Jose Silva - Martinez 1 Typical Transimpedance - limiting amplifier interfaces Many circuits are discussed in Razavi’s textbook “ Design of Integrated Circuits for Optical Communications”; check it for detailed discussions Use previously discussed techniques for analyzing frequency limitations, power consumption and noise floor Peaking techniques may help, but pay special attention to group delay! Power consumption is a major issue, even if dynamic circuits are used!
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Jose Silva-Martinez
1
Typical Transimpedance-limiting amplifier
interfaces
Many circuits are discussed in Razavi’s textbook “ Design of
Integrated Circuits for Optical Communications”; check it for detailed
discussions
Use previously discussed techniques for analyzing frequency
limitations, power consumption and noise floor
Peaking techniques may help, but pay special attention to group
delay!
Power consumption is a major issue, even if dynamic circuits are
used!
Jose Silva-Martinez
2
Typical Transimpedance-limiting interface
Iin {0.1-50 mA}; RF {0.5-1 kΩ}
Limited TIA bandwidth smoothes the input current Iin and hence V1
Small signal at the output of TIA
4-5 stages with voltage gain of 4-8 dB/stage is required
Offset voltages are accumulated offset canceller is required
Gain/stage is not well controlled under PVT variations AGC system
Vb
Jose Silva-Martinez
3
Transimpedance amplifier based on Resistive Feedback
Stability could be an issue
3 poles in a loop!
Poles are not far from each other
Rf is usually in the range of 500-1kΩ
Does this help?
If so, trade-offs?
Always advisable?
Jose Silva-Martinez
4
Transimpedance amplifier based on Resistive
Feedback
Interesting output stage!
RE1<<RF
Ve3~ IinRF
Additional gain due to R2?
Jose Silva-Martinez
5
Transimpedance amplifier based on Resistive Feedback
Option II
Peaking network can be a simple capacitor
Double check amplifier’s group delay
Pole-zero matching is an issue: Significant ripple may result
Jose Silva-Martinez
6
Transimpedance amplifier based on Resistive Feedback
Option II + single-differential output stage
Single-ended output at R6
Fully-differential at nodes A and B
Why source degeneration?
Why R6 is split in two pieces?
Jose Silva-Martinez
7
Transimpedance amplifier based on Resistive Feedback
With adjustable feedback resistor and DC control
R1C1 are a low-frequency filter to extract the DC value (offset) of Vout; it can be compared with a reference (not shown) to have better control.
You may want to use a power detector and control the gain
M3 and M4 provides a DC feedback
Base current of Q1 can be provided by M2 to minimize VRF
Jose Silva-Martinez
8
Cherry-Hooper amplifier
Fully differential amplifier
First topology requires a CMFB
2nd topology does not require CMFB but it is not isolated from VDD!•Voltage headroom forces us to reduce RD•Trade-off with voltage gain of M3
Stability issues for very HF applications (to be discussed in class)
power consumption could be an issue
Jose Silva-Martinez
9
Cherry-Hooper amplifier
Fully differential amplifier
First topology does not require CMFB
2nd topology requires CMFB!
Noise and stability are always relevant issues
Again: two poles in a loop!
Stability issues for very HF applications; power consumption
Small
signal
swing
Jose Silva-Martinez
10
Modified Cherry-Hooper amplifier having
low-output impedance and current re-use
Gain is still dominated by
gm1RF
Output impedance is
dominated by re3!
Bias current of Q1 is re-used by
Q3 to provide low output
impedance
Nice topology but may require
excessive voltage headroom
Jose Silva-Martinez
11
Amplifiers using CMOS Transistors!
Inductive peaking increases frequency response but please evaluate group delay ripple!