• Tidak ada hasil yang ditemukan

Trade-off and Design optimization of the Notch filter for ultralow power ECG application

N/A
N/A
Protected

Academic year: 2023

Membagikan "Trade-off and Design optimization of the Notch filter for ultralow power ECG application"

Copied!
81
0
0

Teks penuh

I declare that this written submission represents my ideas in my own words, and where the ideas or words of others are included, I have adequately cited and referenced the original sources. I also declare that I have complied with all principles of academic honesty and integrity and have not misrepresented, fabricated or falsified any idea/data/fact/source in my submission. First of all, I want to thank God Almighty for keeping me in good health and for all the grace.

Ashudeb Dutta for his valuable guidance, consistent moral support and impetus that helped me in the achievement of my dissertation. I would like to thank my senior Pravanjan Patra Sir who guided me in my project and helped me understand many concepts. I would also like to thank my dear friends Prakash Kumar Lenka, Narendra Nath Ghosh and Pankaj Kumar Jha Sir for being there, both professionally and personally, at all times.

Taking an EKG, several leads combined with signals from different parts of the body (ie, from the right wrist and left wrist) are used to track the electrical activity of the heart. Therefore, we need to design a power line interference filter to avoid signal contamination.

Motivation

It consists of AFE, including instrumentation amplifiers, followed by two PGAs, a mixed-signal AGC and a 10-bit SAR ADC. The ECG acquisition board translates the body signal into six leads and processes the signal using a low-pass filter (LPF) and SAR ADC. The acquisition board consists of: an instrumentation amplifier, a high-pass filter, a 60 Hz notch filter and a common-level regulator.

Requirement for designing a Notch Filter

Design of Passive Notch Filter

Why do we need Active Notch Filter?

Active Notch Filter

Tow-Thomas Architecture

  • Notch Frequency
  • Notch Depth
  • Quality Factor

Modelling of each block

Operation of circuit at lower frequencies

For lower frequencies, as we saw, the input is high-pass filtered by the 1st amplifier. When the loop gain goes to unity, it gives -3dB point for the lower cutoff frequency. As R1 is increased, the cutoff frequency can be shifted, but this leads to a decrease in peaking as the attenuation decreases.

Here the loop gain is found by applying a Vtest at the input to R4 and finding feedback at Vfeed.

Operation of circuit at higher frequencies

Gain: For any CMOS topology, the gain of the circuit is given by the product of its transconductance and the output resistance of the load structure. Common Mode Input Range: This is an important parameter at the input to the circuit. The saturation voltage of the load structure mainly defines the output swing of the amplifier.

Common Mode Rejection Ratio (CMRR): The common mode gain is given by the gain of the common mode input to the output. The common mode rejection ratio is the difference between the differential gain and the common gain. Rout is the output impedance and Cload is the output capacitance of the telescopic differential amplifier.

Non-dominant pole-: Non-dominant pole exists in the cascode node of the differential amplifier. To improve the gain of the differential amplifier, the transistors are cascoded, which further leads to the reduction of the oscillation of the output signal. Gmfc represents the transconductance and Routfc represents the output impedance of the folded OTA cascode.

The dominant pole is at the output node and the non-dominant pole is at the cascode node of the folded cascode OTA. The signal flowing through the current mirror node becomes larger compared to that of the RFC due to the increased impedance of the small signal path. The discharge current of the weak-inversion MOS is proportional to the exponential of the effective gate-source voltage.

59 To address this inefficiency, a modified FC is presented in Fig. The proposed changes are intended to use M3 and M4 as driving transistors. Putting all the values ​​in the above equation, the transconductance of the input transistor is 1.05 mS. The notch gain decreases as the loop gain of the high-pass path reaches unity.

Fig 3.1 Telescopic Differential Amplifier
Fig 3.1 Telescopic Differential Amplifier

Optimizing R and C Values

OTA Terminology

Where 𝑓𝑢 represents the unity gain bandwidth and 𝑔𝑚 represents the transconductance of the circuit and 𝐶𝑙 represents the load capacitance. Higher PM values ​​will allow the output signal to reach steady state without much ringing. The saturation voltage of the bias architecture, the saturation voltage and the voltage space 𝑉𝐺𝑆 of the transconductance structure determine the CMIR.

In general, the cascode structures result in low output swing as more number of transistors are stacked under each other and the 𝑉𝐷𝑆𝑠𝑎𝑡 subtracted from the 𝑉𝐷𝐷 provides the output voltage swing. 𝑂𝐷𝑅 = 𝑉𝐷𝐷− ∑ 𝑉𝐷𝑆𝑠𝑎𝑡𝑉𝑆𝑠𝑎𝑡 (3.5) Static Current consumption defines all the product of the current branch and the supply of the current specific branch. 3.6) The speed of an amplifier is dependent on the equivalent RC constant at the output node.

So it is a strong function of the internal capacitances and currents in the amplifier branches. The CMRR must be very high so that the cancellation of the signals is good.

Different OTA Configuration

Reduction in conduction efficiency and increase in drain-source saturation voltage are disadvantages of MOS compared to weak inversion. The lower cut-off frequency is equal to the UGB of the global gain of the low-pass filter loop.

Recycling folded cascode OTA

Improved Recycling folded cascode OTA

Comparison of all the OTA

Low compared to folded cascode OTA due to the current mirror node in the signal path. Gm/ID is the ratio obtained from full process characterization for NMOS and PMOS transistors. In saturation, there are three regions of operation that depend on the rms voltages applied to the gate of the transistor.

The transconductance of the 2nd stage should be 10 times higher compared to the transconductance of the 1st stage to have resistanceless compensation. By choosing the same operating range, the DC current through M3b is the same as that of M1a and is 8nA. 69 6.5 Relationship between closed-loop forward gain and notch depth with different values ​​of q.

70 6.6 Relationship between Closed-Loop Front Path Gain and Notch Depth with Different Values ​​of q. 71 6.7 Relation between forward path UGB loop gain and band drop for various values ​​of q .

Fig. 4.1 The Problem while designing
Fig. 4.1 The Problem while designing

Application of OTA

Introduction

Different regions of operation

Performance Comparison among Different regions of operation

Different types of Model

Recently, one of the most commonly used architectures, whether a single stage or the first stage of multistage amplifiers, had been the folded cascode (FC) amplifier due to its high gain and reasonably large signal swing in the current and future low voltage CMOS processes. Since we are working in weak inversion, gm/ID is chosen to be very high, so the current consumption is very low and the parasitic capacitance is very high. The transconductance of M1a and M2a is assumed to be equal, so the current consumed by them is also the same.

By choosing the same 𝑔𝑚⁄𝐼𝐷 for M1a and M3a, the DC current required is three times the input, so the DC current is 24nA. Using KCL at the cascode junction, the current flowing in M1a is 8nA and the current flowing in M3a is 24nA, so the current flowing through M5 and M7 is 16nA. Since the path M2b, M11 and M3b act as a unity gain amplifier, the transconductance of M3b should be.

Since the current flowing through M5, M7 and M9 is the same, the current consumed by them is also the same 16nA. The gain of the OTA decreases with the increase of the channel inversion, since the channel inversion increases the gate-to-source voltage, leading to an increase in the drain-to-source potential, and the resistance becomes dependent on the drain-to-source potential, leading to a decrease in the gain of OTA. It can be seen that at the crossing point of low-pass loop gain and high-pass transfer function we get a notch.

Fig: 5.1 Folded Cascode Amplifier
Fig: 5.1 Folded Cascode Amplifier

Recycling Folded Cascode Amplifier

Two Stage Recycling Folded Cascode Amplifier

OTA Design requirements

OTA Design Parameter

Power Consumption of First Stage OTA

All transistors in the RFC OTA operate in a weak inversion region. It can also be observed that the lower passband and upper passband have zero dB gain up to the bandwidth requirement.

Recycling Folded Cascode

Design of RFC

Gambar

Figure 1. Bio-medical sensor interface
Fig: 2.3 Closed Loop block diagram
Fig 3.1 Telescopic Differential Amplifier
Fig: 3.3 Half circuit for calculation of common mode gain of the Telescopic OTA
+7

Referensi

Garis besar

Dokumen terkait

ANNEX 2-D – GENERAL NOTES – MEXICO – 3 year 1 through year 5 and shall be eliminated in 11 annual stages beginning in year 6, and these goods shall be duty-free effective January 1 of