Olaniyi_Jinadu_CAD_Lab_2_Full

DREXEL UNIVERSITY

CAD Experiment II: RF Filters Objective: To design, simulate, and realize LP and BP micro-strip transmission

line filters

Olaniyi Q. Jinadu

2/9/2015

CAD Experiment II: RF Filters Objective: To design, simulate, and realize LP and BP microstrip transmission line filters A. Simulations Set-up A: 1. Design a fourth-order maximally flat filter low-pass filter at corner

frequency of 950 MHz 2. Show schematic and physical realization of this filter using lumped

elements. 3. Simulate insertion loss, return loss, and group delay up to 2 GHz. B. Simulations Set-up B: 1. Convert the lumped element circuit design of the low-pass filter to a

commensurate transmission line using Richard’s transform and Kuroda identities.

2. Simulate insertion loss, return loss, and group delay up to 2 GHz using ideal model of transmission lines.

3. Realize this filter in a microstrip using a h=59 mils, t=0.005 mm, r= 4.0, and loss tangent of 0.015. (Tune the line lengths to get the corner frequency at 950 MHz.)

4. Show physical layout of this filter and prepare the layout to be delivered to TA, Yaaqoub Malallah by February 13.

5. Simulate insertion loss, return loss, and group delay up to the frequency range of 2.5 GHz. Compare these results with the lumped element, ideal transmission line, and the true physical characteristic of this filter (i.e., with dielectric and Ohmic losses) including the step discontinuities.

C. Simulations Set-up C: 1. Design a third order Tchebychev bandpass filter with a ripple of 0.5 dB, at

center frequency of 850 MHz and 10% bandwidth. 2. Show schematic and physical realization of this filter using lumped

elements. 3. Simulate insertion loss, return loss, and group delay over the frequency

range of 750-950 MHz. D. Simulations Set-up D: 1. Convert the lumped element circuit design of the bandpass filter to a

microstrip line realization using a coupled-line structure. Realize a physical realization of this coupler using a coupled microstrip line on the same FR4 substrate. (Tune the line lengths to get the center frequency at 950 MHz.)

2. Show physical layout of this filter and prepare the layout to be delivered to TA, Yaaqoub Malallah by February 13.

3. Simulate insertion loss, return loss, and group delay over the frequency range of 750-950 MHz. Compare these results with the lumped element and the case that the microstrip lines are loss-less (i.e., no dielectric and Ohmic losses).

Questions and Homework Problems: 1) Describe the mathematical function describing the attenuation of the LPF

filter as a function of frequency. How much is the expected loss at 1.3GHz? How much was achieved? What is the reason for this difference?

2) Justify the resultant S11, S2l, and group delay for the lumped element realization. (Hint: use the poles and zeroes of the Butterworth filter.)

3) Discuss the differences in group delay as a function of frequency. What are the sources for non-flat performance? Would this filter appropriate for 2.5Gb/s data links?

4) Describe the mathematical function describing the attenuation of the BPF filter as a function of frequency. How much is the expected loss at 650MHz and 1.3GHz?

5) Justify the resultant Sll, S2l, and group delay for the lumped element realization. (Hint: use the poles and zeroes of the Tchebychev filter.)

SIMULATION SET-UP A: The goal is to design a fourth-order maximally flat filter low pass filter at corner frequency of 950 MHz similar to the fifth order maximally flat filter seen below;

Using the table below, the values for the capacitors and inductors can be gotten;

Since N = 4; g1 = 0.7654; g2 = 1.8478; g3 = 1.8478; g4 = 0.7654; g5 = 1; Zo = 50;

𝐶1 = 𝑔1𝑍𝑜𝑤𝑐

= 2.565𝑝𝐹

𝐿2 = 𝑔2𝑍𝑜𝑤𝑐

= 15.48𝑛𝐻

𝐶3 = 𝑔3𝑍𝑜𝑤𝑐

= 6.2𝑝𝐹

𝐿4 = 𝑔4𝑍𝑜𝑤𝑐

= 6.4𝑛𝐻

S11 & S21 RL & IL GROUP DELAY

SIMULATIONS SET-UP B: The lumped element circuit representation of the fourth order maximally flat filter above was transformed using the Richard’s transform and Kuroda identities as shown below;

The above transformation is represented using ideal transmission lines below;

The resulting parameters of this circuit are given as;

S11 & S21 RL & IL GROUP DELAY

The physical realization of this circuit is given as;

Note the circuit was tuned to achieve a corner frequency of 950 MHz with 3 dB return loss.

S11 & S21 RL & IL GROUP DELAY

The physical layout using FR4 substrate will be;

The comparison of lumped circuit, ideal transmission line and micro-strip realization is given below;

LUMPED ELEMENTS IDEAL TRANSMISSION LINE

MICRO-STRIP REALIZATION

RETURN LOSS 3.042 3.010 4.118 INSERTION LOSS 2.979 3.010 3.049

GROUP DELAY (S21) in ns 0.62 1.37 1.32

SIMULATIONS SET-UP C: The goal is to design a third-order Tchebychev filter at center frequency of 850 MHz with ripple of 0.5 dB and 10% bandwidth. The design can be seen below;

The resulting parameters of the above circuits can be seen below;

S-PARAMETERS RETURN LOSS & INSERTION LOSS GROUP DELAY

SIMULATION SET-UP D: The lumped circuit design of the band pass filter was converted to a micro-strip realization using a coupled line structure with FR4 substrate. The design of the coupled line couplers can be seen below;

The simulated results of this micro-strip line coupled line realization can be seen below;

S-PARAMETERS RETURN LOSS & INSERTION LOSS GROUP DELAY

Questions and Homework Prob]ems:1) Describe the mathematical function describing the attenuation of the LPE

filter as a function of frequency. How much is the expected loss at 1.3GHz?How much was achieved? Vihat is the reason for this difference?

m1.0MHz

s(2,1))=s(1,1)= 4.1',|8

m1 m2

m21.300GH2

1))=-19.2091,'l))=-9.523

-_r:-Naadl(IlEE

-20

-40

€0

€0

3.01.00.0 1.5 2.O 2.5

freq, GHz

8L($D = lt"&sJg

= r.68

I r_ a- o. 3.+O+ JB(rB)

I

'to=k3)-- L)c

(..--L6

<-!13',ae%

I

13 - AGB6

U t + ,^.8 z 1'3Qt\z-

?so (ttz

L6'z tt 13.a6b6 a l3-aM

\S^,\^ =

T* €.Q"'\.\ \L \\-a<J{3 *L\.T, -,0**J tL rs R'ao1JB .L

\o= ''\S

tt"-t{^oFe te..=oo S g^.o{t. \.,qr &" t- S*.**^*\ { tt" '^"*$ifl [* -

2) Justify therealization.

resultant Stt, S2L,(Hint: use the poles

and group delayand zeroes of the

for the lumped elementButterworth filter. )

m3=950.0MH2

,11)Y-2.

m4l.000GHz1,1)=4.911

1

NF

<DOdt drEt,

0,0 0.5 '1.0 1.5 2.O

treq, Gl-lz

{S^ *-A* A$K-''-{,/. ([*;

?-\.= =.Sro : -Tgt{ i :e=Caf$e[-. r, z--,) &.n ?,t

; Z*. z- S2p =6C=' *)

+ ^r cos EIrJz

1 \c"< flE,

Vz\

3uz

Sr, = -B^3f.t = -=..\,^.

S-. Szo

S - S.u-

S*

f c= -srDV-.\

9r J"* ,A \** q'\.s 6','e -j.l' .fr,-

"A zs{b *-\ f*= ' % A"e'^ g^ is o) g,

d"= G.-- .-^^^g\.= f-\ 'S^

i< +- -r\

^B

1La

)

([""

arefor

?) Discuss thethe sources2.SGb/ s data

differences infor non-fIatlinks ?

group delayperformance?

as a functionWouId this

of frequency. Whatfilter appropriate

50MHz

,1F6.538E-10

6'g€-lo

t-{q&uB

-6-'ggns E

fk'<- ? !-t.13 -Q g,S.,-.*$^ \ cu-g.\^, Gk=

a:.',-5 \^a "sA GPSk- ^d.^t$.16W, e-En1g1 tf=

fltr /t.\

;€ - \Stps

L"--FBk Jrga^e \e o,aB36n* G' .-R'o. S- iE* =+Ue*

T

l-a-86't\a

o'f 'ns

LI

I

I

I

I

I

I

4) Describe the mathemat.ical functionfilter as a function of frequency.and 1 .3GHz?

describing theHow much is the

attenuation of the BPEexpected loss at 650MHz

m3i50.0MHz1,1))=-9.092

dB(S(2.1))=-46.791

=1.999GH21.1)].,4.371

1 972

C.,lFou)trt d)EE

1.2 1-4

fieq, GHz

I.(*)-Q.'-3-5L* z- \ J f. rmf)

uc--> I / 1^)A ( **=o'Ix, - g'.1B "\rar

z- R.aX dta-,^

. o'St?S

.^2 - 6Wr< (r

u." z aG(c

L*e

Nr-3

C"\Ae"

; *rtk

6 Bor"]ilc&6\{z

fQo" =-l]\A5Bo \

?-))Lo\<) -/ (g

"r1n",.^

L*

\S^[

1L =

\

LA-

\U \*)

-nL

IL

,,4\S,.,\

\) .e- -- 3''t<

= (115-al - E31E

U[S.fr.. \^? t- tT 'at

L4 ? ?<s63(ol'71't9.

Z +1JBQ- gcsB

ee

ft{tn*+* 4 &.1 c-.U

It f \.G) t^"[ € e