1 of 17

ECE 218B �5G Receiver

JUSTIN KIM

1

2 of 17

Introduction

 

2

Receive Band

Channel Bandwidth

Desired SNR

Maximum Jammers

EVM or IRR

Minimum IF/BB Output Power

26.50 – 29.50 GHz

200 MHz

8 dB for 16QAM, ½ FEC

-20 dBm at 29.5 GHz,

-40 dBm at 200 MHz offset

> 30 dB

-20 dBm

3 of 17

Low-Noise Amplifier

  • Procedure:
    1. Bias for minimum noise.
    2. Stabilize the transistor.
    3. Input match to noise.
    4. Output match to 50 Ohms.

3

 

4 of 17

LNA Results

  • Gain = 9.50 dB
  • NF = 1.514 dB
  • Input Return Loss = -4.503 dB
  • Output Return Loss = -57.310 dB
  • IIP3 = -2.251 dBm
  • OIP3 = 7.246 dBm

4

5 of 17

BiCMOS Mixer

Design Considerations

  • CMOS gm cell offers high IIP3.
  • Bipolar quad offers high-speed switching.
  • Increasing transistor width improves CG but decreases linearity.
  • Inductive degeneration improves linearity but decreases CG.
  • Get at least IIP3 > 7.3 dBm

5

RF_freq = 28 GHz LO_freq = 27.999 GHz IF_freq = 1 MHz

 

[1]

6 of 17

Mixer Results

The optimum LO power for IIP3 was -1.0 dBm.

  • CG = 0.261 dB
  • IIP3 = 8.946 dBm
  • OIP3 = 9.207 dBm
  • NF = 10.507 dB

6

7 of 17

IF Amplifier

  • Two stage differential amplifier with a series feedback resistor.
  • Differential stage with active load 🡪 high gain.
  • Had to sacrifice a lot of gain to meet linearity requirement.

7

 

8 of 17

IF Amplifier Results

  • Gain = 19.73 dB (Needed 53 dB)
  • 3dB BW = 12 GHz
  • IIP3 = 9.252 dBm
  • NF = 3.4 dB

8

9 of 17

In Band (IMD) and Out of Band (XMD) �Jammer Tests

  • IMD: Two Tones: -40 dBm at 28.2 GHz, -40 dBm at 28.4 GHz
    • Need a receiver IIP3 of -18.5 dBm to pass this test.
  • XMD: Three Tones: -40 dBm at 28.5 GHz, -20 dBm at 29.5 GHz, -20 dBm at 30 GHz

🡪 Need a receiver IIP3 of 4.5 dBm to pass this test.

  • My receiver only has a IIP3 of -5.9 dBm
  • In order to pass the XMD test I will add a Butterworth BSF.

9

IIP3 of my IQ receiver

with no filters.

10 of 17

Butterworth Band Select Filter

10

[2]

11 of 17

Jammer Test Results

  • We want the unwanted 3rd order IMD and XMD tones to be below output noise floor (-83 dBm + 29.5 dB = -53.5 dBm)

11

IMD Test Results

XMD Test Results

12 of 17

Cascaded FOM

 

12

Measured IIP3 of total receiver with BSF.

13 of 17

Project Review

PROS

  • Low NF on LNA.
  • Linear Mixer.
  • Good overall IIP3 for receiver.
  • Beat IMD and XMD test.

​

CONS

  • Low gain on IF amplifier.
  • No image rejection filter.

13

14 of 17

Future Improvements

  • Focus more on the IF amplifier.

🡪 Review more analog circuits / op-amp design in order to improve gain and linearity.

  • Add an image rejection filter.
  • Look at gm3 curves in order to bias for linearity.

14

15 of 17

Sources

[1] http://rfic.eecs.berkeley.edu/ee242/pdf/Module_5_3_MixerDesign.pdf

[2] https://rf-tools.com/lc-filter/

​

15

16 of 17

Supplementary Slides

16

17 of 17

IF Amplifier without Resistive Feedback

  • Gain = 55.5 dB
  • 3dB BW = 660 MHz
  • IIP3 = -41 dBm
  • NF = 1.959 dB

​

17