Advanced Design System

DESIGN AND ANALYSIS OF A SINGLE-STUB NOTCH FILTER USING AGILENT’S ADVANCE DESIGN SYSTEM (ADSTM)

ABSTRACT:

The purpose of this case study is to acquire an idea on the design of single-stub notch filters using Agilent advanced design system (ADSTM). By properly calculating the required width, length and insertion loss of the single stub notch filter using ADS one can design a notch filter which can block frequencies not required. In the micro strip layout when wavelength of the stub is , the open circuit of the stub is converted to short circuit and signals along the notch filter are blocked. By adjusting the width and using various functions like line calc the parameters of the filter are calculated and the filter is designed and analysed. Agilent advanced system is an effective software for the analysis of the microwave links.

INTRODUCTION:

Advanced Design System (ADSTM): Advanced Design System is the industry leader in high-frequency domain design. It supports electronic systems and RF design engineers developing all types of RF designs, from simple to the most complex, from RF or microwave modules to be integrated MMICs for communications and aerospace/defense applications.

ADS is With a complete set of simulation technologies ranging from frequency, time, numeric and physical domain simulation to electromagnetic field simulation, ADS lets designers fully characterize and optimize designs. The single, integrated design, GUI graphical user interface environment provides system, circuit, and electromagnetic simulators, along with schematic capture, layout, and verification capability eliminating the starts and stops associated with changing design tools in mid-cycle.

ADS can be used for virtual prototyping, debugging, or as an aid in manufacturing test. To enhance engineering productivity and shorten time-to-market, ADS software offers a high level of design automation and applications intelligence. This proven software environment is easily extensible: we can customize ADS by adding features focused on your particular application needs. An AD runs on PCs and workstations, with complete file compatibility between platforms and across networks. [8]

Advanced Design Systemis a powerful electronic design automation software used by leading companies in the wireless communication & networking and aerospace & defence industries. For WiMAX™, LTE, multi-gigabit per second data links, radar, & satellite applications, ADS provides full, standards-based design and verification with Wireless Libraries and circuit-system-EM co-simulation in an integrated platform.

Key Benefits of ADS

· Complete, integrated set of fast, accurate and easy-to-use system, circuit & EM simulators enable first-pass design success in a complete desktop flow.

· Application-specific Design Guides encapsulate years of expertise in an easy-to-use interface.

Components used in (ADSTM) system

Term (Port Impedance for S-parameters):

Parameters:

Name

Description

Units

Default

Num

Port number

Integer

1

Z

Reference impedance, use 1+j*0 for complex

Ohm

50

Noise

Enable/disable port thermal noise: yes, no (for AC or harmonic balance analysis only; not for S-parameter analysis)

None

yes

V(DC)

Open circuit DC voltage

None

None

Temp

Temperature

oC

None

Table1: Parameters of Term

Note:

‘Term’ can be used in all simulations. For S-parameter simulations it is used to define the impedance and location of the ports. When not in use, it is treated as an impedance with the value R + JX. The reactance is ignored for dc simulations.

MLOC (Micro strip Open-Circuited Stub):

MLOC symbol

MLOC Illustration

Parameters:

Name

Description

Units

Default

Subst

Substrate instance name

None

MSub1

W

Line width

mil

25.0

L

Line length

mil

100.0

Wall1

Distance from near edge of strip H to first sidewall; Wall1 > 1/2 × Maximum( W, H)

mil

1.0e+30

Wall2

Distance from near edge of strip H to second sidewall; Wall2 > 1/2 × Maximum( W, H)

mil

1.0e+30

Temp

Physical temperature (see Notes)

°C

None

Mod

Choice of dispersion model

None

Kirschning

Table 2: Parameters of MLOC

Range of Usage:

1≤Er ≤128 ; 0.01 ≤ ≤ 100

Where, Er = dielectric constant (from associated Subst)

H = substrate thickness (from associated Subst)

Recommended Range for different dispersion models

Kirschning and Jansen: 1≤Er ≤20; 0.1×H≤W ≤100×H

Kobayashi: 1 ≤Er ≤128; 0.1×H ≤ W ≤10×H; 0 ≤H≤0.13×λ

Yamashita: 2 ≤Er ≤16; 0.05×H ≤W ≤16×H

Where, λ= wavelength; freq ≤ 100 GHz

Notes and Equations :

1. The frequency-domain analytical model uses the Kirschning and Jansen formula to calculate the static impedance, Zo, and effective dielectric constant, Eeff. The attenuation factor, α, is calculated using the incremental inductance rule by Wheeler. The frequency dependence of the skin effect is included in the conductor loss calculation. Dielectric loss is also included in the loss calculation.

2. Dispersion effects are included using either the improved version of the Kirschning and Jansen model, the Kobayashi model, or the Yamashita model, depending on the choice specified in Mod. The program defaults to using the Kirschning and Jansen formula.

3. For time-domain analysis, an impulse response obtained from the frequency analytical model is used.

4. The “Temp” parameter is only used in noise calculations.

5. For noise to be generated, the transmission line must be lossy (loss generates thermal noise).

6. To turn off noise contribution, set Temp to −273.15°C.

7. When the Hu parameter of the substrate is less than 100×H, the enclosure effect will not be properly calculated if Wall1 and Wall2 are left blank.

8. Wall1 and Wall2 must satisfy the following constraints:

Min(Wall1) > 1/2×Maximum(W, H)

Min(Wall2) > 1/2×Maximum(W, H)

MLIN (Micro strip Line):

MLIN symbol

MLIN Illustration

Parameters:

Name

Description

Units

Default

Subst

Substrate instance name

None

MSub1

W

Line width

mil

25.0

L

Line length

mil

100.0

Wall1

Distance from near edge of strip H to first sidewall; Wall1 > 1/2 × Maximum( W, H)

mil

1.0e+30

Wall2

Distance from near edge of strip H to second sidewall; Wall2 > 1/2 × Maximum( W, H)

mil

1.0e+30

Temp

Physical temperature (see Notes)

°C

None

Mod

Choice of dispersion model

None

Kirschning

Table 3: Parameters of MLIN

Range of Usage: 1 ≤ ER ≤ 128; 0.01 ≤ ≤ 100

Where, ER = dielectric constant (from associated Subst)

H = substrate thickness (from associated Subst)

Recommended Range for different dispersion models

Kirschning and Jansen: 1 ≤ Er ≤ 20; 0.1 × H ≤ W ≤ 100 × H

Kobayashi: 1 ≤ Er ≤ 128; 0.1 × H ≤ W ≤ 10 × H; 0 ≤ H ≤ 0.13 × λ

Yamashita: 2 ≤ Er ≤ 16; 0.05 × H ≤ W ≤ 16 × H

Where λ= wavelength; freq ≤ 100 GHz

Notes and Equations:

1. The frequency-domain analytical model uses the Hammerstad and Jensen formula to calculate the static impedance, Zo, and effective dielectric constant, Εeff. The attenuation factor, α, is calculated using the incremental inductance rule by Wheeler. The frequency dependence of the skin effect is included in the conductor loss calculation. Dielectric loss is also included in the loss calculation.

2. Dispersion effects are included using either the improved version of the Kirschning and Jansen model, the Kobayashi model, or the Yamashita model, depending on the choice specified in Mod. The program defaults to using the Kirschning and Jansen formula.

3. For time-domain analysis, an impulse response obtained from the frequency analytical model is used.

4. The “Temp” parameter is only used in noise calculations.

5. For noise to be generated, the transmission line must be lossy (loss generates thermal noise).

6. To turn off noise contribution, set Temp to −273.15°C.

7. When the Hu parameter of the substrate is less than 100 × H, the enclosure effect will not be properly calculated if Wall1 and Wall2 are left blank.

8. Wall1 and Wall2 must satisfy the following constraints:

Min(Wall1) > 1/2 × Maximum(W, H)

Min(Wall2) > 1/2 × Maximum(W, H)

MTEE (Microstrip T-Junction):

MTEE symbol

MTEE Illustration

Parameters:

Name

Description

Units

Subst

Microstrip substrate name

None

W1

Conductor width at pin 1

Mil

W2

Conductor width at pin 2

Mil

W3

Conductor width at pin 3

Mil

Temp

Physical temperature

°C

Table 4: Parameters of MTEE

Range of Usage:

0.05 × H ≤ W1 ≤ 10 × H; 0.05 × H ≤ W2 ≤ 10 × H; 0.05 × H ≤ W3 ≤ 10 × H

Er ≤ 20

Wlargest/Wsmallest≤ 5

where

Wlargest, Wsmallest are the largest, smallest width among W2, W2, W3

f(GHz) × H (mm) ≤ 0.4 × Z0

Z0 is the characteristic impedance of the line with Wlargest

Notes and Equations:

1. The frequency-domain model is an empirically based, analytical model. The model modifies E. Hammerstad model formula to calculate the Tee junction discontinuity at the location defined in the reference for wide range validity. A reference plan shift is added to each of the ports to make the reference planes consistent with the layout.

2. The center lines of the strips connected to pins 1 and 2 are assumed to be aligned.

3. For time-domain analysis, an impulse response obtained from the frequency-domain analytical model is used.

4. The “Temp” parameter is only used in noise calculations.

5. For noise to be generated, the transmission line must be lossy (loss generates thermal noise).

Single-stub notch filter:

In Radio Communication Systems, undesired harmonics are generated. A micro strip notch filters undesired harmonics in a narrow band device like a mobile phone.

A Notch filter is a device that passes all frequencies except those in a stop band centred on a centre frequency. The quality factor plays a major role in eliminating the frequencies undesired. Quality factor (Q) of a band pass or notch filter is defined as the centre frequency of a filter divided by the bandwidth.

Where, bandwidth is the difference between frequency of the upper 3dB roll off point and frequency of the lower 3dB roll off point.

TRANSMISSION LINE THEORY:

place to another for directing the transmission of energy, such as electromagnetic waves or acoustic waves, as well as electric power transmission. Components of transmission lines include wires, coaxial cables, dielectric slabs, optical fibres, electric power lines, and waveguides.

Consider the micro strip layout of a notch filter,

In the designing of the micro-strip circuits (i.e. filters), the basic parameters are impedance Z0 and guide wavelength λg which are considered as TEM transmission line.

The impedance in the open circuit stub Zin is as given below,

Zin = ZS}

Where ZL=∞, so we ignore ZS

Zin = ZS} = ZS { } = ZS { } = – j ZS cot β l

However,

l = , β l = =

Therefore, cot β l=0

So, Zin = -j ZS cot β l=0

hence ┌L = = ∞/∞ = 1

so VSWR = = 2/0 =

This indicates that the signal whose wavelength is will have very low impedance and hence it is a short circuit

Thus Insertion loss response at frequency f0 is high except for other frequencies, this is because cot β l is no longer zero.

Insertion loss and return loss are two important data to evaluate the quality of many passive fiber optic components, such as fiber optic patch cord and fiber optic connector and many more.

Insertion loss:

Definition –

The Insertion Loss of a line is the ratio of the power received at the end of the line to the power transmitted into the line.

Insertion loss refers to the fibre optic light loss caused when a fibre optic component insert into another one to form the fibre optic link. Insertion loss can result from absorption, misalignment or air gap between the fibre optic components. We want the insertion loss to be as less as possible. Our fibre optic components insertion loss is less than 0.2dB typical, less than 0.1dB types available on request.

An expression for insertion loss is

IL= 10log10 [1 +(YS/2)2]

Return loss:

Return Loss is a measure of the reflected energy from a transmitted signal. It is commonly expressed in positive dB’s. The larger the value, the less energy that is reflected.

Return loss can be calculated using the following equation:

IMPRLT10.gif (1294 bytes)

Return loss is a measure of VSWR (Voltage Standing Wave Ratio), expressed in decibels (db). The return-loss is caused due to impedance mismatch between two or more circuits. For a simple cable assembly, there will be a mismatch where the connector is connected to the cable. There may be an impedance mismatch caused by bends or cuts in a cable. At microwave frequencies, the material properties as well as the dimensions of the cable or connector plays important role in determining the impedance match or mismatch. A high value of return-loss denotes better quality of the system under test (or device under test). For example, a cable with a return loss of 21 db is better than another similar cable with a return loss of 14 db, and so on.

Phase Response of the notch filter:

The phase response of a notch filter shows the greatest rate of change at the centre frequency. The rate of change becomes more rapid as the Q of the filter increases. The group delay of a notch filter is greatest at the centre frequency, and becomes longer as the Q of the filter increases.

EXPERIMENT SUB PARTS

CASE-STUDY PART 1:

Aim:

Designing and simulation of a notch filter at 3 GHz using Agilent’s ADSTM for the given design specifications.

Requirement:

Electrical performance:

Centre frequency: 3.0 GHz

Insertion loss: >25.0 dB

Input/output Impedance: 50 Ω

Substrate specifications:

Material type: 3M Cu-clad

Dielectric constant (εr): 2.17

Thickness (h): 0.794mm

Conductor thickness (t): 35um

Conductivity (σ): 5.84e+7 S/m

tanδ: 0.0009

MLIN, MLOC and MTEE are micro strip elements defined in ADSTM which is used to construct the circuit

Explanation:

We need to simulate and design a notch filter at 3 GHz here, using Agilent’s ADS. When the above specifications are used in ADS, the width of the microstrip lines is obtained as 2.42mm corresponding to 50 ohms transmission line using Line calc function.

The Line Calc function is also used to determine the effective dielectric constant (Keff) of 3M Cu-clad Substrate at 3.0GHZ from which the initial, length of the open circuit stub can be calculated.

εr = 2.1 Keff = 1.854 at 3.0GHZ (from line calc) , λ0 = 100 m (at 3.0 GHZ)

λg = λ0 /(Keff)1/2 = 100/(1.854)1/2 =73.44mm; λg/4 =18.36 mm

The initial design length of the open circuit stub is 18.354 mm.

Thus we obtain the following substrate specifications at Centre frequency: 3.0 GHz, Insertion loss: greater than 25.0 dB and Input/output Impedance:

Material type: 3M Cu-Clad, Dielectric constant (εr): 2.17, Thickness (h): 0.794m,

Conductor thickness (t): 35um, Conductivity (σ): 5.84e+7 S/m, tanᶿ=0.0009, l = 18.36mm W(Width of the micro strip lines)=2.42mm

From these specifications we obtain the plot of Insertion Loss Response(S21) indicating about 49.234 dB attenuation near 3 GHz which is shown in 8

To observe the effect of varying the length of the open circuit stub , the same procedure of simulation is repeated twice or thrice with different values of length of open circuit stub given as follows L1=20, L2=18.34, L3=16.As we can see in the 9 that as the length of open stub increases the frequency decreases. As the length of open stub must be λg/4 and so the 50Ω micro strip line is blocked and hence the signal is passed and if there is change in the length then the micro strip is not blocked hence the signal is blocked.

Analysis of the case study 1:

From the case study1, it proves that at wavelength λg/4 the open circuit at point S of the stub is transformed to short circuit and the signals passing along AB micro strip is blocked. Thus we design a filter at 3 GHz frequency.

When the wavelength is λg/4 the signal will see very low impedance to ground at point S and hence is short circuited. This signal will be absorbed from the signals applied at input A, which will manifest high attenuation in its insertion loss at 3GHz.All other signals remain unaffected, hence low insertion loss accept near 3GHz.

CASE-STUDY PART 2:

Aim:

Using the ADSTM Tuning facility, investigate the effect of varying the width of the stub filter. Determine the width of line which provides minimum out of band loss whilst maintaining the original filter specifications (i.e.>25 db at 3.0 GHz)

Requirement:

Electrical performance:

Centre frequency: 3.0 GHz

Insertion loss: >25.0 dB

Input/output Impedance: 50 Ω

Substrate specifications:

Material type: 3M Cu-clad

Dielectric constant (εr): 2.17

Thickness (h): 0.794mm

Conductor thickness (t): 35um

Conductivity (σ): 5.84e+7 S/m

tanδ: 0.0009

CS2 10: Circuit Diagram of Stub Notch filter obtained by ADS Simulation

Explanation:

When the width of the stub is 5mm and length is 18.8mm the response obtained is as shown below

Now we vary the width of the stub to investigate the effect. . In this process the width of the stub filter is changed at different values from w1=5mm, w2=2.5mm, w3=2mm, w4=1mm, w5=0.2mm as shown in 12. Here we also note that when varying the width of line, both the width of the stub line and corresponding width on the MTEE section must is varied.

After varying the width using tuning fork function of the ADS facility we obtain a response at 3GHz and width is noted as 0.2mm.The 13 shows the following.

Analysis of case study 2:

The width of the line determines its impedance. If the impedance is high thinner the line and viceversa.When the width of the i/o transmission line is equivalent to the width of the stub then Insertion loss is at 0Db and when width of the i/o transmission line is greater than the width of the stub then Insertion loss tends to 0Db.

In the above case thus we vary the width of the stub and transmission line and when centre frequency is 3 GHz and the width is 0.2mm the insertion loss is very low. Lower the insertion loss more is the signal transmitted.

CASE- STUDY PART 3:

Aim

To design a notch filter at centre frequency of 4.5GHZ and it should cancel the spurious signal and unwanted harmonics by at least 24db with minimum out of band loss with the specifications given below

Requirement:

Electrical specifications:

Centre frequency: 4.5 GHz

Insertion loss: >25.0 dB

Input/output Impedance: 50 Ω

Substrate specifications:

Material type: 3M Cu-clad

Dielectric constant (εr): 2.17

Thickness (h): 0.794mm

Conductor thickness (t): 35um

Conductivity (σ): 5.84e+7 S/m

tanδ: 0.0009

Explanation: In the responses shown below we have obtained the 24 dB difference by adjusting the frequency at 4.5 GHz. In CS3 14 the length and width are adjusted to obtain the particular response

Analysis of case study 3:

In case study 3 we understand the way of designing a notch filter to cancel the spurious signals generated by wireless communication systems.

CONCLUSION:

This case study helps us analyse the notch filter. The notch filter is designed and its basics and working are understood. The tool ADS proves very effective in this learning. To conclude, this experiment gives us a broader knowledge about transmission theory. The concept is deeply understood. In wireless communications the unwanted harmonics and spurious signals generated are cancelled by this notch filter enabling a better reception. Thus designing of such a notch filter is learnt.