S INGLE -B ALANCED M IXER P ROJECT F INAL P RESENTATION RIT Senior Project Jared Burdick May 17, 2012 Multidisciplinary Senior Design Conference Kate Gleason.

SINGLE-BALANCED MIXER PROJECT

FINAL PRESENTATION

RIT Senior Project

Jared Burdick

May 17, 2012

Multidisciplinary Senior Design Conference

Kate Gleason College of Engineering

Rochester Institute of Technology

Rochester, New York 14623

INTRODUCTION What is a mixer?

A device used to convert frequencies. Mixer is a term generally associated with converting higher

frequencies to lower frequencies. Where are they used?

Communication systems. Radar applications.

How does a mixer work? They take advantage of the non-linear properties of diodes. The signal (RF) is “mixed” with another fixed (or tunable)

frequency (LO) and a “difference” frequency (IF) is produced along with a number of predictable inter-modulation products.

There are several different configurations for mixers. A single-balanced configuration was selected for this project.

RF

LO

IF

PROJECT GOALS

Research Mixers Understand theory, applications, configurations, and

design trade-offs. Design, Simulate, Prototype Mixer

Choose an appropriate configuration. Develop design and simulation skills. Mitigate risks and follow project plan. Test mixer and compare simulated to actual

performance. Analyze results and offer possible future improvements /

implementations.

CUSTOMER NEEDS

Need to Update

SPECIFICATIONS

Need to Update

Several specifications were modified during the development (with customer approval)

SYSTEM BLOCK DIAGRAM

Need to Update

COMPONENTS USED Anaren 90º Hybrid Coupler (XC0900A-3S)

Avago Schottky-Diode (HSMS-2822)

Coilcraft Chip Inductors (0805HT-12NTJB)

DLI Chip Capacitors (C06UL120G and C04UL2R7)

Gigalane SMA Connector (PAF-S05-007)

Murata Chip Inductor (LQW18AN39NG00D)

Rogers Substrate Material (RO4003C)

AWR MODELSIN D QID = L 1L = 1 2 n HQ = 4 0F Q = 0 .1 G H zA L P H = 0

C A P QID = C 1C = 1 2 p FQ = 5 0F Q = 0 .1 G H zA L P H = 0

IN D QID = L 2L = 3 9 n HQ = 4 0F Q = 0 .1 G H zA L P H = 0

C A P QID = C 2C = 1 2 p FQ = 5 0F Q = 0 .1 G H zA L P H = 0

IN D QID = L 3L = 1 2 n HQ = 4 0F Q = 0 .1 G H zA L P H = 0

P O R TP = 1Z= 5 0 O h m

P O R TP = 2Z= 5 0 O h m

L u m p ed E lem en t F ilt e r

L u m p ed L P F

M ax im a lly F la t

D eg r ee = 5

F p = 4 0 0 M H z

Single-Balanced Mixer

5th Order LPF

DESIGN TRADE-OFFS & DECISIONS MADE Configuration

Use commercially available components wherever possible. Removed BPF’s from the RF and LO paths due to not readily available. Went to lumped-element LPF in the IF path for the same reason.

LO Leakage (LO to IF Isolation) Increased to 5th order of LPF at IF port

Better rejection (approx. 20dB more) at 1GHz, which improved LO/IF isolation (SBM configuration offers no natural reduction of the LO).

Conversion Loss Flatness Added micro-strip quarter-wave transformer to help match the impedance

coming out of the diodes and going into LPF Varied width of micro-strip line to see which gave the best conversion loss result

Changed the radial RF micro-strip choke into a shorted quarter-wave micros-trip stub

Tried Various angles for the radial choke and line width and found there was little improvement

Finally went to a true shorted quarter-wave stub Gave the best result in simulation Easy to provide ground to stub for physical layout

Added RF bypass capacitor shorted to ground after the diode provide additional filtering prior to the impedance transformation.

Improved conversion loss level and flatness

SIMULATION RESULTS

0.8 0.9 1 1.1 1.2Frequency (GHz)

Mixer C onversion Loss

-8

-7

-6

-5

-4

Co

nv

ers

ion

Lo

ss

(d

Bm

)

0.8 1.2Frequency (GHz)

M ixer Return Losses

-40

-30

-20

-10

0

Re

turn

Lo

ss

(d

B)

LO Return Loss, dB

RF Return Loss, dB

0.8 0.9 1 1.1 1.2Frequency (GHz)

Mixer Isolation

-70

-60

-50

-40

-30

-20

Ou

tpu

t L

eve

l (d

Bm

)

PRF@IF (dBm)

PLO@IF (dBm)

0 1 2 3 4 4.8Frequency (GHz)

IF Outpu t P o w e r S p e c tru m

-80

-60

-40

-20

0

Po

we

r (d

Bm

)

0.6 GHz-63.94 dBm

0.4 GHz-61.92 dBm

0.8 GHz-51.21 dBm

1 GHz-40.24 dBm

0.2 GHz-16.86 dBm

RF = 0.8 GHZLO = 1.0 GHz

SIMULATION RESULTS

0 0.5 1 1.5 2 2.5Frequency (GHz)

LPF IL RL

-80

-60

-40

-20

0

Inse

rtio

n L

oss

-40

-30

-20

-10

0

Re

turn

Lo

ss

DB(|S(2,1)|) (L)LPF

DB(|S(1,1)|) (R)LPF

0.8 0.9 1 1.1 1.2 1.3Frequency (GHz)

XC0900A3 Coupling

-3.6

-3.4

-3.2

-3

-2.8

Co

up

lin

g (

dB

)

1.1994 GHz-2.8599 dB

1.1995 GHz-3.5242 dB

1.0006 GHz-3.0466 dB

1.0003 GHz-3.1613 dB

DB(|S(3,1)|)Coupler

DB(|S(4,1)|)Coupler

Power CompressionApproximate 1dB Compression Points

CIRCUIT LAYOUT &ASSEMBLED UNIT

SMA Conn Launch(RF In)

SMA Conn Launch(LO In)

SMA Conn Launch (IF Out)

LPF

λ/4 transformerDiode Pair

Coupler

λ/4 shorted stub

λ/4 shorted stub

RF Bypass Cap

Circuit Layout

Assembled Unit

TEST RESULTS – SUMMARY COMPARISON

All specifications were met by both units built.Both units had very similar performance.

Unit 1 Unit 2S1 RF Frequency GHz 0.8 - 1.2 OK OK OKS2 LO Frequency GHz 1 OK OK OKS3 IF Frequency MHz DC - 200 OK OK OK

S4Maximum Conversion Loss

dB 10 7.2 7.6 8.0

S5 RF/IF Isolation dB 30 42 37.0 38.5S6 LO/IF Isolation dB 35 50 41.0 43.0 LPF rejection less than simulated.S7 Minimum LO Power dBm 10 OK OK OKS8 Maximum RF Power dBm -10 OK OK OK

S9Minimum 1dB Compression

dBm 5 8 8.6 8.6 Approximate Value

S10 Maximum RF VSWR Ratio 2.0:1 1.40:1 1.29:1 1.25:1S11 Maximum LO VSWR Ratio 2.5:1 1.09:1 1.05:1 1.08:1S12 Maximum IF VSWR Ratio 1.5:1 - 1.46:1 1.43:1 Not simulated

LO & RF Port VSWR's response similar - LO port defined at 1.0 GHz only (null)

All parameters tested at these levels.

Spec #

Specification (metric)

Unit of Measure

Spec Simulated Measured Comments

All parameters tested over these ranges. Several tests looked at performance beyond specified.

TEST RESULTS

-8.0

-7.5

-7.0

-6.5

-6.0

-5.5

-5.0

-4.5

-4.0

Conv

ersi

on G

ain

(dB)

RF Frequency (GHz)

Conversion Gain (dB)Unit #1

Conversion Gain (dB)

Simulation (dB)

0

10

20

30

40

50

60

0.800 0.900 1.005 1.100 1.200

Isol

ation

(dB)

RF Frequency (GHz)

RF - IF Isolation - Unit #1

Measured Simulated

4.0

6.0

8.0

10.0

12.0

14.0

16.0

18.0

-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

Conv

ersi

on L

oss

(dB)

RF Input Power Level (dBm)

Conversion Loss vs. RF Input PowerUnit #1

0.8 GHz

0.9 GHz

1.005 GHz

1.1 GHz

1.2 GHz

-60

-50

-40

-30

-20

-10

0

Spur

Leve

l(dB

c)

Frequency (GHz)

Highest Spur Output Level Unit #1

Measured Simulated

Conversion Loss RF to IF Isolation

1-dB Compression

Spurious Output

TEST RESULTS

Unit #1IF Output SpectrumLO = 1000 MHzRF = 850 MHzHoriz. Scale: 200 MHz/div

0 1 2 3 4 4.85Frequency (GHz)

IF Outpu t P o w e r S p e c tru m

-80

-60

-40

-20

0

Po

we

r (d

Bm

)

0.7 GHz-77.27 dBm

0.3 GHz-72.65 dBm 0.85 GHz

-54.26 dBm

1 GHz-40.24 dBm

0.15 GHz-16.27 dBm

RF = 0.85 GHZLO = 1.0 GHz

Spurious Output

CONCLUSIONS

The prototype mixer met the target specifications.

There were differences between the simulated performance and the actual measured performance.

In general, the actual measured performance was consistent with the model.

LO to IF Isolation about 7-9 dB less. Suspect that the LPF roll-off (rejection at higher frequencies) was less

than modeled – this will need further evaluation to confirm. RF to IF Isolation 3-5 dB less – LPF roll-off would contribute here as well. Conversion Loss was slightly higher – connectors not modeled could be a

contributor.

Future Iterations / Investigations. Add BPF to the LO and RF input paths. Investigate LPF performance. Refine the AWR model (connectors, HFSS sub-models, etc.). Work on final mechanical packaging concept.

LESSONS LEARNED

Do your homework before starting to design There are many trade-offs that need to be considered and decisions that

need to be made in order to best match the expected performance to the application and requirements.

Ability to model the circuits accurately was key and greatly increased the probability of success.

Time is a scare resource Valuable lessons can be learned even in non-ideal circumstances. Figuring out project limitations early on in the process helped reduce risk

and deliver the final product on time. Look at contingency plans (alternate parts, fabrication alternatives etc.) Identifying concrete action items helped to focus efforts and reduce

wasted time.

Make use of all available resources Eliciting feedback from other knowledgeable people proved invaluable. There was a significant amount of information available on-line (technical

papers, forums, etc.).