Si-Vision Academy Analog Labs Lab 03

Common Drain Frequency Response

PART 1: Sizing Chart
2) Setup for PMOS characterization on schematic editor was as follows:

with VGS, VDS, W and L parameterized.

3) An OP analysis was conducted to calculate PMOS’ VTH. Through OP Points Results
Report, |VTH,P| = 0.411770

4) We now have our parameters as follows:

with DC sweep on VGS from 0 to 820 mV with 10 mV step size.

1
Si-Vision Academy Analog Labs Lab 03

6) VOV and V* overlaid vs. VGS is as follows:

7) At V* = 200 mV (this time with V* as x-axis as opposed to y-axis), VGS,Q = 607 mV and
VOV,Q = 196 mV

2
Si-Vision Academy Analog Labs Lab 03

8) Below are curves for ID, gm and gds vs. VGS:

At VGS,Q = 607 mV, ID,X = 12 µA, gm,X = 120 µS and gds,X = 928 nS
𝑊𝑋
9) Using formula 𝑊𝑄 = 𝐼𝐷,𝑄 𝐼 in a save.text file as follows:
𝐷,𝑋

where IDQ = 10 µA (which we want as spec) and IDX is ID at V* = 200 mV; W (in log
scale in y-axis) is plotted vs. V* (in x-axis), we find that WQ = 8.32 µm

3
Si-Vision Academy Analog Labs Lab 03

10) Likewise with gm and gds, gm,Q and gds,Q were obtained through following expressions in
save.text file:

with calculated values of 100 µS and 774 nS respectively at V* = 200 mV.

PART 2: CD Amplifier
1. OP Analysis
1) Schematic of PMOS CD amplifier is as follows:

4
Si-Vision Academy Analog Labs Lab 03

with parameters as shown:

2) Below are the OP Analysis Points Results:

CSB ~ CStot – CGS ~ 15.3037 fF

CDB ~ CDtot – CGD ~ 6.63238 fF
3) Printed Points Results show “2” for region, indicating M0 is in saturation.
2. AC Analysis
2) Bode plot for magnitude response is shown below:

3) Peaking in freq. domain is quite noticeable, and determined to be 4.28293 dB, thanks
to evaluation of ymax(db20(v(/Vout))) expression.
4) Neglecting 1/gds and gmb,
√𝑏2
𝑄=
𝑏1
𝐶𝐺𝑆 +𝐶𝐿 (𝐶𝐺𝑆 +𝐶𝐺𝐷 )𝐶𝐿 +𝐶𝐺𝑆 𝐶𝐺𝐷
Where 𝑏1 = 𝐶𝐺𝐷 𝑅𝑠𝑖𝑔 + and 𝑏2 = ( ) 𝑅𝑠𝑖𝑔 . Plugging in
𝑔𝑚 𝑔𝑚
values, we get b1 = 31.7n and b2 = 2.49f, giving us Q = 1.58 > 0.5, definitely an
underdamped system.

5
Si-Vision Academy Analog Labs Lab 03

5) We get the following magenta, light mauve and orange Bode plots for magnitude
responses for CL at 2, 4 and 8 pF overlaid respectively:

We note that:
 Peaking happens at lower frequencies as CL is increased
 Peaking above 0 dB reaches lower values as CL is increased
This is because:
 As CL is increased, both b1 and b2 increase, but due to b2 being under square root
(a weak function), increase in numerator is less than increase in denominator,
meaning Q falls meaning lower peaking as CL is increased.
 As CL is increased, b2 increases, ω0 decreases, increasing output pole’s dominance
by lowering its frequency, hence explaining why peaking happens at lower
frequencies.
6) We get the following light blue, red and magenta Bode plots for magnitude
responses for Rsig at 20 k, 200 k and 2 MOhms overlaid respectively:

As Rsig is increased, peaking is increased. This is because, with Rsig input pole gets closer
to dominant output pole, losing the advantage of having a dominant and nondominant
pole; resulting in Q increase, meaning higher peaking.

6
Si-Vision Academy Analog Labs Lab 03

3. Transient Analysis
2) Vin and Vout overlaid across 10 µs:

3) DC shift between Vin and Vout = 607 mV = |VGS|. We note that this is a DC shift
upwards. A DC shift downwards could have been achieved through using an NMOS
CD amplifier.
4) Ringing in time domain in Vout waveform is noticed with each step in Vin. An
attempt was made to use “overshoot” from “Measurement Tool”, choosing “target
signal topline/baseline” for percentage lock level, but there was lack of understanding
on how to set margin thresholds for rise and fall. However, plugging in value
calculated for Q in 4) in Part 2. AC analysis, 100% overshoot = 35.1%.
5) Performing a parametric sweep for CL = 2 p, 4 p and 8 pF nested in transient analysis:

We see that as CL is increased, overshoot decreases. Least overshoot (light mauve) is for
CL = 8 pF.
6) Conversely with Rsig parametric sweep, we see that increasing Rsig increases
overshoot, with highest (light mauve) being for Rsig = 2 MOhm.

7
Si-Vision Academy Analog Labs Lab 03

4. Zout (Inductive Rise) (optional)

1) Configuration for calculating Zout frequency response in schematic editor is as
follows:

3) We indeed notice the inductive rise. This is because we are first met by ωz =
1/(RsigCGS), which is << ωp1 = 1/((1/gm)CGS)

8
Si-Vision Academy Analog Labs Lab 03

4) Zout does not remain flat after ωp1 counteracts inductive rise of ωz. Another pole at ωp2
manifests as CGD shunts Rsig at high frequencies. It manifests earlier than ωp1 due to
(1/gm) << Rsig in denominator of both poles.
5) Obtaining zeros, poles and |Zout| at both low and high freq analytically:
1
ωz = 𝑅 𝐶 = 8.88 Mrad/s = 1.41 MHz.
𝑠𝑖𝑔 𝐺𝑆
𝑔
ωp1 = 𝐶 𝑚 = 1.76 Grad/s = 281 MHz
𝐺𝑆
1
ωp2 = 𝑅 = 91.2 Mrad/s = 14.5 MHz
𝑠𝑖𝑔 𝐶𝐺𝐷
|Zout| @ low freq ~ 1/gm = 10.0 kOhm = 80.0 dB
|Zout| @ high freq ~ Rsig = 2 MOhm = 126 Db

Obtaining zeros, poles and |Zout| at both low and high freq from simulation:
ωz > On phase Bode plot, add H-cursor > Show H-crossing values > Set cursor’s y-
value at 45 > first H-crossing value = 1.58 MHz
ωp2 > Measure tool > Level > Maximum > 55.2° > 55.2 – 45 = 10.2° > Apply On
phase Bode plot, add H-cursor > Show H-crossing values > Set cursor’s y-value at
10.2 > second H-crossing value = 28.9 MHz
ωp1 > On phase Bode plot, add H-cursor > Show H-crossing values > Set cursor’s y-
value at 10.2 – 45 = -34.8° > first H-crossing value = 107 MHz
|Zout| @ low freq: Measure tool > Level > Baseline > Apply > 80 dB
|Zout| @ high freq: Measure tool > Level > Topline > Apply > 100 dB

Analytical Calculation Simulation

ωz (MHz) 1.41 1.58
ωp1 (MHz) 281 107
ωp2 (MHz) 14.5 28.9
|Zout| @ low freq (dB) 80 80
Zout| @ high freq (dB) 126 100