Gain a prediction model of chaos and then subtract the predicted chaotic signal from the received signals. Finally, we will extract the measured signal from chaotic background signals [47].

        3. Chaos system is overly sensitive to initial conditions [48]. Every initial value determined the only trajectory. Small changes of initial values may make a great difference to chaotic trajectory. So, we can deduce small changes in the initial value by detecting the great changes of the trajectory. According to this theory, when measured signal enters into chaos system, its dynamic behavior will change a lot. After appropriate signal processing, we can achieve various parameters of the measured signal [49]

The research of chaos theory has decades of history. But its application in weak signal detection is just beginning. Nanoampere current detection based on chaos theory will become a branch of chaos application. And it will have wide application space and broad development prospect in the field of signal detection with its unique advantages. 

      In addition to all of the above, there are a variety of detection methods such as AD converter method, in theory, the minimum detection current can reach femtoampere [50]. However, the minimum detection range can only reach picoampere which is limited by the instrument. Another example is the capacitor charging and discharging method. This method can reduce the effects of accidental factors, reduce noise and improve the precision due to the integration effect of capacitor charging. Meanwhile, in order to filter out the noise, all kinds of filters are widely used in current detection system. For example, in [27, 50], it needs combine a variety of detection methods so as to obtain accurate measurement results. We should combine hardware circuit with computer software, bringing the noise, the measurement error and interference from outside down to the minimum value, forming a high precision detection system. 

2.6 Parallel Test System (PTS) With Picoampere Measurement Capabilities 

        In 2011, Acharyya group proposed a parallel automatic detection system that can detect DC current signal varying from 5 pA to 500 mA, its detection speed can reach up to 50Mbps [51]. The block diagram of this system is shown in Figure 11 [51], the system consists of a V93K test head, pico-ampere cards, DC parametric cards(VI32), digital cards, pogo cables, pogo-tower, wafer prober interface board (WPI BOARD)and a probe card (POGO TOWER). According to the practical demand, system channels can be configured into four different patterns: digital signal mode, digital ground model, VI direct mode and picoampere mode.

Figure 11 Block diagram of pico-ampere parallel test system. The system consists of a V93K test head, pico-ampere cards, DC parametric cards (VI32), digital cards, pogo cables, pogo tower, wafer prober interface board (WPI BOARD) and a probe card (POGO TOWER).

Figure 11 Block diagram of pico-ampere parallel test system. The system consists of a V93K test head, pico-ampere cards, DC parametric cards (VI32), digital cards, pogo cables, pogo tower, wafer prober interface board (WPI BOARD) and a probe card (POGO TOWER).

     When a channel is configured into the digital signal mode, channel relay matrix connects the output port to a channel on digital cards in the V93 test head. In the digital signal mode, the channel can provide digital I/O operation up to 50Mbps or measure frequency up to 50MHz by embedding the time interval analyzer in internal digital card. If higher digital processing capability is needed, low bandwidth pico-ampere card can be bypassed by using a different probe card. In digital ground model, the output channel is connected to digital ground.

      In the picoampere mode, it supports two ranges namely 2 nA and 200nA. The 2nA range has an accuracy of ±5 pA, a 100 fA resolution and 100 ms setting time. The 200 nA range has an accuracy of ±300 pA, a 10 fA resolution and 15 ms setting time. Linux workstation controls picoampere cards and channel relay matrix through USB, and controls V93K tester using standard optical fiber communication. 

2.6.1 Pico-ampere Instrument Card (PIC) 

        Picoampere instrument card is one of the most important components of picoampere subsystem, as shown in Figure 12 [51].

Figure 12 Pico-ampere instrument card (a) picture (b) block diagram

Figure 12|Pico-ampere instrument card (a) picture (b) block diagram

     The PIC consists of a motherboard and four picoampere daughter cards. It supports eight parallel channels and occupys a tester slot, and the physical dimensions of the card are about 15 inches × 8.5 inches. Each picoampere daughter card supports two independent channels, each channel also has its own dedicated relay matrix and picoammeters. And this topology is the key to design a real parallel system. Each picoampere card is controlled through a USB interface. The motherboard handles all USB communication and provides DC power for pico-ampere card. 

2.6.2 Picoampere Relay Matrix (PRM) 

      Designing a low leakage switch matrix requires special techniques to reduce parasitic leakage. The block diagram of picoampere relay matrix is shown in Figure 13 [42].

Figure 13 Picoampere relay matrix

Figure 13 Picoampere relay matrix

       The switch matrix is in charge of configuring each channel into different modes. A digital channel is configured by the path through relays K08-K02. Channel relay matrix connects the output port of picoampere channel to a channel on digital cards in the V93 test head. In digital signal mode, relay K03 is closed thereby grounding the guard shield. Channel can also be configured as ground through shutting relays K07-K02 and relay K03. VI direct mode can be obtained by shutting relays K09-K02 and relay K03. In this mode, picoampere channel is connected to a channel of VI32 by a single voltage force and sense lines. In picoampere mode, the voltage for device under test is generated by ‘VI-32 CH1’, as shown in Figure 13 [51]. The output of picoampere measurement system is measured by another VI32 channel, indicated by ‘VI-32 CH2’ in Figure 13 [51]. 

2.6.3 Picoampere Cable 

Reasonable cable is a very important aspect of designing a sensitive instrument. The parasitic characteristics of the cable can be neglected for measuring the current of mA level. However, these parasitic characteristics cannot be ignored for measuring current varying from picoampere current to nanoampere.

Figure 14 Picoampere cable

Figure 14 Picoampere cable

Figure 14 [51] (a) shows a coaxial cable with ground shield. A parasitic leakage path will be created due to the potential difference between the signal conductors and the ground shield. According to the environmental conditions and the type of cable insulation, the magnitude of parasitic leakage can be similar or greater than the measurement current of device under test.This source of error can affect the measurement accuracy. One method to reduce the error is by choosing the cable with insulating material made ​​of Teflon material, another electrical method is by nulling the magnitude of parasitic leakage, as shown in Figure 14(a) (bottom). Figure 14(b) shows the cross section of the cable used in low leakage path. Figure 14(c) shows the complete pico-ampere cable assembly. Figure 14(d) shows the custom made low leakage pogo block. Each pogo block provides 16 channels, each channel has its own force-pin, sense-pin and guard-pin. The guard-pin is required because it is necessary to propagate the guard signal through the entire low-leakage signal path. 

2.6.4 Wafer Prober Interface Board (WPIB) 

On the one hand, the WPIB shown in Figure 15 [51] (a) connects the pico-ampere instrument to the VI32 and digital sources. On the other hand, it connects the required pico-ampere channel to pogo tower interface. As shown in Figure 15(b), the WPIB has two different materials, uses the Rogers material as the substrate. These signal layers are used for routing the various low leakage signal traces, as shown in Figure 15(c). The second substrate material is regular FR-4 used for routing the general power and digital signal.

Figure 15 Wafer prober interface board

Figure 15 Wafer prober interface board

2.6.5 Pogo Tower 

        Pogo tower is shown in Figure 16(a) [51]. It is a standard component of V93K. Pogo-group-segment and individual-pogo-segment are shown in Figure 16(b)[51] and Figure 16(c) [51], respectively. The WPIB connects to the probe card through the pogo tower. The housing of the pogo-group-segment serving a subset of the channels is made of Ultem material. Individual-pogo-segment, serving each picoampere channel, includes guard-pin surrounded by metal housing connected to guard signal, force-pin and sense-pin. Force-pin and sense-pin are isolated by Teflon material from guard-pin. In addition to low leakage pogo-group-segment, in picoampere bypass mode, pogo tower also includes the V93K standard digital pogo-group-segment used in high speed digital test.

Figure 16 CAD drawing of pogo-tower and low-leakage pogo segments. (a) Pogo Tower. (b) Pogo-group-segment. (c) Individual-pogo-segment.

Figure 16 CAD drawing of pogo-tower and low-leakage pogo segments. (a) Pogo Tower. (b) Pogo-group-segment. (c) Individual-pogo-segment.

2.6.6 Probe Card 

The probe card is responsible for connecting all digital and analog signals to the semiconductor wafer. The probe card of picoampere system is dedicated. For measuring low current, the design of picoampere card must obey the design methodology for low-leakage including reasonable circuit layout and the selection of the correct PCB board substrate material with low leakage current, as shown in Figure 15(c). Contaminants produced by circuit board manufacturing process may not be cleaned after manufacting, and the slightest amount of contaminants will form a parasitic leakage path that affect the measurement. So, the probe card should be carefully cleaned before using. 

Conclusion and Perspectives 

Different signals and noise characteristics need different approaches for nano ampere signal detection. We have to study newer and proper techniques for various types of signals and noises. There are two aspects for nano ampere signal detection: theory and technology. 

On theoretical aspect, we may expect these breakthroughs:

  1. Noise theory, equivalent noise model and its overcoming defects. If we can clear the mechanism of noise or can accurately estimate the noise model, the noise can be restrained from the source so that the detection accuracy can be improved. If only starting from the angle of the observation, following problems remain to be solved: how to accurately determine the source signal types in observed signal, under what conditions can furthest separate source signal, how much is the distance between the numerical upper bound and the theoretical upper bound, etc.
  2. The theory of nonlinear noise reduction method. Nonlinear system theory has instability, self-organization, irreversibility, and complexity. These characteristics make it possible to detect weak current signal under the instable and non-equilibrium state. We eager to deal with the noise through nonlinear methods such as the higher order spectrum, neural network, empirical mode decomposition, chaos theory, difference vibrator, stochastic resonance. Nonlinear denoising method will be more and more widely used in related fields. At present, there are still some problems need to study, for example, the convergence rate of the neural network needs to be improved, looking for chaos detection system which is more stable and better than the Duffing oscillator method, etc.
  3. Adaptive stochastic resonance (ASR) theory. In view of different input signals, through analyzing the shape parameters’ changes of SNR and the potential, the system will automatically adjust its parameters to generate stochastic resonance, thus the strength of the useful signal will be improved. The field of non-equilibrium weak signal detection using ASR theory need to be further broadened. And it will be an inevitable trend to study how to combine all kinds of technology with different application fields.
  4. Greatly improve the SNR by average accumulation of small amount. According to the principle of statistics, average accumulation of small amount can effectively improve the SNR.
  5. Reduce the measuring time and average the random signal. Averaging the random signal can mitigate the error causing by random noise, weaken the interference of accidental factors, and improve the SNR. 

On technology aspect, some new approaches may be expected for nano ampere signal detection:

  1. Pre-amplifier: it is suggested to make us of a pre-amplifier device that has good noise characteristics and large input impedance, because the noise figure of the preamplifier has a decisive effect on noise characteristics of the entire detection circuit.
  2. Shielding measures: taking effective shielding measures to the devices whose output signal is susceptible to interfere can greatly mitigate a variety of external electromagnetic interference, and then improve the accuracy.
  3. Suppressing noise: it must take the approach of installing filters to suppress common mode noise and high frequency power supply ripple so as to effectively improve the quality of the output signal because of the existing of switching power supply circuit.
  4. Resolution: sensors are the core of detection system, and its resolution directly determined the resolution of the detection system. So we need to improve the detection precision of the sensor and mitigate noise. 

        Above all, we should design the most suitable nanoampere signal detecting instrument and improve the old instrument according to the latest testing theory. Most of equipment uses a certain unique nanoampere measurement methodology. A combination of various methodologies is thus suggested to build the equipment to achieve higher SNR and lower detection threshold. Combining various techniques enriches the signal detection and raises it up to a new height.

Besides, software and virtual instrument have promoted nano ampere signal detection technology by aid of the digital technology. We could use various processing software to further mitigate noise and outer interference together with the hardware for noise and interference suppression. There are software such as Lab VIEW, MATLAB, C languages in visual fault detect, medical electronical signal detection, and the application of DSP, FPGA, MCU in signal detecting and processing. Only under the conditions of effectively suppressing the noise and increasing the amplitude of nanoampere signals can it become possible to accurately extract the useful signal. The level of detection will be higher with the development of software, hardware and their combination. 

References

  1. G. Ferrari, F. Gozzini, A. Molari, et al., Transimpedance amplifier for high sensitivity current measurements on nanodevices. IEEE J. Solid-St Circ. 44, 1609 (2009). doi:10.1109/JSSC.2009.2016998
  2. Y. Hakamata, Y. Ohno, K. Maehashi, et al., External-noise-induced small-signal detection with solution-gated carbon nanotube transistor. Appl. Phys. Express 4, 045102 (2011). doi:10.1143/APEX.4.045102
  3. H.B. Li, G.Q. Zhang, X.G. Cai, et al., Research on small signal detection of optical voltage/current transformer. Proc. SPIE 8914, International Symposium on Photoelectronic Detection and Imaging 2013: Fiber Optic Sensors and Optical Coherence Tomography 2013, 8914, 8. doi:10.1117/12.2035253
  4. N. Verma, A.P. Chandrakasan, A High-density 45 nm SRAM using small-signal non-strobed regenerative sensing. IEEE J. Solid-St. Circ. 380 (2008). doi:10.1109/ISSCC.2008.4523216
  5. G. Svirskis, J. Rinzel, Influence of subthreshold nonlinearities on signal-to-noise ratio and timing precision for small signals in neurons: minimal model analysis. Network: Computation in Neural Systems 14, 137 (2003).
  6. Y. Hu, Study on the Weak signal detection approaches based on the specific chaotic system (M.S., Jilin University (People’s Republic of China) 2009)
  7. P.Y. Liang, Weak DC voltage signal detection (M.S., University of Electronic Science and Technology (People’s Republic of China) 2013)
  8. Y.E. Wang, Research on measurement of micro current (M.S., Xi'an University of Technology (People’s Republic of China) 2005)
  9. J.Z. Gao, Detection of weak signals (Tsinghua University Press (People’s Republic of China) 2004)
  10. Y.F. Zhang, L.F. Duan, Y.Z. Zhang, et al., Advances in conceptual electronic nanodevices based on 0D and 1D nanomaterials. Nano-Micro Lett. 6, 1 (2014). doi:10.5101/nml.v6i1.p1-19
  11. C.X. Chen, D. Xu, E.S. Kong, et al., Multichannel Carbon-nanotube FETs and complementary logic gates with nanowelded contacts.IEEE Electr. Device L. 27, 852 (2006). doi:10.1109/LED.2006.882530
  12. W.C. Ding, F. Fang, J.B. Zhou, et al., The study on micro-direct current preamplifier.Nuclear Electronics & Detection Technology 29, 853 (2009).
  13. Z.H. Huang, Weak current signal detection and reception of quasi-electrostatic field. The Ninth International Conference on Electronic Measurement & Instruments 2009, 61. doi:10.1109/ICEMI.2009.5274218
  14. Y. Li, C.Q. Song, Y.X. Hou, et al., Methods of improving performance of weak current integrated amplifier. Nuclear Electronics & Detection Technology 27, 978 (2007).
  15. Z.H. Zeng, X.Z. Chen, R.H. Hu, Design of programmable weak-current amplifier. Chinese J. Sci. Instrum. 25, 10 (2004).
  16. S.J. Yan, F. Tang, X.H. Wang, et al., Design of micro current detection circuit for FAIMS. Journal of electronic measurement and instrument 25, 711 (2011). doi:10.3724/SP.J.1187.2011.00711
  17. Y.J. Wang, L. Li, On picking-up and processing method of micro-current chemical signals. Computer Applications and Software 29, 240 (2012).
  18. G. Wielgoszewski, T. Gotszalk, M. Woszczyna, et al., Conductive atomic force microscope for investigation of thin-film gate insulators. B. Pol. Acad. Sci. 56, 39 (2008).
  19. Y.N. Yu, Research on low photocurrent measurement system for position sensitive detector (M.S., Wuhan University of Technology (People’s Republic of China) 2012)
  20. W.X. Wang, Research on detection of weak direct current (M.S., Xi'an University of Technology (People’s Republic of China) 2008)
  21. S.J. Liu, J.C. Fan, J.Z. Yan, et al., Design of wideband DC amplifier. Journal of Hubei University for Nationalities (Natural Science Edition) 29, 103 (2011).
  22. T.Y. Lin, R.J. Green, P.B. O’Connor, A low noise single-transistor transimpedance preamplifier for Fourier-transform mass spectrometry using a T feedback network. Rev. Sci. Instrum. 83, 094102 (2012). doi:10.1063/1.4751851
  23. X.L. Lu, Embedded system design for biochip electrochemical detection (Shanghai Jiao Tong University(People’s Republic of China) 2010)
  24. H.L. Huang, L.Y. Wang, Y.H. Dai, et al., Weak current detection technique for electrostatic droplet ejection. 2009 International IEEE Workshop on Intelligent Systems and Applications, Xiamen, China 2009, 1. doi:10.1109/IWISA.2009.5073236
  25. F.X. Yan, Design of an extremely weak photoelectronic current measuring circuit. Journal of WUT (Information & Management Engineering) 28, 114 (2006).
  26. J. Wang, B.K. Li, L.B. Ruan, et al., Automatic weak-current measurement system with high precision. High power laser and particle beams 24, 1975 (2012).
  27. L.Q. Wei, S.J. Lei, M.H. Fang, et al., An I-V converter in picoampere current measurement.Instrumentation Analysis Monitoring 3, 28 (2010).
  28. B. Linares-Barranco, T. Serrano-Gotarredona, R. Serrano-Gotarredona, et al., Current mode techniques for sub-pico-Ampere circuit design.Analog Integr. Circ. S. 38, 103 (2004). doi:10.1023/B:ALOG.0000011162.52504.39
  29. B. Linares-Barranco, T. Serrano-Gotarredona, R. Serrano-Gotarredona, et al., On the design and characterization of femto-ampere current-mode circuits. IEEE J. Solid-St. Circ. 38, 1353 (2003). doi:10.1109/JSSC.2003.814415
  30. J.M. Chang, A.A. Abidi, C.R. Viswanathan, Flicker noise in CMOS transistors from subthreshold to strong inversion at various temperatures. IEEE T. Electron Dev. 41, 1965 (1994). doi:10.1109/16.333812
  31. L. Zhang, X.Q. He, Z.Q. Yu, Circuit design and verification for ultra low current sensing amplifier aimed at bio-sensor applications. 2005 6th International Conference On ASIC, Shanghai, China 2005, 431. doi:10.1109/ICASIC.2005.1611353
  32. L. Zhang, Z.Q. Yu, X.Q. He, An ultra-low current mode amplifier aiming at biosensor applications. 2007 7th International Conference on ASIC 2007, 477. doi:10.1109/ICASIC.2007.4415671
  33. Y.W. Zhang, Z.W. Zhang, A kind of high-precision I/F converter. Life Sci. Instrum. 7, 52 (2009).
  34. C.F. Dong, H. Hong, J.P. Xing, et al., The I-F converter in the weak current measurement. Nuclear Electronics & Detection Technology 24, 488 (2004).
  35. G.R. Wang, The I-F converter design for very weak current measurement. Nuclear Electronics & Detection Technology 25, 358 (2005).
  36. C.F. Lin, X.B. Du, G. Li, Development of a kind of weak current tester based on I-F conversion.Journal of Shanghai Jiao Tong University 40, 1516 (2006).
  37. X.M. Liu, J. Peng, T. Liu, The design of I-F converter in weak current measurement. Journal of Lanzhou Jiao Tong University 30, 99 (2011).
  38. R. Yadav, K.R. Raghunandan, A. Dodabalapur, et al., Operational current to frequency converter. 2013 IEEE 56th International Midwest Symposium on Circuits and Systems (MWSCAS) 2013, 900. doi:10.1109/MWSCAS.2013.6674795
  39. W. Vigano, B. Dehning, E. Effinger, et al., Comparison of three different concepts of high dynamic range and dependability optimised current measurement digitisers for beam loss systems. CREAN- ATS. 279 (2012).
  40. L. Yu, Y.P. Wang, A class of high-precision I/F converter based on ADuC847. New PRODUCT & TECH 13, 53 (2013).
  41. H. Jiang, G.B. Liu, Y. Xue, Intelligentized multi channels data acquisition system based on current frequency converter.Instrum. Technol. 14 (2007).
  42. C.Y. Nie, Research on the weak signal detection method based on the chaos theory and specific chaotic system (M.S., Jilin University(People’s Republic of China) 2006)
  43. P. Grassberger, I. Procaccia, Measuring the strangeness of strange attractors. Physica D: Nonlinear Phenomena 9, 189 (1983). doi:10.1016/0167-2789(83)90298-1
  44. C.Y. Nie, Z.W. Wang, Application of chaos in weak signal detection. 2011 Third International Conference on Measuring Technology and Mechatronics Automation (ICMTMA) 2011, 1, 528. doi:10.1109/ICMTMA.2011.134
  45. Y. Li, B. Yang, The chaotic detection of periodic short-impulse signals under strong noise background.J. Electron. 19, 431 (2002). doi:10.1007/s11767-002-0078-z
  46. C. Huang, L. Chen, C.C. Yuan, et al., Chaos study of vehicle lateral dynamics based on perturbation parameter. Journal of Southeast University (Natural science edition) 42, 1111 (2012).
  47. G.W. Zhang, W.K. Shi, Q.H. Lu, Application of chaos theory in information detection. Computer Measurement & Control 10, 634 (2002).
  48. Z.B. Lin, D.H. Guo, X.L. Yu, et al., Experimental investigation on the chaotic phenomena in the wake of a natural thermal convection flow.ACTA MECHANICA SINICA (English Series) 16, 121 (2000). doi:10.1007/BF02486703
  49. D.H. Wu, Q.J. Li, P. Yang, Research on the influence of noise to weak signal detection based on duffing equation. Computer Measurement & Control 18, 61 (2010).
  50. X.Y. Yu, E. Farquhar, B. Blalock, A novel frequency based current-to-digital converter with programmable dynamic range.53th IEEE International Midwest Symposium on Circuits and Systems 2010, 869. doi:10.1109/MWSCAS.2010.5548667
  51. D. Acharyya, K. Miyao, D. Ting, et al., Architecture and implementation of a truly parallel ATE capable of measuring pico ampere level current.2011 IEEE International Test Conference (ITC) 2011, 1. doi:10.1109/TEST.2011.6139133

Acknowledgements

Acknowledgements XXX

References

 

  1. G. Ferrari, F. Gozzini, A. Molari, et al., Transimpedance amplifier for high sensitivity current measurements on nanodevices. IEEE J. Solid-St Circ. 44, 1609 (2009). doi:10.1109/JSSC.2009.2016998
  2. Y. Hakamata, Y. Ohno, K. Maehashi, et al., External-noise-induced small-signal detection with solution-gated carbon nanotube transistor. Appl. Phys. Express 4, 045102 (2011). doi:10.1143/APEX.4.045102
  3. H.B. Li, G.Q. Zhang, X.G. Cai, et al., Research on small signal detection of optical voltage/current transformer. Proc. SPIE 8914, International Symposium on Photoelectronic Detection and Imaging 2013: Fiber Optic Sensors and Optical Coherence Tomography 2013, 8914, 8. doi:10.1117/12.2035253
  4. N. Verma, A.P. Chandrakasan, A High-density 45 nm SRAM using small-signal non-strobed regenerative sensing. IEEE J. Solid-St. Circ. 380 (2008). doi:10.1109/ISSCC.2008.4523216
  5. G. Svirskis, J. Rinzel, Influence of subthreshold nonlinearities on signal-to-noise ratio and timing precision for small signals in neurons: minimal model analysis. Network: Computation in Neural Systems 14, 137 (2003).
  6. Y. Hu, Study on the Weak signal detection approaches based on the specific chaotic system (M.S., Jilin University (People’s Republic of China) 2009)
  7. P.Y. Liang, Weak DC voltage signal detection (M.S., University of Electronic Science and Technology (People’s Republic of China) 2013)
  8. Y.E. Wang, Research on measurement of micro current (M.S., Xi'an University of Technology (People’s Republic of China) 2005)
  9. J.Z. Gao, Detection of weak signals (Tsinghua University Press (People’s Republic of China) 2004)
  10. Y.F. Zhang, L.F. Duan, Y.Z. Zhang, et al., Advances in conceptual electronic nanodevices based on 0D and 1D nanomaterials. Nano-Micro Lett. 6, 1 (2014). doi:10.5101/nml.v6i1.p1-19
  11. C.X. Chen, D. Xu, E.S. Kong, et al., Multichannel Carbon-nanotube FETs and complementary logic gates with nanowelded contacts. IEEE Electr. Device L. 27, 852 (2006). doi:10.1109/LED.2006.882530
  12. W.C. Ding, F. Fang, J.B. Zhou, et al., The study on micro-direct current preamplifier. Nuclear Electronics & Detection Technology 29, 853 (2009).
  13. Z.H. Huang, Weak current signal detection and reception of quasi-electrostatic field. The Ninth International Conference on Electronic Measurement & Instruments 2009, 61. doi:10.1109/ICEMI.2009.5274218
  14. Y. Li, C.Q. Song, Y.X. Hou, et al., Methods of improving performance of weak current integrated amplifier. Nuclear Electronics & Detection Technology 27, 978 (2007).
  15. Z.H. Zeng, X.Z. Chen, R.H. Hu, Design of programmable weak-current amplifier. Chinese J. Sci. Instrum. 25, 10 (2004).
  16. S.J. Yan, F. Tang, X.H. Wang, et al., Design of micro current detection circuit for FAIMS. Journal of electronic measurement and instrument 25, 711 (2011). doi:10.3724/SP.J.1187.2011.00711
  17. Y.J. Wang, L. Li, On picking-up and processing method of micro-current chemical signals. Computer Applications and Software 29, 240 (2012).
  18. G. Wielgoszewski, T. Gotszalk, M. Woszczyna, et al., Conductive atomic force microscope for investigation of thin-film gate insulators. B. Pol. Acad. Sci. 56, 39 (2008).
  19. Y.N. Yu, Research on low photocurrent measurement system for position sensitive detector (M.S., Wuhan University of Technology (People’s Republic of China) 2012)
  20. W.X. Wang, Research on detection of weak direct current (M.S., Xi'an University of Technology (People’s Republic of China) 2008)
  21. S.J. Liu, J.C. Fan, J.Z. Yan, et al., Design of wideband DC amplifier. Journal of Hubei University for Nationalities (Natural Science Edition) 29, 103 (2011).
  22. T.Y. Lin, R.J. Green, P.B. O’Connor, A low noise single-transistor transimpedance preamplifier for Fourier-transform mass spectrometry using a T feedback network. Rev. Sci. Instrum. 83, 094102 (2012). doi:10.1063/1.4751851
  23. X.L. Lu, Embedded system design for biochip electrochemical detection (Shanghai Jiao Tong University(People’s Republic of China) 2010)
  24. H.L. Huang, L.Y. Wang, Y.H. Dai, et al., Weak current detection technique for electrostatic droplet ejection. 2009 International IEEE Workshop on Intelligent Systems and Applications, Xiamen, China 2009, 1. doi:10.1109/IWISA.2009.5073236
  25. F.X. Yan, Design of an extremely weak photoelectronic current measuring circuit. Journal of WUT (Information & Management Engineering) 28, 114 (2006).
  26. J. Wang, B.K. Li, L.B. Ruan, et al., Automatic weak-current measurement system with high precision. High power laser and particle beams 24, 1975 (2012).
  27. L.Q. Wei, S.J. Lei, M.H. Fang, et al., An I-V converter in picoampere current measurement. Instrumentation Analysis Monitoring 3, 28 (2010).
  28. B. Linares-Barranco, T. Serrano-Gotarredona, R. Serrano-Gotarredona, et al., Current mode techniques for sub-pico-Ampere circuit design. Analog Integr. Circ. S. 38, 103 (2004). doi:10.1023/B:ALOG.0000011162.52504.39
  29. B. Linares-Barranco, T. Serrano-Gotarredona, R. Serrano-Gotarredona, et al., On the design and characterization of femto-ampere current-mode circuits. IEEE J. Solid-St. Circ. 38, 1353 (2003). doi:10.1109/JSSC.2003.814415
  30. J.M. Chang, A.A. Abidi, C.R. Viswanathan, Flicker noise in CMOS transistors from subthreshold to strong inversion at various temperatures. IEEE T. Electron Dev. 41, 1965 (1994). doi:10.1109/16.333812
  31. L. Zhang, X.Q. He, Z.Q. Yu, Circuit design and verification for ultra low current sensing amplifier aimed at bio-sensor applications. 2005 6th International Conference On ASIC, Shanghai, China 2005, 431. doi:10.1109/ICASIC.2005.1611353
  32. L. Zhang, Z.Q. Yu, X.Q. He, An ultra-low current mode amplifier aiming at biosensor applications. 2007 7th International Conference on ASIC 2007, 477. doi:10.1109/ICASIC.2007.4415671
  33. Y.W. Zhang, Z.W. Zhang, A kind of high-precision I/F converter. Life Sci. Instrum. 7, 52 (2009).
  34. C.F. Dong, H. Hong, J.P. Xing, et al., The I-F converter in the weak current measurement. Nuclear Electronics & Detection Technology 24, 488 (2004).
  35. G.R. Wang, The I-F converter design for very weak current measurement. Nuclear Electronics & Detection Technology 25, 358 (2005).
  36. C.F. Lin, X.B. Du, G. Li, Development of a kind of weak current tester based on I-F conversion. Journal of Shanghai Jiao Tong University 40, 1516 (2006).
  37. X.M. Liu, J. Peng, T. Liu, The design of I-F converter in weak current measurement. Journal of Lanzhou Jiao Tong University 30, 99 (2011).
  38. R. Yadav, K.R. Raghunandan, A. Dodabalapur, et al., Operational current to frequency converter. 2013 IEEE 56th International Midwest Symposium on Circuits and Systems (MWSCAS) 2013, 900. doi:10.1109/MWSCAS.2013.6674795
  39. W. Vigano, B. Dehning, E. Effinger, et al., Comparison of three different concepts of high dynamic range and dependability optimised current measurement digitisers for beam loss systems. CREAN- ATS. 279 (2012).
  40. L. Yu, Y.P. Wang, A class of high-precision I/F converter based on ADuC847. New PRODUCT & TECH 13, 53 (2013).
  41. H. Jiang, G.B. Liu, Y. Xue, Intelligentized multi channels data acquisition system based on current frequency converter. Instrum. Technol. 14 (2007).
  42. C.Y. Nie, Research on the weak signal detection method based on the chaos theory and specific chaotic system (M.S., Jilin University(People’s Republic of China) 2006)
  43. P. Grassberger, I. Procaccia, Measuring the strangeness of strange attractors. Physica D: Nonlinear Phenomena 9, 189 (1983). doi:10.1016/0167-2789(83)90298-1
  44. C.Y. Nie, Z.W. Wang, Application of chaos in weak signal detection. 2011 Third International Conference on Measuring Technology and Mechatronics Automation (ICMTMA) 2011, 1, 528. doi:10.1109/ICMTMA.2011.134
  45. Y. Li, B. Yang, The chaotic detection of periodic short-impulse signals under strong noise background. J. Electron. 19, 431 (2002). doi:10.1007/s11767-002-0078-z
  46. C. Huang, L. Chen, C.C. Yuan, et al., Chaos study of vehicle lateral dynamics based on perturbation parameter. Journal of Southeast University (Natural science edition) 42, 1111 (2012).
  47. G.W. Zhang, W.K. Shi, Q.H. Lu, Application of chaos theory in information detection. Computer Measurement & Control 10, 634 (2002).
  48. Z.B. Lin, D.H. Guo, X.L. Yu, et al., Experimental investigation on the chaotic phenomena in the wake of a natural thermal convection flow. ACTA MECHANICA SINICA (English Series) 16, 121 (2000). doi:10.1007/BF02486703
  49. D.H. Wu, Q.J. Li, P. Yang, Research on the influence of noise to weak signal detection based on duffing equation. Computer Measurement & Control 18, 61 (2010).
  50. X.Y. Yu, E. Farquhar, B. Blalock, A novel frequency based current-to-digital converter with programmable dynamic range. 53th IEEE International Midwest Symposium on Circuits and Systems 2010, 869. doi:10.1109/MWSCAS.2010.5548667
  51. D. Acharyya, K. Miyao, D. Ting, et al., Architecture and implementation of a truly parallel ATE capable of measuring pico ampere level current. 2011 IEEE International Test Conference (ITC) 2011, 1. doi:10.1109/TEST.2011.6139133

How to Citation

How to citation XXX

History

Article History XXX

 


Additional Info

  • Type of Publishing: JOUR - Journal
  • Title:

    Review on Signal Detection of Low DC Current in Nanodevices and Various Sensors

  • Author: Yantao Liang, Jiapin Chen, Franklin Li Duan, Dong Xu, Jun Wang, Hanling Yang, Yafei Zhang
  • Year: 2014
  • Volume: 1
  • Issue: 1
  • Journal Name: Materials and Electronics Engineering
  • Publisher: Nicety Press Company Limited
  • ISSN: 2410-1648
  • URL: http://www.meej.org/volume-1/december-2014/item/363-review-on-signal-detection-of-low-dc-current-in-nanodevices-and-various-sensors
  • Abstract:

    As the operating current of various new nanoelectronic devices and signals of various sensors are scaled down to nano-ampere range, technologies for ultralow current detection thus become critical for the development of nanodevices and related fields such as sensors. Low current characterization is roughly divided into two categories: direct methods and indirect methods. The direct approach uses the high precision current meter / equipments to detect the small signals which may reach the resolution of 10-15 A. This sort of equipment is usually expensive and not quite suitable for general large scale application. The indirect method makes use of various electronic circuitry and theory to measure the ultralow electric current and can reach the resolution up to 10-14A. In this paper, we mainly review and compare a few popular indirect methods, including: sampling resistors, feedback operational amplifiers, biochip method based on MOSFET operating in sub-threshold region, I-F transformation, chaos theory and parallel test system with pico-ampere measurement capabilities. Pros and cons of direct approaches are also compared both for direct and indirect categories.

  • Publish Date: Wednesday, 10 December 2014
  • Start Page: 1
  • DOI: 10.11605/mee-1-1