Lab Activity: Amplifier Gain & Frequency Response Using AI-Assisted Analysis

使用 AI 輔助分析放大器增益與頻率響應

Chung Yuan Christian University 中原大學 電子工程學系 Department of Electronic Engineering

Course: Technical English and Popular Science Communication / 科技英文與科普傳播 Instructor: Prof. Yao-Jen Chang (張耀仁)

Student Name / 姓名: ___________________　　Student ID / 學號: ___________________ Date / 日期: ___________________

1. Objectives / 學習目標

This lab teaches you two complementary skills simultaneously: (1) the engineering methodology for analyzing amplifier circuits, and (2) effective AI prompting strategies for technical problem-solving. By the end of this activity, you will be able to:

1. Analyze a Common-Emitter (CE) BJT amplifier from DC bias through frequency response
2. Apply the Miller Effect to predict high-frequency bandwidth limitations
3. Use the Short-Circuit Time Constant (SCTC) method for low-frequency pole analysis
4. Generate and interpret Bode plots using Python
5. Develop structured AI prompting skills for engineering analysis

本實驗同時教授兩項互補技能：(1) 放大器電路分析的工程方法，(2) 技術問題求解的有效 AI 提示策略。

2. Background / 背景知識

2.1 Why Step-by-Step Prompting?

When solving complex engineering problems with AI, a structured step-by-step approach offers significant advantages over a single monolithic prompt:

• Verification at each stage — You can catch errors early before they propagate through subsequent calculations.
• Deeper understanding — Each step requires you to evaluate the AI's output, reinforcing your engineering judgment.
• Reproducibility — A documented prompting strategy can be reused and refined across different circuits.
• Technical communication — Learning to specify precise engineering constraints (e.g., "do not ignore IB") is itself a professional skill.

當使用 AI 解決複雜工程問題時，結構化的逐步提示方法比單一提示具有顯著優勢：可在每個階段驗證結果、加深工程理解、提高可重複性，並培養專業溝通技能。

2.2 The Standard Strategy

The analysis follows a well-established engineering workflow. Each step builds on the previous result:

3. Circuit Parameters / 電路參數

Use the following Common-Emitter (CE) BJT amplifier parameters for this lab:

3.1 Circuit Schematic / 電路圖

Show Image

Figure 1. Common-Emitter BJT Amplifier with Voltage Divider Bias

4. Procedure / 實驗步驟

Follow each step below. Copy the suggested prompt into Claude, then verify the output using the checkpoint provided. Record both the AI output and your verification notes.

請依序執行以下步驟。將建議的提示複製到 Claude，然後使用檢查點驗證輸出。記錄 AI 輸出和您的驗證筆記。

Step 1: Solve for DC Operating Point (Q-Point)

求解直流工作點

Suggested Prompt: (Copy and paste into Claude)

Given the following Common-Emitter amplifier circuit:
VCC = 12V, R1 = 47kΩ, R2 = 10kΩ, RC = 4.7kΩ, RE = 1kΩ, β = 150

Solve for the DC operating point (ICQ, VCEQ).
Do NOT ignore base current (IB). Use the exact base current in your KVL calculation.
Show all work step by step.

✅ Verification Checkpoint: Check: Is I_CQ in the expected mA range? Is V_CEQ between 0 and V_CC? Compare the exact solution with the simplified ("I_B ≈ 0") result — what is the percentage error?

檢查：I_CQ 是否在合理的 mA 範圍內？V_CEQ 是否介於 0 與 V_CC 之間？比較精確解與簡化解的百分比誤差。

✍ Student Notes:

(Record key values and observations here)

 

 

Step 2: Calculate Midband Voltage Gain

計算中頻電壓增益

Suggested Prompt: (Copy and paste into Claude)

Using the DC operating point from Step 1, now calculate:
1. The transconductance gm
2. The input resistance rπ
3. The midband voltage gain Av (with load RL = 10kΩ)

Use the hybrid-π small-signal model. Show all equations.

✅ Verification Checkpoint: Check: Does |A_v| seem reasonable for a CE amplifier (typically 50–300)? Is the gain negative (inverting)?

檢查：|A_v| 對於 CE 放大器是否合理（通常 50–300）？增益是否為負值（反相）？

✍ Student Notes:

(Record key values and observations here)

 

 

Step 3: Determine Miller Effect & High-Frequency Poles

求解 Miller 效應與高頻極點

Suggested Prompt: (Copy and paste into Claude)

Now calculate the Miller effect and high-frequency response:
Given Cπ = 20pF, Cμ = 2pF (from datasheet)

1. Calculate the Miller capacitance (CM)
2. Determine the input pole frequency (fH_in)
3. Determine the output pole frequency (fH_out)
4. Identify the dominant high-frequency pole

Include the source resistance RS = 1kΩ in your calculation.

✅ Verification Checkpoint: Check: Is C_M significantly larger than C_μ? Is f_H_in lower than f_H_out (usually the dominant pole)?

檢查：C_M 是否顯著大於 C_μ？f_H_in 是否低於 f_H_out（通常是主導極點）？

✍ Student Notes:

(Record key values and observations here)

 

 

Step 4: Calculate Low-Frequency Poles

計算低頻極點

Suggested Prompt: (Copy and paste into Claude)

Calculate the low-frequency poles for the amplifier:
Given C1 (input coupling) = 1μF, C2 (output coupling) = 4.7μF, CE (bypass) = 47μF

1. Calculate the pole due to C1 (input coupling capacitor)
2. Calculate the pole due to C2 (output coupling capacitor)
3. Calculate the pole due to CE (emitter bypass capacitor)
4. Identify the dominant low-frequency pole

Use the short-circuit time constant method.

✅ Verification Checkpoint: Check: Is the C_E bypass pole typically the dominant (highest frequency) low-frequency pole? Are all poles in the Hz–kHz range?

檢查：C_E 旁路極點是否通常是主導低頻極點？所有極點是否在 Hz–kHz 範圍？

✍ Student Notes:

(Record key values and observations here)

 

 

Step 5: Generate Bode Plots in Python

用 Python 繪製 Bode 圖

Suggested Prompt: (Copy and paste into Claude)

Using all the values calculated above, generate Python code to plot:
1. Bode magnitude plot (Gain in dB vs. Frequency in Hz)
2. Bode phase plot (Phase in degrees vs. Frequency in Hz)

Requirements:
- Use matplotlib with two subplots (magnitude on top, phase on bottom)
- Frequency range: 1 Hz to 100 MHz (log scale)
- Mark the -3dB bandwidth, midband gain, and all pole frequencies
- Label axes in English
- Use a professional plotting style
- Save as both PNG and the Python .py source file

✅ Verification Checkpoint: Check: Does the midband gain match your Step 2 calculation? Are the −3dB points at the correct pole frequencies? Does the phase shift from 0° to −180° (or equivalent) correctly?

檢查：中頻增益是否與步驟 2 計算結果一致？−3dB 點是否在正確的極點頻率？相位變化是否正確？

✍ Student Notes:

(Record key values and observations here)

 

 

5. Advanced Prompting Tips / 進階提示技巧

5.1 Single-Prompt Alternative

Once you are comfortable with the step-by-step approach, you may combine everything into a single comprehensive prompt. This is efficient for experienced users but sacrifices intermediate verification:

Given the following CE amplifier: VCC=12V, R1=47k, R2=10k, RC=4.7k, RE=1k,
RS=1k, RL=10k, β=150, Cπ=20pF, Cμ=2pF, C1=1μF, C2=4.7μF, CE=47μF

Solve for: (1) DC bias point including exact base current,
(2) midband voltage gain, (3) Miller capacitance & high-frequency poles,
(4) all low-frequency poles using SCTC, and (5) generate Python Bode plots
(gain & phase vs. Hz). Do NOT neglect base current.

5.2 Key Prompting Principles

5.3 When AI Makes Mistakes

AI models can and will make computational errors. If you detect an error in the output, try these correction prompts:

• "I believe the calculation of g_m is incorrect. V_T = 26mV at room temperature. Please recalculate."
• "The Miller capacitance should use |A_v| not A_v. The gain magnitude is [your value]. Please redo this step."
• "You forgot to include the parallel combination of R₁‖R₂ in the input resistance. Please recalculate."

Learning to identify and correct AI errors is an essential professional skill.

學會識別和糾正 AI 錯誤是一項重要的專業技能。

6. Deliverables / 繳交項目

Submit the following items as a single PDF or Word document:

7. Grading Rubric / 評分標準

8. Important Notes / 注意事項

1. Academic Integrity: Using AI tools is explicitly permitted and encouraged in this lab. However, you must demonstrate understanding through your verification notes and reflection. Simply copying AI output without verification will result in reduced marks.
2. Language: Prompts to the AI should be written in English (this is part of the "Technical English" learning objective). Your student notes may be in either English or Chinese.
3. Collaboration: You may discuss strategies with classmates, but each student must submit their own AI conversation, verification work, and reflection.
4. Time Estimate: This lab is designed to take approximately 2–3 hours. If you encounter persistent AI errors, document them — this strengthens your reflection section.

CYCU EE — Technical English and Popular Science Communication / 科技英文與科普傳播