---
Project: LNA-24G Characterization
Author: Hardware Engineering
Date: 2026-08-12
Revision: 1.0
---

# RF Low Noise Amplifier (LNA) Test Report

## 1. Test Setup Architecture

The following block diagram illustrates the hardware pipeline for evaluating the LNA gain (S21) and noise figure (NF). The setup uses a calibrated noise source and a spectrum analyzer with a noise figure measurement personality.

```mermaid
flowchart LR
    NS[Noise Source] --> ENR[ENR Table]
    NS --> LNA[LNA Under Test]
    LNA --> BP[Bandpass Filter]
    BP --> SA[Spectrum Analyzer]

    classDef hardware fill:#eef,stroke:#333,stroke-width:2px;
    class NS,LNA,SA hardware;
```

> [!WARNING]
> Maximum continuous input power to the Spectrum Analyzer RF port must not exceed +10 dBm to prevent front-end saturation and permanent mixer damage.

## 2. Theoretical Cascaded Noise Figure

When cascading the LNA with the measurement equipment, the total system noise figure is calculated using the Friis formula. To isolate the LNA performance, we must de-embed the spectrum analyzer's noise floor.

The total noise factor $F_{total}$ is defined as:

$$
F_{total} = F_{LNA} + \frac{F_{SA} - 1}{G_{LNA}}
$$

Where:
* $F_{LNA}$ is the noise factor of the amplifier under test (linear).
* $F_{SA}$ is the noise factor of the spectrum analyzer (linear).
* $G_{LNA}$ is the linear gain of the amplifier.

## 3. Swept Measurement Results

The table below summarizes the measured Gain and Noise Figure across the target 2.4 GHz ISM band. Data is averaged over 100 sweeps with a resolution bandwidth (RBW) of 1 MHz.

| Frequency (GHz) | S21 Gain (dB) | Noise Figure (dB) | Test Status |
| :--- | :--- | :--- | :--- |
| 2.400 | 18.2 | 1.15 | PASS |
| 2.425 | 18.4 | 1.12 | PASS |
| 2.450 | 18.5 | 1.10 | PASS |
| 2.475 | 18.3 | 1.14 | PASS |
| 2.500 | 17.9 | 1.21 | PASS |

\pagebreak

# Digital Demodulator: Firmware State Machine

## 1. State Machine Overview

The signal acquisition and tracking logic for the digital QPSK demodulator is implemented as a finite state machine (FSM). This logic runs in the primary DSP thread.

```mermaid
stateDiagram-v2
    [*] --> IDLE
    IDLE --> AGC_SETTLE : RSSI > Threshold
    AGC_SETTLE --> PLL_LOCK : AGC Locked
    PLL_LOCK --> DEMODULATING : Phase Locked
    DEMODULATING --> IDLE : Signal Lost (Timeout)
    PLL_LOCK --> IDLE : PLL Timeout
```

## 2. NCO Tuning Word Calculation

The Numerically Controlled Oscillator (NCO) frequency $f_{out}$ is determined by the 32-bit Phase Accumulator width and the system clock $f_{clk}$. The Tuning Word ($TW$) is calculated as follows:

$$
TW = \text{round}\left( \frac{f_{out} \cdot 2^{32}}{f_{clk}} \right)
$$

For a system clock of 122.88 MHz and a target IF of 10.7 MHz, the firmware calculates and updates this tuning word dynamically within the PLL loop filter constraint.

\pagebreak
-

## 3. Baseband Register Map

| Register Name | Address | Width | Access | Description |
| :--- | :--- | :--- | :--- | :--- |
| `DSP_CTRL` | `0x0000` | 32-bit | R/W | DSP core control bits (Enable, Reset) |
| `FSM_STATUS` | `0x0004` | 32-bit | R/O | FSM current state and lock flags |
| `NCO_TW_REG` | `0x0010` | 32-bit | R/W | Phase accumulator tuning word |
| `AGC_GAIN` | `0x0014` | 32-bit | R/O | Current variable gain amplifier state |
