Accurate Mixer Measurements Using Multi-tone
X-parameter - Models
Mihai Marcu
Radoslaw M. Biernacki
Agilent Technologies, Inc.
© Copyright Agilent Technologies 2010
Page 1
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
What We Will Talk About
This presentation focuses on the applicability of X-parameter*
technology
h l
to mixers.
i
System-level models need to protect the manufacturer IP and to
accurately predict the HW nonlinear behavior in the simulation
environment.
X-parameter models have this unique capability.
Their applicability to mixers is demonstrated and the accuracy is
shown based on some fundamental measurements, such as
compression and LO-starvation.
* "X-parameters" is a registered trademark of Agilent Technologies.
The X-parameter format and underlying equations are open and documented.
For more information visit http://www.agilent.com/find/eesof-x-parameters-info
© Copyright Agilent Technologies 2010
Page 2
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
Mixer X-Parameter Models
Accuracy of X-parameter models for mixer is evaluated for
severall practical
i l scenarios:
i
– multi-tone
– multi-port
– swept RF level
– swept LO level
– swept RF frequency
– mixer sub
sub-system
system = mixer + Band Pass Filter (BPF)
© Copyright Agilent Technologies 2010
Page 3
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
Mixer Circuit
• standard Gilbert-cell mixer
R
R1 0
R=40 0 Oh m
R
R11
R=8 0 0 Oh m
BJ T_ NPN
BJ T9
M o d el =BJ TM 4
Are a =1
Re gi on =
Tri s e =
BJ T_ NPN
BJ T1 0
M o de l =BJ TM 3
Are a=1
Re gi o n =
Tri s e =
R
R8
R=4 00 Ohm
R
R12
R=7 0 0 Oh m
R
R7
R=1 00 Ohm
Port
L O_ PORT
Num =4
C
C4
C=1 u F
L
L2
L =0 .5 n H
R=
BJ T_NPN
BJ T7
M od el =BJ TM 1
Area =1
Re g i on =
Tri s e=
R
R3
R=5 0 Oh m
BJ T_ NPN
BJ T5
M o d el =BJ TM 1
Are a =1
Re gi on =
Tri s e =
C
C3
C=1.0 uF
C
C1
C=1 u F
L
L1
L=0.5 nH
R=
BJ T_NPN
BJ T4
M od e l =BJ TM 2
Area =1
Reg i o n=
Tri s e=
Is c =1 .1 61 6 00 00 E-12
C4=
Nc =2 .00 E+0 0
Cbo =
Gbo =
Vbo =
Rb=6 .12 4 58 79 1 E+01
Rbm =1.91 7 85 71 4E+01
Re=3 .57 1 42 85 7 E+00
Rc =6 .96 2 32 09 6 E+01
Rc v =
Rc m =
Dop e =
Cex =
Cc o =
Im ax =1.0 A
Im el t=
Cj e =9.67 68 00 1 7E-1 4
Vj e =8.49 99 99 8 7E-0 1
M j e=3 .99 9 99 99 4E-0 1
Cj c =6.05 00 00 1 1E-1 4
Vj c =7.49 99 99 8 9E-0 1
M j c =4 .99 9 99 99 3E-0 1
Xc j c =2 .5 25 61 9 80 E-01
Cj s =8.68 00 00 1 5E-1 4
Vj s =6.99 99 99 9 0E-0 1
M j s =4 .99 9 99 99 3E-0 1
Fc =7 .9 99 9 99 88 E-01
Xtf=3 .3 54 7 22 24 E+0 0
Tf=8 .9 12 92 80 3 E-1 2
Itf=1 .0 8 50 00 02 E-02 Tri s e =
Eg =1 .1 1
Ptf=1.80 E+0 1
Tr=1 .6 0 00 00 05 E-09 Xtb=2.20 E+00
Xti =8 .0 0E+00
Kf=0 .0 0 E+00
Af=1 .0 0 E+00
Ab=1.00
Fb=1.00
Rbn oi =
Is s =0 A
Ns =1.00
Nk =0.50
Ffe=1 .0 0
RbM o de l =M DS
App ro x q b=y es
Tno m =2 5.
BJ T_ M od e l
BJ TM 2
NPN=y es
PNP=n o
Is =2 .1 48 22 9 3E-1 6
Bf=1 .3 00 35 6 47 E+02
Nf=1 .0 3E+0 0
Va f=2.50 E+0 1
Ik f=1 .0 85 0 00 02 E-02
Is e=7.56 17 3 43 4E-1 3
C2 =
Ne =2 .0 0E+00
Br=5 .1 22 95 0 82 E+00
Nr=1 .0 0E+0 0
Ik r=5 .6 00 0 00 01 3 E-0 2
Ke =
Kc =
© Copyright Agilent Technologies 2010
Page 4
BJ T_ NPN
BJ T2
M o de l =BJ TM 3
Are a=1
Re gi o n =
Tri s e =
Im a x =1 .0 A
Im e l t=
Cj e =1 .9 35 3 60 03 E-13
Vj e =8 .4 99 9 99 87 E-01
M j e =3.9 9 99 99 9 4E-0 1
Cj c =1 .0 45 0 00 02 E-13
Vj c =7 .4 99 9 99 89 E-01
M j c =4.9 9 99 99 9 3E-0 1
Xc j c =2.92 4 40 18 7E-0 1
Cj s =1 .0 92 0 00 02 E-13
Vj s =6 .9 99 9 99 90 E-01
M j s =4.9 9 99 99 9 3E-0 1
Fc =7 .99 9 99 98 8 E-0 1
Xtf=3 .35 4 72 22 4 E+00
Tf=8.91 29 2 80 3E-1 2
Im a x =1 .0 A
Im e l t=
Cj e =3 .8 70 72 0 07 E-13
Vj e =8 .4 99 99 9 87 E-01
M j e =3.9 9 99 99 94 E-0 1
Cj c =1 .9 25 00 0 03 E-13
Vj c =7
7 .4
4 99 99 9 89 E
E-01
01
M j c =4.9 9 99 99 93 E-0 1
Xc j c =3.17 50 6 48 9E-0 1
Cj s =1 .5 40 00 0 03 E-13
Vj s =6 .9 99 99 9 90 E-01
M j s =4.9 9 99 99 93 E-0 1
Fc =7 .99 9 99 98 8E-0 1
Xtf=3 .35 4 72 22 4E+00
Tf=8.91 29 28 0 3E-1 2
Itf=4 .3 40 00 01 0 E-02
Ptf=1 .8 0E+01
Tr=1 .6 00 00 00 5 E-09
Kf=0 .0 0E+0 0
Af=1 .0 0E+0 0
Ab =1 .0 0
Fb =1
1 .0
00
Rb no i =
Is s =0 A
Ns =1 .0 0
Nk =0 .5 0
Ffe=1.00
Rb M od el =M DS
Ap prox qb =y e s
Tn om =25 .
Tri s e =
Eg =1 .1 1
Xtb =2.2 0 E+00
Xti =8 .0 0E+0 0
C
C5
C=1 uF
L
L3
L=0.5 n H
R=
Port
IF_ po rt
Num =3
R
R13
R=5 0 0 Oh m
BJ T_M o de l
BJ TM 3
NPN=y es
PNP=n o
Is =2 .0 43 1 59 24 E-16
Bf=1 .6 61 7 89 34 E+0 2
Nf=1 .0 3E+00
Vaf=2 .50 E+0 1
Ik f=1.16 25 00 0 3E-0 2
Is e=6 .09 1 23 88 8 E-1 3
C2=
Ne=2.00 E+00
Br=5 .1 22 9 50 82 E+0 0
Nr=1 .0 0E+00
Ik r=6.00 00 00 1 3E-0 2
Ke=
Kc =
Is c =2 .3 18 40 0 01 E-12
C4=
Nc =2.00 E+0 0
Cbo =
Gbo =
Vbo =
Rb=1.17 1 66 37 9E+02
Rbm =2.70 22 1 48 4E+0 1
Re=1.66 6 66 66 7E+00
Rc =3.58 0 72 51 2E+01
Rc v =
Rc m =
Dop e=
Cex =
Cc o =
Im a x =1 .0 A
Im e l t=
Cj e =1 .5 61 60 0 03 E-13
Vj e =8 .4 99 99 9 87 E-01
M j e =3.9 9 99 99 94 E-0 1
Cj c =1 .2 07 50 0 02 E-13
Vj c =7 .4 99 99 9 89 E-01
M j c =4.9 9 99 99 93 E-0 1
Xc j c =1.95 44 5 13 2E-0 1
Cj s =1 .1 89 00 0 02 E-13
Vj s =6 .9 99 99 9 90 E-01
M j s =4.9 9 99 99 93 E-0 1
Fc =7 .99 9 99 98 8E-0 1
Xtf=3 .16 1 17 21 5E+00
Tf=7.97 47 77 1 2E-1 2
Itf=2 .3 25 00 00 5 E-02
Ptf=1 .8 0E+01
Tr=1 .6 00 00 00 5 E-09
Kf=0 .0 0E+0 0
Af=1 .0 0E+0 0
Ab =1 .0 0
Fb =1 .0 0
Rb no i =
Is s =0 A
Ns =1 .0 0
Nk =0 .5 0
Ffe=1.00
Rb M od el =M DS
Ap prox qb =y e s
Tn om =25 .
Tri s e =
Eg =1 .1 1
Xtb =2.2 0 E+00
Xti =8 .0 0E+0 0
BJ T_ NPN
BJ T1 2
M o de l =BJ TM 3
Are a=1
Re gi o n =
Tri s e =
BJ T_NPN
BJ T1
M od e l =BJ TM 3
Area =1
Reg i o n=
Tri s e=
R
R4
R=17 0 Oh m
Is c =2 .0 06 40 0 01 E-12
C4=
Nc =2 .00 E+0 0
Cbo =
Gbo =
Vbo =
Rb=3 .16 6 46 06 2E+01
Rbm =1.06 30 9 52 4E+01
Re=1 .78 5 71 42 9E+00
Rc =3 .75 7 04 75 6E+01
Rc v =
Rc m =
Dop e =
Cex =
Cc o =
Is c =3 .6 96 00 0 01 E-12
C4=
Nc =2.00 E+0 0
Cbo =
Gbo =
Vbo =
Rb 1 61 7 95 25 3E
Rb=1.61
3E+01
01
Rbm =5.66 26 9 84 1E+0 0
Re=8.92 8 57 13 0E-0 1
Rc =2.27 4 41 08 7E+01
Rc v =
Rc m =
Dop e=
Cex =
Cc o =
BJ T_ NPN
BJ T1 1
M o de l =BJ TM 3
Are a=1
Re gi o n =
Tri s e =
C
C2
C=1 u F
R
R6
R=20 Ohm
BJ T_ M od el
BJ TM 1
NPN=y e s
PNP=no
Is =1.07 41 1 14 7E-1 6
Bf=1.30 03 5 64 7E+0 2
Nf=1.03 E+0 0
Va f=2 .5 0E+01
Ik f=5 .4 24 99 97 4 E-0 3
Is e =3 .7 80 8 67 17 E-13
C2 =
Ne =2 .0 0E+0 0
Br=5.12 29 5 08 2E+0 0
Nr=1.00 E+0 0
Ik r=2 .8 00 00 00 6 E-0 2
Ke =
Kc =
BJ T_NPN
BJ T8
M od el =BJ TM 1
Area =1
Re g i on =
Tri s e=
R
R2
R=50 Ohm
BJ T_ NPN
BJ T3
M o de l =BJ TM 2
Are a=1
Re gi o n =
Tri s e =
BJ T_M o de l
BJ TM 4
NPN=y es
PNP=n o
Is =4 .2 96 4 45 87 E-16
Bf=1 .3 00 3 56 47 E+0 2
Nf=1 .0 3E+00
Vaf=2 .50 E+0 1
Ik f=2.17
f 2 17 00 00 0 5E-0
5E 0 2
Is e=1 .51 2 34 68 5 E-1 2
C2=
Ne=2.00 E+00
Br=5 .1 22 9 50 82 E+0 0
Nr=1 .0 0E+00
Ik r=1.11 99 99 9 8E-0 1
Ke=
Kc =
R
R9
R=10 0 Oh m
BJ T_NPN
BJ T6
M o de l =BJ TM 1
Are a=1
Reg i o n=
Tri s e=
R
R1
R=50 Ohm
Port
RF_ po rt
Num =1
Port
Bi as _ po rt
Num =2
BJ T_ NPN
BJ T1 4
M o d el =BJ TM 5
Are a =1
Re gi on =
Tri s e =
R
R5
R=1 70 Ohm
Itf=2 .1 70 00 0 05 E-02
Ptf=1 .8 0E+01
Tr=1 .6 00 00 0 05 E-09
Kf=0 .0 0E+0 0
Af=1 .0 0E+0 0
Kb =
Ab =1 .0 0
Fb =1 .0 0
Rb no i =
Is s =0 A
Ns =1 .0 0
Nk =0 .5 0
Ffe=1.00
Rb M od el =M DS
Ap prox qb =y e s
R
R1 7
R=45 Ohm
BJ T_ NPN
BJ T1 3
M o de l =BJ TM 3
Are a=1
Reg i o n=
Tri s e =
R
R15
R=1 0 00 Ohm
R
R14
R=2 0 0 Oh m
BJ T_M o de l
BJ TM 5
NPN=y es
PNP=n o
Is =4 .0 86 3 18 47 E-16
Bf=1 .6 61 7 89 34 E+0 2
Nf=1 .0 3E+00
Vaf=2 .50 E+0 1
Ik f=2.32 50 00 0 5E-0 2
Is e=1 .21 8 24 77 6 E-1 2
C2=
Ne=2.00 E+00
Br=5 .1 22 9 50 82 E+0 0
Nr=1 .0 0E+00
Ik r=1.19 99 99 9 98 E-01
Ke=
Kc =
Is c =4.0 8 48 00 02 E-12
C4 =
Nc =2 .0 0E+0 0
Cb o=
Gb o=
Vb o=
Rb =5 .9 17 84 2 77 E+01
Rb m =1 .4 10 63 12 3 E+01
Re =8 .3 33 33 3 21 E-01
Rc =2 .2 01 44 3 69 E+01
Rc v =
Rc m =
Do pe =
Ce x =
Cc o=
Im ax =1 .0 A
Im el t=
Cj e =3.1 2 32 00 0 5E-1 3
Vj e =8.4 9 99 99 8 7E-0 1
M j e=3.99 99 9 99 4E-0 1
Cj c =2.1 2 75 00 0 4E-1 3
Vj c =7.4 9 99 99 8 9E-0 1
M j c =4.99 99 9 99 3E-0 1
Xc j c =2 .2 18 56 6 36 E-01
Cj s =1.6 5 30 00 0 3E-1 3
Vj s =6.9 9 99 99 9 0E-0 1
M j s =4.99 99 9 99 3E-0 1
Fc =7 .9 99 99 9 88 E-01
Xtf=3 .1 61 17 2 15 E+0 0
Tf=7 .97 4 77 71 2 E-1 2
Itf=4 .6 50 0 00 10 E-02
Ptf=1.80 E+01
Tr=1 .6 00 0 00 05 E-09
Kf=0 .0 0E+00
Af=1 .0 0E+00
Ab=1.00
Fb=1.00
Rbn oi =
Is s =0 A
Ns =1.00
Nk =0.50
Ffe=1 .00
RbM o de l =M DS
App ro x q b=y e s
Tno m =2 5.
Tri s e=
Eg =1.1 1
Xtb=2.20 E+0 0
Xti =8.0 0 E+00
Tn om =25 .0
Tri s e =
Eg =1 .1 1
Xtb =2.20 E+00
Xti =8 .0 0E+00
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
Performance to Be Measured
• spectral content – magnitude and phase
• compression curve
• LO starvation curve
• leakage: LO-to-RF, LO-to-IF, RF-to-IF
• frequency response
© Copyright Agilent Technologies 2010
Page 5
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
X-Parameter Model Extraction
X-parameter model is extracted
using multiport
multiport, multi
multi-tone
tone ADS
ADSsimulation.
DC
XP_Bias
PORT4
Num=4
DC_mode=Voltage
DC_value=5 V
Bias_port
• RFfreq = 2 GHz
• LOfreq = 1.75 GHz
RF_port
XP_Source
PORT1
IF_port
LO_port
x_2_GilCellMix
X1
XP_Load
PORT2
XP_Source
PORT3
• IFfreq = 250 MHz
• RF-power sweep: -30 dBm to 0 dBm
• LO-power
LO power sweep: -25
25 dB to 0 dBm
• power supply – one point: 5 V
X-Parameters
X_Param
XP1
MaxOrder=5
Freq[1]=1.75 GHz
Freq[2]=2 GHz
Order[1]=5
Order[2]=5
XParamMaxOrder=5
© Copyright Agilent Technologies 2010
Page 6
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
Spectral Content
• measurement made at single LO and RF power levels
• magnitude and phase of output signal at all frequencies
monitored at the output port
© Copyright Agilent Technologies 2010
Page 7
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
Spectral Content
Measurement Setup
• RFpwr = -50 dBm
• LOpwr = -5 dBm
HARMONIC BALANCE
V_DC
SRC1
Vdc=5 V
HarmonicBalance
HB1
MaxOrder=5
MaxOrder
5
Freq[1]=1.75 GHz
Freq[2]=2 GHz
Order[1]=5
Order[2]=5
Bias_port
V_DC
SRC2
Vdc=5 V
X4P
XNP1
File="a0_Mixer.ds"
4
RF_port
P_1Tone
PORT1
Num=1
Z=50 Ohm
P=polar(dbmtow(-50),0)
Freq=2 GHz
IF_port
LO_port
© Copyright Agilent Technologies 2010
Page 8
x_2_GilCellMix
X1
P_1Tone
PORT2
Num=2
Z=50 Ohm
P=polar(dbmtow(-5),0)
Freq=1.75 GHz
v out_ckt
out ckt
1
R
R1
R=50 Ohm
P_1Tone
PORT3
Num=3
Z=50 Ohm
P=polar(dbmtow(-50),0)
Freq=2 GHz
v out_mdl
out mdl
2
3
Re f
P_1Tone
PORT4
Num=4
Z=50 Ohm
P=polar(dbmtow(-5),0)
Freq=1.75 GHz
R
R2
R=50 Ohm
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
Spectral Content
Measurement Results - Magnitude
• RFpwr = -50 dBm
m4
indep(m4)= 3.000E8
plot_vs(dBm(vout_mdl), freq+(0.05 GHz))=-38.914
• LOpwr = -5 dBm
m3
indep(m3)= 2.500E8
plot_vs(dBm(vout_ckt), freq)=-38.914
-20
m3
m4
-40
dBm(vou
ut_mdl)
dBm(vout_ckt)
-60
Small display offset
-80
100
-100
-120
-140
-160
160
-180
1.1E10
1.0E10
9.0E9
8.0E9
7.0E9
6.0E9
5.0E9
4.0E9
3.0E9
2.0E9
1.0E9
0.0
Ckt
Xpar
p
freq Hz
freq,
freq+(0.05 GHz)
• intentional display
p y offset of 50 MHz – to distinguish
g
the traces
© Copyright Agilent Technologies 2010
Page 9
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
Spectral Content
Measurement Results - Phase
• RFpwr = -50 dBm
m2
indep(m2)= 3.000E8
plot vs(phase(vout mdl) freq+(0
plot_vs(phase(vout_mdl),
freq+(0.05
05 GHz))=-50
GHz))=-50.105
105
• LOpwr = -5 dBm
m1
indep(m1)= 2.500E8
plot_vs(phase(vout_ckt), freq)=-50.105
200
150
phase(vout_mdl)
phase(vo
out_ckt)
100
Small display offset
50
0
-50
m1
m2
-100
-150
150
-200
1.1E10
1.0E10
9.0E9
8.0E9
7.0E9
6.0E9
5.0E9
4.0E9
3.0E9
2.0E9
1.0E9
0.0
Ckt
Xpar
p
freq Hz
freq,
freq+(0.05 GHz)
• intentional display
p y offset of 50 MHz – to distinguish
g
the traces
© Copyright Agilent Technologies 2010
Page 10
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
Compression Curve
• compression curve measured by maintaining a constant LO
power and
d sweeping
i th
the RF power llevell
• magnitude and phase of output signal at the IF frequency
monitored at the output port
© Copyright Agilent Technologies 2010
Page 11
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
Compression Curve
Measurement Setup
• RFpwr = -50 dBm to 0 dBm
• LOpwr = -5 dBm
V_DC
SRC1
Vdc=5 V
V_DC
SRC2
Vdc=5 V
X4P
XNP1
File="b0_Mixer_Compression.ds"
Bi a s _ p o rt
RF_ p o rt
P_1Tone
PORT1
Num=1
Z=50 Ohm
P=polar(dbmtow(RFpwr_dBm),0)
Freq=2 GHz
LO_ p o rt
R
R1
R=50 Ohm
P_1T one
PORT 2
Num=2
Z=50 Ohm
P=polar(dbmtow(-5),0)
F
Freq=1.75
1 75 GH
GHz
HarmonicBalance
HB1
M Od 5
MaxOrder=5
Freq[1]=1.75 GHz
Freq[2]=2 GHz
Order[1]=5
Order[2]=5
Page 12
1
x_2_GilCellMix
X1
HARMONIC BALANCE
© Copyright Agilent Technologies 2010
4
vout_ckt
o t ckt
IF_ p o rt
Var
Eqn
VAR
VAR1
RFpwr_dBm=-20 _dBm
P_1Tone
PORT 3
Num=3
Z=50 Ohm
P=polar(dbmtow(RFpwr_dBm),0)
Freq=2 GHz
2
3
vout_mdl
o t mdl
Ref
P_1T one
PORT 4
Num=4
Z=50 Ohm
P=polar(dbmtow(-5),0)
F
Freq=1.75
1 75 GHz
GH
R
R2
R=50 Ohm
PARAMETER SWEEP
ParamSweep
Sweep1
S
SweepVar="RFpwr_dBm"
V "RF
dB "
SimInstanceName[1]="HB1"
SimInstanceName[2]=
SimInstanceName[3]=
SimInstanceName[4]=
SimInstanceName[5]=
SimInstanceName[6]=
Start=-50
Stop=0
Step=1
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
Compression Curve
10
10
0
0
dBm(vout_m
mdl[::,1])+1
dBm(vout_
_ckt[::,1])
dBm(vout_m
mdl[::,1])
dBm(vout_
_ckt[::,1])
Measurement Results
• small intentional display offset – to distinguish traces
Mag
-10
-20
-30
-50
-45
-40
-35
No display offset
-30
-25
-20
-15
-10
-5
-20
-30
-50
0
-45
-40
-35
Small display offset
RFpwr_dBm
-30
-25
-20
-15
-10
-5
0
-15
-10
-5
0
RFpwr_dBm
-44
-46
Phase
-48
-50
Ckt
Xpar
-52
phase(v
vout_mdl[::,1])+0.2
phas
se(vout_ckt[::,1])
-44
phase
e(vout_mdl[::,1])
phas
se(vout_ckt[::,1])
-10
-40
-40
-46
Phase
-48
-50
-52
-54
-54
-50
-45
-40
-35
-30
-25
-20
RFpwr_dBm
© Copyright Agilent Technologies 2010
Page 13
Mag
-15
-10
-5
0
-50
-45
-40
-35
-30
-25
-20
RFpwr dBm
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
LO Starvation Curve
• starvation curve measured by maintaining a constant RF
power and
d sweeping
i th
the LO power llevell
• magnitude and phase of output signal at the IF frequency
monitored at the output port
© Copyright Agilent Technologies 2010
Page 14
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
LO Starvation Curve
Measurement Setup
• RFpwr = -50 dBm
• LOpwr = -15 dBm to 0 dBm
V_DC
SRC1
Vdc=5 V
V_DC
SRC2
Vdc=5 V
X4P
XNP1
File="c0_Mixer_Starvation.ds"
Bi as _ po rt
4
vin ckt
vin_ckt
RF_ p o rt
P_1Tone
PORT1
Num=1
Z=50 Ohm
P=polar(dbmtow(-50),0)
Freq=2 GHz
vout ckt
vout_ckt
IF_p o rt
L O_ p ort
x_2_GilCellMix
X1
HarmonicBalance
HB1
MaxOrder=5
Freq[1]=1.75 GHz
Freq[2]=2 GHz
Order[1]=5
Order[2]=5
© Copyright Agilent Technologies 2010
Page 15
1
2
3
P_1Tone
PORT2
Num=2
Z=50 Ohm
P=polar(dbmtow(LOpwr_dBm),0)
Freq=1 75 GHz
Freq=1.75
HARMONIC BALANCE
vin mdl
vin_mdl
Var
Eqn
R
R1
R=50 Ohm
VAR
VAR1
LOpwr_dBm=-20 _dBm
vout mdl
vout_mdl
Re f
P_1Tone
PORT3
Num=3
Z=50 Ohm
P=polar(dbmtow(-50),0)
Freq=2 GHz
P_1Tone
PORT4
Num=4
Z=50 Ohm
P=polar(dbmtow(LOpwr_dBm),0)
Freq=1 75 GHz
Freq=1.75
R
R2
R=50 Ohm
PARAMETER SWEEP
ParamSweep
Sweep1
SweepVar="LOpwr
SweepVar=
LOpwr_dBm
dBm"
SimInstanceName[1]="HB1"
SimInstanceName[2]=
SimInstanceName[3]=
SimInstanceName[4]=
SimInstanceName[5]=
SimInstanceName[6]=
Start=-15
Stop=0
Step=1
p
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
LO Starvation Curve
-38
38
-38
38
-39
-39
dBm(vout_m
mdl[::,1])+0.1
dBm(vout_
_ckt[::,1])
dBm(vout_
_mdl[::,1])
dBm(vout_
_ckt[::,1])
Measurement Results
• small intentional display offset – to distinguish traces
Mag
-40
-41
-42
-43
-16
-14
-12
No display offset
-10
-8
-6
-4
-2
-42
-43
-16
0
-14
-12
-10
Small display offset
LOpwr_dBm
-8
-6
-4
-2
0
-6
-4
-2
0
LOpwr_dBm
-44
48
-48
Phase
-50
-52
-54
Ckt
Xpar
-56
phase((vout_mdl[::,1])+0.2
2
phas
se(vout_ckt[::,1])
-46
phase
e(vout_mdl[::,1])
phas
se(vout_ckt[::,1])
-41
41
-44
44
-44
-46
Phase
-48
-50
-52
-54
-56
-58
-58
-16
-14
-12
-10
-8
LOpwr_dBm
© Copyright Agilent Technologies 2010
Page 16
Mag
-40
-6
-4
-2
0
-16
-14
-12
-10
-8
LO
LOpwr_dBm
dB
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
LO to RF and LO to IF Leakage
• LO to RF and LO to IF leakage measured by maintaining a
constant
t t RF power and
d sweeping
i the
th LO power level
l
l
• magnitude and phase of output signal at the LO frequency
monitored at both the input and output ports
© Copyright Agilent Technologies 2010
Page 17
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
LO to RF and LO to IF Leakage
Measurement Setup
• same as LO starvation curve
• RFpwr = -50 dBm
• LOpwr = -15
15 dBm to 0 dBm
© Copyright Agilent Technologies 2010
Page 18
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
LO to RF Leakage
-30
-30
-32
-32
-34
dBm(mix(vin_m
mdl,{1,0}))+0.2
dBm(mix(vin_
_ckt,{1,0}))
dBm(mix(vin_
_mdl,{1,0}))
dBm(mix(vin_
_ckt,{1,0}))
Measurement Results
• small intentional display offset – to distinguish traces
Mag
-36
-38
38
-40
-42
-44
-46
-16
-14
-12
-10
No display offset
-8
-6
-4
-2
-38
-40
-42
-44
-46
-16
0
-14
-12
-10
Small display offset
LOpwr_dBm
-8
-6
-4
-2
0
-6
-4
-2
0
LOpwr_dBm
4.0
35
3.5
Phase
3.0
2.5
20
2.0
Ckt
Xpar
1.5
1.0
phase(m
mix(vin_mdl,{1,0}))+
+0.05
phas
se(mix(vin_ckt,{1,0})))
4.0
phase((mix(vin_mdl,{1,0})))
phase(mix(vin_ckt,{1,0}))
Mag
-36
-48
48
-48
3.5
Phase
3.0
2.5
2.0
1.5
1.0
-16
-16
-14
-12
-10
-8
LOpwr dBm
LOpwr_dBm
© Copyright Agilent Technologies 2010
Page 19
-34
-6
-4
-2
0
-14
-12
-10
-8
LOpwr_dBm
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
LO to IF Leakage
Measurement Results
• small intentional display offset – to distinguish traces
-24
dBm(mix(vout_m
mdl,{1,0}))+0.2
dBm(mix(voutt_ckt,{1,0}))
dBm(mix(vout_
_mdl,{1,0}))
dBm(mix(vout_
_ckt,{1,0}))
-24
-26
Mag
-28
-30
-32
-34
-36
-38
-16
-14
-12
No display offset
-10
-8
-6
-4
-2
-30
-32
-34
-36
-38
0
-16
-14
-12
Small display offset
LOpwr_dBm
128
-10
-8
-6
-4
-2
0
-6
-4
-2
0
LOpwr_dBm
129
127
Phase
126
125
124
Ckt
Xpar
123
122
phase(mix
x(vout_mdl,{1,0}))+0
0.1
phase(m
mix(vout_ckt,{1,0})))
phase(m
mix(vout_mdl,{1,0}))
phase(m
mix(vout_ckt,{1,0})))
Mag
-28
-40
-40
128
Phase
127
126
125
124
123
122
-16
-14
-12
-10
-8
LO
LOpwr_dBm
dB
© Copyright Agilent Technologies 2010
Page 20
-26
-6
-4
-2
0
-16
-14
-12
-10
-8
LOpwr_dBm
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
RF to IF Leakage
• RF to IF leakage measured by maintaining a constant LO
power and
d sweeping
i th
the RF power llevell
• magnitude and phase of output signal at the RF frequency
monitored at the output port
© Copyright Agilent Technologies 2010
Page 21
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
RF to IF Leakage
Measurement Setup
• same as compression curve
• RFpwr = -50 dBm to 0 dBm
• LOpwr = -5 dBm
© Copyright Agilent Technologies 2010
Page 22
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
RF to IF Leakage
Measurement Results
• small intentional display offset – to distinguish traces
-10
dBm(mix(vout_
_mdl,{0,1}))+1
dBm(mix(vou
ut_ckt,{0,1}))
dBm(mix(voutt_mdl,{0,1}))
dBm(mix(vout_ckt,{0,1}))
-10
-20
Mag
-30
-40
-50
-60
-50
-45
-40
-35
No display offset
-30
-25
-20
-15
-10
-5
-40
-50
-60
-50
0
-45
-40
-35
Small display offset
RFpwr_dBm
-30
-25
-20
-15
-10
-5
0
-15
-10
-5
0
RFpwr_dBm
115
110
Phase
105
100
Ckt
Xpar
95
90
-50
-45
-40
-35
-30
-25
-20
RFpwr dBm
RFpwr_dBm
© Copyright Agilent Technologies 2010
-15
-10
-5
0
phase(m
mix(vout_mdl,{0,1}))+0.5
phase
e(mix(vout_ckt,{0,1}}))
115
phase(mix(vout_mdl,{0,1}))
phase((mix(vout_ckt,{0,1})))
Mag
-30
-70
70
-70
Page 23
-20
110
Phase
105
100
95
90
-50
-45
-40
-35
-30
-25
-20
RFpwr
p _dBm
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
Modeling a Mixer Sub-System
The X-parameters may be used to model more than just the
mixer
i
it
itself.
lf
The following example is a simple system where the mixer is
cascaded with a band-pass filter.
The X-parameters are extracted for the entire cascade
encompassing both the mixer and the band-pass filter.
© Copyright Agilent Technologies 2010
Page 24
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
Mixer Sub-System – X-Parameter Extraction
Mixer and Band
Band-Pass
Pass Filter
• one set of power levels is verified
• similar tests as for the single mixer could be configured
DC
XP_Bias
PORT4
Num=4
DC_mode=Voltage
DC_value=5 V
Bi
Bias_port
t
RF_port
XP_Source
PORT1
Num=1
Z0=(50+j*0) Ohm
LS_freqHarms[1]=0,1
LS_format[1]=Mag/Phase
LS sw pType[1]=Single point
LS_sw
LS_value[1,Mag]=dbmtow (-50)
LS_value[1,Phase]=0
LS_start[1,Mag]=
LS_stop[1,Mag]=
LS_numPts[1,Mag]=
IF_port
LO_port
x_2_GilCellMix
X1
XP_Source
PORT3
Num=3
Z0=(50+j*0) Ohm
LS_freqHarms[1]=1,0
LS_format[1]=Mag/Phase
LS_sw pType[1]=Single point
LS_value[1,Mag]=dbmtow (-5)
LS_value[1,Phase]=0
LS_start[1,Mag]=
LS_stop[1,Mag]=
LS_numPts[1,Mag]=
BPF_Chebyshev
BPF1
Fcenter=250 MHz
BWpass=100 MHz
Ripple=0.5 dB
N=15
XP_Load
PORT2
Num=2
Z0=(50+j*0) Ohm
Load_mode=Impedance
LS_freqHarms[1]=1
LS format[1]=Mag/Phase
LS_format[1]=Mag/Phase
LS_sw pType[1]=Use sw eep
LS_value[1,Mag]=
LS_start[1,Mag]=
LS_stop[1,Mag]=
LS_numPts[1,Mag]=
X-Parameters
X
Parameters
X_Param
XP1
MaxOrder=5
Freq[1]=1.75 GHz
Freq[2]=2 GHz
Order[1]=5
Order[2]=5
XParamMaxOrder=5
© Copyright Agilent Technologies 2010
Page 25
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
Mixer Sub-System Spectrum Measurement
• output port now located after the BPF
• measurement made at single LO and RF power levels
• magnitude and phase of output signal at all frequencies
monitored at the output port
© Copyright Agilent Technologies 2010
Page 26
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
Mixer Sub-System
Spectrum Measurement Setup
• RFpwr = -50
V_DC
SRC1
Vdc=5 V
• LOpwr = -5 dBm
Bias_port
RF_port
P_1Tone
PORT1
Num=1
Z=50 Ohm
P=polar(dbmtow(-50),0)
Freq=2 GHz
v out_ckt
IF_port
x_2_GilCellMix
x
2 GilCellMix
X1
LO_port
P_1Tone
PORT2
Num=2
Z=50 Ohm
P=polar(dbmtow(-5),0)
Freq=1.75 GHz
BPF_Cheby shev
BPF1
Fcenter=250 MHz
BWpass=100 MHz
Ripple=0.5 dB
N=15
R
R1
R=50 Ohm
V_DC
SRC2
Vdc=5 V
HARMONIC BALANCE
HarmonicBalance
HB1
MaxOrder=5
Freq[1]=1.75 GHz
Freq[2]=2 GHz
Order[1]=5
Order[2]=5
© Copyright Agilent Technologies 2010
Page 27
X4P
XNP1
File="d0_MixerBPF.ds"
4
1
P_1Tone
PORT3
Num=3
Z=50 Ohm
P=polar(dbmtow(-50),0)
Freq=2 GHz
v out_mdl
2
3
Re f
P_1Tone
PORT4
Num=4
Z=50 Ohm
P=polar(dbmtow(-5),0)
Freq=1.75 GHz
R
R2
R=50 Ohm
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
Mixer Sub-System
0
0
-200
-200
Mag
dBm(vo
out_mdl)
dBm(vo
out_ckt)
dBm(vou
ut_mdl)
dBm(vou
ut_ckt)
Spectrum Measurement Results
• small intentional display offset – to distinguish traces
-400
-600
Mag
-400
-600
-800
-800
-1000
1
1.0E10
1
1.1E10
1.0E10
1.1E10
No display offset
9.0E9
10
9
9.0E9
9
8.0E9
8
8
8.0E9
7
7
7.0E9
6
6
6.0E9
5
freq, GHz
5
5.0E9
4
4
4.0E9
3
3
3.0E9
2
2
2.0E9
1
1
1.0E9
0
0
0.0
-1000
000
Small display offset
freq, Hz
freq+(0.05 GHz)
200
Phase
100
phase(vout_mdl)
phase(vout_ckt)
ph
hase(vout_mdl)
ph
hase(vout_ckt)
200
0
-100
freq, GHz
© Copyright Agilent Technologies 2010
Page 28
6
7
8
9
10
7.0E9
5
6.0E9
4
5.0E9
3
4.0E9
2
-200
3.0E9
1
-100
100
2.0E9
0
0
1.0E9
-200
Phase
0.0
Ckt
Xpar
100
freq, Hz
ffreq+(0.05
(0 05 GHz)
GH )
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
Mixer Sub-System Frequency Response
• output port now located after the BPF (BW = 100 MHz)
• measurement made at single LO and RF power levels
• RF frequency swept from 1.9 GHz to 2.1 GHz
• magnitude and phase of output signal at all frequencies
monitored at the output port
© Copyright Agilent Technologies 2010
Page 29
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
Mixer Sub-System – Frequency Sweep
Mixer and Band
Band-Pass
Pass Filter X-Parameter
X Parameter Extraction
• cross-frequency non-linear response
DC
XP_Bias
PORT4
Num=4
DC_mode=Voltage
DC_value=5 V
Bias_port
RF_port
XP_Source
XP
Source
PORT1
Num=1
Z0=(50+j*0) Ohm
LS_freqHarms[1]=0,1
LS_format[1]=Mag/Phase
LS_sw pType[1]=Single point
LS_value[1,Mag]=dbmtow (-50)
LS value[1 Phase]=0
LS_value[1,Phase]=0
LS_start[1,Mag]=
LS_stop[1,Mag]=
LS_numPts[1,Mag]=
X-Parameters
X_Param
XP1
MaxOrder=5
Freq[1]=1.75 GHz
Freq[2]=RFfreq
Order[1]=5
Order[2]=5
XP
XParamMaxOrder=5
M Od 5
© Copyright Agilent Technologies 2010
Page 30
IF_port
LO port
LO_port
Var
Eqn
x_2_GilCellMix
X1
BPF Ch b h
BPF_Chebyshev
BPF1
Fcenter=250 MHz
BWpass=100 MHz
Ripple=2 dB
N=9
XP_Load
XP
Load
PORT2
Num=2
Z0=(50+j*0) Ohm
Load_mode=Impedance
LS_freqHarms[1]=1
LS_format[1]=Mag/Phase
LS_sw pType[1]=Use sw eep
LS value[1 Mag]=
LS_value[1,Mag]=
LS_start[1,Mag]=
LS_stop[1,Mag]=
LS_numPts[1,Mag]=
XP_Source
PORT3
Num=3
Z0=(50+j*0) Ohm
LS_freqHarms[1]=1,0
LS_format[1]=Mag/Phase
LS_sw pType[1]=Single point
LS_value[1,Mag]=dbmtow (-5)
LS_value[1,Phase]=0
LS_start[1,Mag]=
LS_stop[1,Mag]=
LS_numPts[1,Mag]=
VAR
SWEEP PLAN
VAR1
RFfreq=2 GHz
Sw eepPlan
Sw pPlan1
Start=1.9 GHz Stop=2.1 GHz Step=0.001 GHz Lin=
UseSw eepPlan=
Sw eepPlan=
Reverse=no
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
Mixer Sub-System – Frequency Sweep
Measurement Setup
• RFfreq = 1.9 GHz to 2.1 GHz
V_DC
SRC1
Vdc=5 V
• RFpwr = -50 dBm
• LOpwr = -5 dBm
Bias_port
RF_port
P_1Tone
PORT1
Num=1
Z=50 Ohm
P=polar(dbmtow(-50),0)
Freq=RFf req
v out_ckt
IF_port
x_2_GilCellMix
X1
LO_port
P_1Tone
PORT2
Num=2
Z=50 Ohm
P=polar(dbmtow(-5),0)
Freq=1 75 GHz
Freq=1.75
BPF_Cheby shev
BPF1
Fcenter=250 MHz
BWpass=100 MHz
Ripple=2 dB
N=9
R
R1
R=50 Ohm
HARMONIC BALANCE
HarmonicBalance
HB1
MaxOrder=5
Freq[1]=1.75 GHz
Freq[2]=RFf req
Order[1]=5
Order[2]=5
Var
Eqn
V_DC
SRC2
Vdc=5 V
X4P
XNP1
File="d2_MixerBPF_Fswp.ds"
VAR
VAR1
RFf req=2 GHz
4
1
PARAMETER SWEEP
ParamSweep
Sweep1
SweepVar="RFf req"
SimInstanceName[1]="HB1"
SimInstanceName[2]=
SimInstanceName[3]=
SimInstanceName[4]=
SimInstanceName[5]=
SimInstanceName[6]=
Start=1.9 GHz
Stop=2.1
p
GHz
Step=0.001 GHz
© Copyright Agilent Technologies 2010
Page 31
P_1Tone
PORT3
Num=3
Z 50 Oh
Z=50
Ohm
P=polar(dbmtow(-50),0)
Freq=RFf req
v out_mdl
2
3
Re f
P_1Tone
PORT4
Num=4
Z=50 Ohm
P=polar(dbmtow(-5),0)
Freq=1.75 GHz
R
R2
R=50 Ohm
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
Mixer Sub-System – Frequency Sweep
Measurement Results
• small intentional display offset – to distinguish traces
-30
-20
dBm(vout_
_mdl[::,1])
dBm(vout_
_ckt[::,1])
dBm(vout_m
mdl[::,1])+1
dBm(vout_
_ckt[::,1])
Mag
-40
-60
-80
-100
-120
Mag
-35
-40
40
-45
-140
-50
-160
Phase
100
50
0
-50
-100
100
Ckt
Xpar
-200
Phase
100
0
-100
100
-200
2.10E9
2.08E9
2.06E9
2.04E9
2.02E9
2.00E9
1.98E9
1.96E9
1.94E9
1.92E9
1.90E9
2.10E9
2.08E9
2.06E9
2.04E9
© Copyright Agilent Technologies 2010
2.02E9
2.00E9
1.98E9
1.96E9
1.94E9
1.92E9
1.90E9
RFfreq
pha
ase(vout_mdl[::,1])+10
0
phase(vout_ckt[::,1])
phase(vout_mdl[::,1])
phase(vout_ckt[::,1])
p
2.06E9
2.05E9
2.04E9
2.03E9
2.02E9
2.01E9
2.00E9
1.99E9
1.98E9
1.97E9
RFfreq
200
-150
Page 32
1.96E9
Small display offset
200
150
1.95E9
RFfreq
1.94E9
2.10E9
2.08E9
2.06E9
2.04E9
2.02E9
2.00E9
1.98E9
1.96E9
1.94E9
1.92E9
1.90E9
No display offset
RFf
RFfreq
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010
Conclusions
We have demonstrated excellent accuracy of X-parameter
models
d l iin mixer
i
applications:
li i
spectrum, compression,
i
starvation, leakage, frequency response.
It is important to understand the X-parameter model needs to
have a proper coverage for the operating conditions under
which it is going to be used (power levels
levels, frequencies
frequencies, loading
loading,
biasing).
X-parameter
X
t models
d l may simulate
i l t significantly
i ifi
tl ffaster
t th
than th
the
transistor level circuits.
X-parameter models provide the IP-protection of the circuit level
designs.
© Copyright Agilent Technologies 2010
Page 33
IMS 2010 - MicroApps: Accurate Mixer Measurements
Using Multi-Tone X-Parameter Models
May 26, 2010