I must thank him for his right help in the right places and at the right times, without which my research journey would not have been fulfilled. The configurability of the downconversion mixers comes in the form of active/passive and bandwidth (RF) and output (IF) reconfigurability.
Introduction
Thus, for a continuous wideband receiver, a tunable RF filter is better suited for dynamically reconfigurable blocking rejection. Our goal is to design a low-noise wideband receiver for radar and satellite applications without detection and rejection of out-of-band blockers in the input phase.
Aim and Motivation
Furthermore, an independently reconfigurable active/passive mixer stage is proposed, and a comparison is made with the previous passive in simulation mode.
Highlights of Research Investigations
Overview of Thesis Contribution
- Blocker Rejection Wireless Receiver Techniques
- High Performance Bandwidth Extension Low Noise Main Path ReceiverReceiver
- Auxiliary Path Ultra low power Out of Band Sensing Subthresh- old Receiver
- Reconfigurable Down Conversion Mixer
Furthermore, to implement a low-power ultra-wideband down-conversion mixer in subthreshold region, the conventional Gilbert cell topology is modified. Furthermore to reduce the area of the chip-down conversion mixer, it is designed with RF bandwidth reconfigurability in single circuits using switches.
Introduction
In this chapter, various metrics of blockers have been discussed in detail, especially their impact on receiver performance.
Effect of Blocker on Wideband Receiver
Blocker Impact on Various Parameter
Moreover, if the blocker has high phase noise, which also contributes to the overall noise floor level. If the blocking is close to the receiver signal, blocking and transmitter leakage will cause 3rd order intermodulation (IMD3) and cross-modulation (XMD) simultaneously due to LNA nonlinearity.
Wideband Radar/Satellite Receiver design Perspec- tivetive
On Chip Blocker Removal Techniques for Wide Band Receiver
One way to reduce the out-of-band blocking is through forward injection shown in fig. In that sense, the interference cancellation loop acts as a control loop that suppresses the blocking of the open loop gain.
Proposed Design Concern on Wireless Receiver Ap- proachproach
Identified Blockers and Strength
Proposed Design Approach
In the auxiliary path, the receiver will tune the notch filter for the out-of-band frequency (3 GHz - 5 GHz) so that the main receiver does not become saturated with interfering factors. If any out-of-band frequency jammer is present, the auxiliary path receiver with a notch filter setting to reject or attenuate the jammer.
Summary
An auxiliary path receiver includes subthreshold LNA down-conversion mixers connected to the main path input that detect the out-of-band interferences and tune the notch filters located at a fixed frequency of 4 GHz and tune left and right from 4 GHz according to the available RF band. After looking at all the techniques, the multipath receiver seems to be a good choice, so for radar and satellite applications, a wideband receiver with low noise and sub-threshold out-of-band detection solution has been approached.
Introduction
An overview of recently reported low noise wideband receiver performance is listed in Table 3.1. Various papers from LNA and blend papers have been reviewed and listed in the sections below.
Wide Band Receiver Design Specification
- Sensitivity
- Noise Figure
- Selectivity
- Linearity
In the sections below, broadband input matching and noise cancellation topologies for LNA and downconversion mixer operation are discussed in detail. Selectivity is the maximum signal level that a receiver can demodulate and decode the information without error in the presence of much stronger interferers (in-band/out-of-band).
Receiver Front-End Architecture
Target Specification
Circuit Implementations
Wide Band Low Noise Amplifier (LNA)
Where Co and Cs are the output and source paracities as shown in Fig. where gm,nL and gm,pL are the conductance of NMOS and PMOS transistor with inductive degeneration as shown in Fig. a) Simulated S11 LNA with and without inductor (b) Simulated Zin(Y0) with and without the inductor. 3.4.1.1.2 Noise analysis The various noise sources that affect the total noise figure of the LNA are shown in the figure. A composite NMOS/PMOS transistor is used as the base cell to reduce the total noise figure of the LNA.
3.2(b) is calculated using the small signal half-circuit model shown in Figure ωps= gmnL+gmpL Cgs,p+Cgs,n+Cs. 3.28) where Av,mid is the midband gain, ωpo is the output pole, and ωps and ωz are due to the parasitic capacitances Cgs,n, Cgs,p and Cs. Cgs,n, Cgs,p are the gate source capacitance of NMOS/PMOS transistor while CoenCs are the output and source capacitances as shown in Figure a) Bandwidth extension with inductor and loss due to buffer of LNA.
Down Conversion Mixer
In the MGTR amplifier, however, this negative peak of the main transistor (MT) can be canceled by the positive peak of an appropriately sized auxiliary transistor whose transfer characteristic is shifted to the right by changing either the gate bias or the threshold voltage as shown in Fig. The biggest advantage of the passive switching is that it does not spread static effect. If the voltage height is not a limitation, the common-mode voltage level can be reduced to allow higher LO voltage swing and higher overdrive voltage of the switches (as high as Vdd/2 for overdrive).
A higher overdrive voltage results in a lower average on-resistance of the switches and increases the linearity, gain and noise performance of the mixer.
Analysis of Down Conversion Mixer
The linearity performance of the mixer depends on: 1) the linearity of the voltage-to-current conversion in the transconductance stage, 2) effects from the coupling stage, and 3) the linearity of the transimpedance amplifier stage. However, the linearity of the transimpedance amplifier strongly depends on the frequency shifts of the blocking signal from the carrier wave, which can be explained as follows. Assuming an amplifier has a forward voltage transfer function of )A(f), the total loop gain of this amplifier can be written as
It is worth mentioning that the linearity of A(f) is not constant as a function of frequency and it also affects the overall linearity of the circuit.
Wide Band Receiver Taped Out Design
Experimental Results
Since the measurement cable is lossless, the measured noise figure number must be subtracted by the input cable attenuation as well as balun loss. If the loss at the output is small, the LNA noise figure can be estimated as follows. As shown in the plot, the minimum measured noise figure is about 1.7 dB at about 1.2 GHz and 12.5 mA bias current.
A differential buffer is added at the output of the receiver to isolate the loading effects seen in the mixer output stage.
Summary
Introduction
Subthreshold MOS Operation
The major advantage of biasing a CMOS transistor in the subthreshold region is the significant increase in the ratio of transconductance to bias current (gm/ID) compared to strong inversion operation. As can be seen from (4.4), increasing W/L without changing ID does not increase the transconductance, in contrast to strong inversion. 4.3(a), gm/ID increases by approx. an order of magnitude from strong inversion to weak inversion in all process nodes.
This is as high as the peakfT of the 180 nm device in strong inversion shown in Fig. 2b.
Power Consumption in Wireless Front-ends
Subthreshold Low Noise Amplifier
Design and Implementation Subthreshold LNA
The LNA contains a common gate (CG) stage, followed by a CS-CG stage with recycled power, gain gain, and feed-forward noise reduction. The CS-CG stage load, Ld3, provides shunt peaking by resonating with the total capacitance at drain from M3, providing bandwidth extension. Achieving this transconductance in a subthreshold regime would require a large device, resulting in a large input capacitance (Cgs), which in turn.
However, by introducing an inductor (Lg1) at the gate of the input device, the input matching dependence on gm is minimized.
Active Gilbert Cell Down Conversion mixer
Transfer Function of Active Gilbert Cell Down Conversion Mixer
A peak capacitor provides a zero in the transfer function, which widens the bandwidth of the transconductance phase with relaxed degeneracy of the resistive source, as shown by the following expression and Fig. By inserting zero into a transfer function, it adds +20dB and cancels the pole as one as a result increases the high-frequency response of the circuit, since from (4.11), increases the high-frequency circuit performance.
Noise in Current-Commutating Mixers
Non-idealities of the Mixer
Compared to the conventional Gilbert cell mixer, degeneration introduces an additional fourth term, as given in (4.14). The third-order nonlinearity of MOS mixers with Gilbert cells is mainly determined by the intrinsic device nonlinearity of the RF transconductor. The balun is multifunctional and is used both in the output of the LNA and to convert the single-ended frequency to differential, due to the fact that the LO port of the mixer must operate differentially.
An effort has been made to minimize as many different bias sources as possible through the use of resistive dividers and blocking capacitors which efficiently channel bias to the necessary points.
Fabricated Fully Subthreshold Receiver
Measurement Results of Auxiliary Receiver
Differential buffer was added to the receiver output to isolate the load effects seen in the mixer's output stage, an off-chip IF balun (mini-circuit) is used for differential-to-single-ended measurements. The front-end voltage gain is approximately 3 dB lower than the power gain. The estimated balun and cable loss (obtained from separate cable and balun measurements using a network analyzer) is 3.1 dB, and the de-embedded DSB NF is 12 dB −3.1 dB = 8.9 dB.
At the input power of -26 dBm, the output voltage is approximately 0.7 Vp-p on either side of.
Performance Summary
Summary
Introduction
Wide-Band Reconfigurable (Active / Passive) Down Conversion MixerConversion Mixer
- Transconductance Amplifier
- Common Source Reconfigurable Mixer
- Transimpedance Amplifier
- Simulated results
The width of PMOS is chosen to provide degeneracy resistance, which changes the overall mixer topology into a passive mode, as shown in figure. The bias voltage can be selected to control parameters of the input stage or switching operation of Gm MOS switch 5-6, as shown in path 2. The Gm of MOS Mn1 and Mn2 can be changed by changing the value of the bias voltage, which increases the gain of the mixer is varied. A simplified schematic of the transimpedance amplifier is shown in figure. TIA consists of an operational transconductance amplifier with a feedback RFCF.
A two-stage mill-compensated OTA topology was chosen for the TIA design, as shown in Fig.
A Low/High Band Parallel path TCA with config- urable (Active / Passive) Down Conversion Mixer
- Transconductance Amplifier
- Active/Passive Mixer
- Transimpedance Amplifier
- Simulated Result
Reconfigurable down-conversion mixer in active mode has the better FOM compared to others. The voltage conversion gain is close to 22/26 dB and 25/31 for LB and HB case respectively where these figures represent passive/active (PLB, PHB, ALB, AHB) mode. Due to high conversion gain at low IF, the output compression point of the OPAMP limits the input-referenced linearity of the circuit. 1dB compression point of the circuit is limited by the output swing and varies with IF frequency.
By varying the RF value and CF gain, you can vary the gain with a constant bandwidth, or by varying only the RF, you can vary the conversion gain with small bandwidth changes.
Reconfigurable High/Low band Passive Down Conver- sion mixer for wide band Receiversion mixer for wide band Receiver
- Transconductance Amplifier
- Switches
- Transimpedance Amplifier
- Simulation Results
It is necessary to increase the transconductance of the Gm stage because the loss at the next passive switching stage degrades the overall mixer NF. LO signals are AC coupled via capacitors while DC bias level of the switches is set to achieve the lowest resistance while preventing overlap on periods. Due to high conversion gain at low IF, the output compression point of the OPAMP limits the input-referenced linearity of the circuit.
The 1dB compression point of the circuit is limited by the output swing and varies with the IF frequency.
Fabrication of Wide-Band Reconfigurable (Active / Passive) MixerPassive) Mixer
Measurement Result
A custom PCB is used to generate control signals for selecting the operating modes shown in Fig. Agilent noise figure analyzer, N8975A is used to measure the noise figure and IF spectrum. a) Micrograph of the reconfigurable down-conversion mixer chip.
Summary
Measured conversion gain shows 3 dB bandwidth from 1 to 4GHz for active cabinet and 0.5 -5 GHz for passive cabinet.
Conclusion
Future Scope
An L-band receiver front-end architecture using adaptive Q enhancement techniques in 65nm CMOS as an enabler for single-SAW GPS receivers. 0.6−3GHz wideband receiver RF front end with a feedforward noise and distortion cancellation, resistive feedback LNA. Multi-band, multi-mode, low-power CMOS receiver front end for sub-GHz ISM/SRD band with narrow channel spacing.
A low-cost, low-power CMOS front-end receiver for ultra-wideband MB-OFDM systems.
