ECE545 REPORT
Cemin Zhang
 5GHz LNA Design
 2.4GHz OSC
 VCO Design Using Varactor
 VCO With Buffer
 Mixer Single Balanced
I.
5GHz LNA Design
Noise Figure (Simulation)：
3.5
3.0
m2
freq=5.000GHz
m2=0.904
nf(2)
2.5
2.0
1.5
m2
1.0
0.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
freq, GHz
Comparison of [S] : Simulation(left) and Measurement(right)
20
dB(S(1,2))
dB(S(2,2))
dB(S(1,1))
dB(S(2,1))
10
0
-10
-20
-30
3.0
3.5
4.0
4.5
5.0
5.5
freq, GHz
6.0
6.5
7.0
6.5
7.0
II.
2.4GHz OSC
III.
2
0
1
ts(HB.Vout), V
dBm(HB.Vout)
100
-100
-200
-300
0
-1
-2
0
1
2
3
4
5
6
7
0
100
200
harmindex
300
400
500
600
700
800
900
time, psec
-40
harmindex
-60
pnfm
-80
m3
-100
-120
-140
-160
1E2
m3
indep(m3)=1.000E5
m3=-106.732
1E3
1E4
1E5
HB.freq
0
1
2
3
4
5
6
7
1E6
0.0000 Hz
2.408GHz
4.815GHz
7.223GHz
9.631GHz
12.04GHz
14.45GHz
16.85GHz
1E7
noisefreq, Hz
Simulation results show the Voltage output in Frequency and Time domain. The
OSC frequency is adjusted 2.408GHz, the simulated phase noise is
-107dBc@100kHz.
III.
VCO Design Using Varactor
Varactor is the units from MicroMetrics, SMT package. Sample units are at hand.
dBm(HB.Vout)
m8
0
m8
harmindex=1
m8=9.034
VR=1.000000
m9
m9
harmindex=2
m9=-46.341
VR=11.500000
3
2
ts(HB.Vout), V
100
-100
-200
1
0
-1
-2
-3
-300
0
1
2
3
4
5
6
0.0
7
0.2
0.4
-40
m1
-60
2.4
m12
pnfm
-80
2.2
m1
indep(m1)=9.000
m1=2.454E9
2.0
1.0
1.2
(b)
(a)
2.6
0.8
time, nsec
harmindex
HB.freq[1], GHz
0.6
m12
indep(m12)=1.000E4
m12=-94.498
VR=5.500000
-100
-120
-140
1.8
0
2
4
6
8
10
12
14
16
18
20
-160
1E2
1E3
1E4
1E5
HB.VR
noisefreq, Hz
(c)
(d)
1E6
(C) shows the frequency output response as the relationship with the tuning voltage: 1.9GHz –
2.5GHz, 1V-20V; (a) (b) (d) show the output in frequency and time domain and the phase noise
performance -95dBc@10kHz
1E7
VCO Measurement Results
(a)
(b)
(c)
The plot (a) and (b) show that the frequency range of VCO is from 2.2GHz to 2.33GHz, with
tuning voltage from 3V(2.2GHz) to 12V(2.33GHz).
Plot (c) shows the detail of the VCO output with tuning voltage at 5V, and the oscillation
frequency is 2.265GHz.
Output power is low here because:
When I increase the bias voltage, some kind of low frequency harmonics will appear. While I
keep increasing the bias, the transistor will be burned.
I will try to figure out how to solve this problem.
VI.
VCO With Buffer
Buffer
Performance is similar with Part III design, but have good isolation between load and VCO,
and has adjustable output power.
V.
Single Balanced Mixer Design
RF: 2.45 GHz (-10dBm)
LO: 2.35 GHz
IF:
100
MHz
m4
-7
0
m5
indep(m5)=4.000
m5=-44.412
-10
-9
RF_LO_Iso
ConvGain
-8
m4
indep(m4)=4.000
m4=-7.524
-10
-11
-20
-30
m5
-40
-12
-13
-50
-6
-4
-2
0
2
4
6
8
10
-6
-4
-2
Pow er_LO
0
2
4
6
8
Pow er_LO
Conversion Gain VS. PLO : ILmin @ PLO=4dBm
Isolation (RF-LO port) VS LO Power
m1
-7
ConvGain
-8
-9
-10
-26.5
-26.6
-26.7
-11
2.0E9
-26.4
RF_LO_Iso
m1
indep(m1)=2.360E9
m1=-7.018
-26.8
2.2E9
2.4E9
2.6E9
2.8E9
3.0E9
Sig
Conversion Gain VS. RF : ILmin @ RF=2.36GHz
2.0E9
2.2E9
2.4E9
2.6E9
2.8E9
3.0E9
Sig
Isolation (RF-LO port) VS. RF Freqency
10
10
m1
Power_RF=3.000
m1=-4.659
Pif_dBm
Pif_ext
5
0
m1
m2
-5
-10
m2
Power_RF=3.000
m2=-5.628
-15
-20
-10
-5
0
5
10
15
Power_RF
P1dB point when PRF=3dBm, so RF power input to the Mixer should be < 3dBm
Mixer Measurement Results
Based on the simulation results, I choose the value below during the measurement:
RF: 2.45GHz, -10dBm
LO: 2.35GHz, 4dBm
(a)
(c)
(b)
(d)
 Figure (a) shows IF signal, with the RF and LO leakage at IF output port. The conversion
gain is -7.7dB, which is very close to the simulation result(-7.56dB)
 Figure (b) shows the detail of IF output signal
 Figure (c) shows LO and RF leakage at IF port, the LO at IF port is -36dBm, which means
the LO to IF isolation is -40dB
 Figure (d) shows the IF harmonics, the 2nd harmonics (200MHz) is -50dBm