Block 6 — RF frontend workflow¶

This block moves the student from digital DSP/FPGA processing to a real RF bench: frequency plan, levels, bandwidth, gain, safe connection and external signal observation.

Main engineering chain¶

flowchart LR
    BB[Baseband signal] --> DBG[Debug waveform]
    DBG --> DAC[DAC / TX digital path]
    DAC --> TXRF[AD9363 TX RF chain]
    TXRF --> PATH[Coax / attenuator / antenna]
    PATH --> RXRF[Receiver RF chain]
    RXRF --> ADC[ADC / IQ capture]
    ADC --> ANALYSIS[FFT, levels, SNR, overload check]
    ANALYSIS --> REPORT[RF report]

Block 6 inputs¶

Artifact from earlier blocks	How it is used in the RF experiment
Sample rate	defines digital bandwidth and FFT interpretation
Digital mixer / NCO	defines signal offset in baseband
FIR/decimator	defines useful bandwidth and out-of-band rejection
IQ metadata	records capture parameters for reproducibility
HDL/AXIS interface	connects FPGA streaming path to the RF chain
Debug waveform format	helps find the signal on air and verify the chain through known signatures

Minimum RF discipline¶

Before enabling transmission, record:

Parameter	Example	Why it matters
TX center frequency	915 MHz	transmitter LO frequency
RX center frequency	915 MHz	receiver LO frequency
Sample rate	2.4 MS/s	digital observation bandwidth
RF bandwidth	2 MHz	analog chain bandwidth
TX gain / attenuation	-20 dB	protects receiver from overload
External attenuation	20–60 dB	safe cabled connection
Expected tone offset	100 kHz	frequency-plan verification
Expected preamble/sync	`1010...` / `0xA5A55A5A`	packet detection verification

Safe first-bench connection¶

flowchart LR
    TX[AD9363 TX] --> ATT[30-60 dB attenuator]
    ATT --> RX[RTL-SDR / receiver]
    RX --> PC[HDSDR / Python / MATLAB]

RF safety

Do not connect TX directly to a sensitive receiver without attenuation. Start with minimum TX gain, external attenuation and overload monitoring in the spectrum.

Debug signal for on-air search¶

For RF bring-up, transmit a dedicated debug frame instead of an arbitrary stream:

silence → lead-in tone → preamble → sync word → header → training → payload/PRBS → CRC → silence

Minimum TX modes worth supporting:

pure tone — find the transmission and verify the frequency plan;
preamble only — verify correlation-based search;
repeated sync word — verify bit/byte order;
PRBS packet — compute BER;
amplitude sweep — find overload and tune gain/attenuator;
frequency sweep — verify NCO sign, I/Q and frequency axis;
two-tone/multitone — verify linearity, bandwidth and spurs.

Detailed guide: Debug Waveform Design for SDR Hardware Bring-up.

Normal operating signs¶

Sign	Meaning
One stable peak	frequency plan is correct
Packet bursts are visible on the waterfall	frame period and silence work correctly
Preamble correlation gives a stable peak	the signal is found automatically
Sync word is found after the preamble	bit/byte ordering is consistent
No broad flat top	no obvious ADC/RF overload
Stable noise floor	gain is reasonable
Side spurs below the useful signal	NCO/LO/quantization do not dominate
Gain changes level predictably	chain is not saturated

Overload signs¶

Symptom	Possible cause	Action
Wide flat spectral top	ADC/RF overload	reduce gain / add attenuation
Many harmonics	amplifier saturation	reduce TX power
Noise floor rises with signal	nonlinearity or AGC	disable AGC, reduce level
Peak does not change with gain	clipping/limiting	check levels and cables
Signal drifts in frequency	LO offset / drift	measure frequency error
Correlation peak exists, but CRC fails	timing/CFO/bit-order error	check sync, training and demodulator

Basic RF report¶

Every Block 6 experiment should contain:

RF experiment goal;
connection diagram;
frequency plan;
gain/bandwidth settings table;
debug waveform or tx_mode description;
screenshot or FFT plot of normal operation;
preamble/sync detection result;
overload signs or their absence;
IQ metadata file;
engineering conclusion.

Connection to later blocks¶

Block 6 prepares the real RF bench for:

TX/RX chain experiments;
modulation and synchronization;
recording and analysis tools;
integrated SDR project;
final report with measured IQ data.