Lab 6.11 — dBm vs dBFS power calibration

Goal

Build a reproducible relationship between RF power in dBm and measured digital amplitude in dBFS for the ZynqSDR + AD9363 platform, using the RTL-SDR as a relative monitoring receiver and an external attenuator or calibrated reference point when available.

This lab sits between the RX-only baseline (Lab 6.6) and the first full over-the-air DDS tone observation (Lab 6.8). Its purpose is to establish a safe TX attenuation range and a documented amplitude scale before committing to OTA experiments, so that subsequent labs can choose attenuation settings with a known link budget.

Why this lab

At the current stage:

• Lab 6.6 confirmed the ZynqSDR RX path is healthy;
• TX experiments require a known safe starting point before enabling the AD9363 PA;
• the RTL-SDR already exists as an independent receiver with a stable tuner_gain_db10 setting;
• the course now needs a clear bridge between analog RF power (dBm) and digital sample amplitude (dBFS).

The table produced here lets later labs (6.8, 7.x, 11.x) set TX attenuation to a point within the RTL-SDR dynamic range without risking overload. It also makes it explicit that dBFS is not an absolute RF power unit unless a calibration reference is recorded.

Status: this is a procedure with empty result tables. No calibrated dBm reference or measured sweep is committed for it yet, so it contains no measured numbers and the course makes no calibrated power claim from it. It was previously numbered 6.7, which collided with the site's Zero-IF artifacts lab, and is now Lab 6.11.

Key definitions

Term	Meaning	Course interpretation
dB	Ratio between two values	Used for attenuation, gain, SNR and relative spectrum differences
dBFS	Decibels relative to digital full scale	FFT/sample amplitude relative to ADC or file full-scale range
dBm	Decibels relative to 1 mW	Absolute RF power at a defined impedance, usually 50 ohm

Important rule:

dBFS is relative to the digital receiver scale.
dBm is absolute RF power at a measurement plane.
A conversion from dBFS to dBm is only valid after calibration.

Calibration model

Use one calibration point measured at the same frequency, gain, sample rate, FFT processing chain and receiver settings:

P_dBm = P_ref_dBm + (P_dBFS - P_ref_dBFS)

where:

• P_ref_dBm is the known RF power at the receiver input or measurement plane;
• P_ref_dBFS is the measured digital peak or integrated power for that same reference signal;
• P_dBFS is the measured digital value for the unknown signal;
• P_dBm is the estimated RF power under the same receiver settings.

This is a relative calibration. Change any of the following and the calibration must be repeated: tuner gain, analog gain, sample rate, FFT window, RBW/bin integration method, center frequency, cables, attenuators or antenna geometry.

Executable files

File	Purpose
blocks/block_06_rf_frontend_and_ad9363/python/lab_6_3_probe_iio_context.py	confirm IIO device list before enabling TX
blocks/block_06_rf_frontend_and_ad9363/python/lab_6_8_capture_zynq_ota_tone.py	used in sweep mode with --tx-attenuation-db argument

No dedicated lab_6_11_*.py script is required: the sweep reuses lab_6_8 with a fixed DDS tone and varied TX attenuation. If a calibrated RF source or power meter is available, record the reference point in the same report table.

Measurement procedure

1. Confirm ZynqSDR IIO context is reachable:

python blocks/block_06_rf_frontend_and_ad9363/python/lab_6_3_probe_iio_context.py \
  --uri ip:192.168.40.1

1. Set up the RTL-SDR at a fixed gain (for example, tuner gain = 200 / 20 dB) and tune to the DDS tone frequency (center + offset).

2. Record the receiver settings before any sweep:

Parameter	Value
Center frequency	__________ Hz
Tone offset	__________ Hz
Sample rate	__________ sps
RTL-SDR tuner gain	__________ dB
FFT size / window	__________
Measurement plane	antenna / cable / attenuator output / other

1. If a calibrated reference is available, record one known point:

Reference RF power (dBm)	Measured digital level (dBFS)	Notes
__________	__________	same gain, frequency and FFT settings

1. For each TX attenuation value in the table below, run:

python blocks/block_06_rf_frontend_and_ad9363/python/lab_6_8_capture_zynq_ota_tone.py \
  --uri ip:192.168.40.1 \
  --center-frequency-hz 915000000 \
  --tone-offset-hz 50000 \
  --tx-hardwaregain-db <VALUE> \
  --out-iq datasets/lab6_8_zynq_ota_tone_observation/raw/tx_cal_<VALUE>dB.ci16 \
  --manifest-out datasets/lab6_8_zynq_ota_tone_observation/manifest_tx_cal_<VALUE>dB.yaml \
  --dataset-id lab6_11_tx_cal_<VALUE>dB

In this script the TX level is the AD9361 hardwaregain of the TX channel: 0 dB is maximum output and -89.75 dB is the minimum. The column called "TX attenuation" in the tables below is that value.

1. Read the RTL-SDR FFT peak for each run and record the received level in dBFS. If a valid reference point exists, estimate the corresponding dBm using the calibration model above.

Suggested attenuation sweep

TX attenuation (dB)	Expected RTL-SDR SNR	Notes
-89.75 (min TX)	noise floor	safe baseline - no signal expected
-60	low	start here
-50	moderate	standard course default
-40	good	increase only with antenna separation > 2 m
-30	high	check RTL-SDR for overload symptoms

Result table template

Run	TX attenuation (dB)	RTL-SDR peak (dBFS)	Estimated power (dBm)	SNR (dB)	Notes
baseline	-89.75	____	n/a	____	noise floor
tx_cal_-60dB	-60	____	____	____
tx_cal_-50dB	-50	____	____	____
tx_cal_-40dB	-40	____	____	____
tx_cal_-30dB	-30	____	____	____	overload check

If no calibrated dBm reference is available, leave the Estimated power (dBm) column as n/a and explicitly state that the result is a relative dBFS calibration only.

Safety rules

• Never raise the TX level above --tx-hardwaregain-db -30 without confirmed antenna separation of at least 2 m and no reflective enclosures.
• Prefer an external attenuator for any conducted or near-field measurement.
• Always restore TX to the safe state (TX LO powerdown = 1, TX attenuation = -89.75 dB) after the sweep.
• Record the dmesg tail after each TX enable to catch any xo_disable or clock_lost warnings from the AD9363 driver.
• Stop the sweep if the FFT peak stops increasing linearly with TX attenuation or if the spectrum shows clipping/overload products.

Expected outputs

Output	Content
TX calibration table (JSON or Markdown)	TX attenuation vs RTL-SDR peak power (dBFS) and optional estimated dBm
docs/assets/lab611_tx_power_calibration_table.json	machine-readable calibration reference
Report conclusion	valid receiver settings, safe TX attenuation range and calibration limitations

Suggested JSON shape:

{
  "lab": "6.11",
  "title": "dBm vs dBFS power calibration",
  "center_frequency_hz": 915000000,
  "tone_offset_hz": 50000,
  "receiver": "RTL-SDR",
  "receiver_gain_db": 20.0,
  "reference": {
    "available": false,
    "power_dbm": null,
    "measured_dbfs": null
  },
  "points": [
    {
      "tx_attenuation_db": -50.0,
      "measured_peak_dbfs": null,
      "estimated_power_dbm": null,
      "snr_db": null,
      "notes": "fill after measurement"
    }
  ]
}

Report checklist

• State RTL-SDR tuner gain used throughout the sweep.
• Record center frequency, sample rate and tone offset.
• Describe whether the measurement is OTA, conducted, attenuated or calibrated with a known reference.
• Attach the full attenuation-vs-dBFS table.
• Attach the optional dBm estimate only if a valid reference point exists.
• Identify the safe TX attenuation range for the available antenna geometry.
• Confirm TX was restored to safe state after sweep.

Engineering conclusion template

TX power calibration at ____ MHz used DDS tone offset ____ kHz and RTL-SDR
tuner gain ____ dB. Received power ranged from ____ dBFS at ____ dB attenuation
to ____ dBFS at ____ dB attenuation. A dBm conversion was / was not applied.
Reference point: ____ dBm = ____ dBFS. Estimated RF power range: ____ dBm to
____ dBm. Safe operating range for subsequent labs: TX attenuation between
____ dB and ____ dB. TX was restored to safe state: TX LO powerdown = 1,
attenuation = -89.75 dB.