Radios
USB radios
Driver requirements, quirks, and platform setup for USB and Ethernet SDRs.
On this page
See Supported radios for specs, charts, and sdrmm --doctor.
RTL-SDR
Built-in driver supports R820T / R828D tuners only (E4000, FC0012/13, FC2580 refused). Rates: 225–300 kHz or 900 kHz–3.2 MHz; BW: 290 kHz–8 MHz. Direct sampling (off/i/q) is hidden on Blog V4 / V4 Lite (built-in HF upconverter). Windows needs WinUSB via Zadig.
Serials
Many dongles ship with 00000001. SDRmm prompts once to write a unique serial to EEPROM (close the dongle, write, replug) so settings and calibration stick across USB ports. Clones without EEPROM cannot persist a serial.
KrakenSDR
Five shared-clock tuners (KrakenSDR <hub> lane 1–5; KerberosSDR has four), max 2.56 MS/s, no vendor Pi image needed. Wire all lanes into one Array in antenna order (tier locks to Shared clock).
- Lane 1 controls the noise source and all five bias tees, so it must stay open and wired.
- Calibration: Array triggers the noise source on retune/gain change and every minute (antennas disconnect and processors pause). Gain steps shift phase up to ~10°.
- Errors:
Noise clips, lower gain(> 75% clipped);Noise needs all lanes(incomplete bank). Check Coherent banks insdrmm --doctor.
Tested on hardware provided by KrakenRF.
HackRF
Built-in. LNA (8 dB steps), VGA (2 dB steps), +14 dB Amp (shared RX/TX), baseband BW, Bias tee. Give 20 MS/s its own USB 2 bus to avoid silent drops.
Airspy
Built-in (omit airspy/airspyhf features to force SoapySDR).
- R2 / Mini: 24 MHz–1.8 GHz; LNA/Mixer/VGA in firmware steps; AGC on LNA, mixer, or both; bias tee.
- HF+ / Discovery: 1 kHz–31 MHz & 60–260 MHz; Amp (+6 dB), Attenuator (6 dB steps to -48 dB), low/high AGC, factory PPM init. Centres below 180 kHz (84 kHz narrow) clamp to that floor while still covering lower signals.
Tested on hardware provided by Airspy.
FUNcube Dongle Pro+
Built-in, fixed 192 kS/s. LNA/Mixer toggles, IF gain 0–59 dB (overload above ~25 dB with LNA+Mixer on), IF BW 200 kHz–8 MHz, RF filter, bias tee, PPM.
OS setup (Linux udev, Windows Zadig, macOS Microphone permission)
- Linux: detaches sound/HID drivers over raw USB. Add udev rule:
SUBSYSTEM=="usb", ATTRS{idVendor}=="04d8", ATTRS{idProduct}=="fb31", MODE="0660", GROUP="plugdev" - Windows: in Zadig, enable List All Devices, disable Ignore Hubs or Composite Parents, pick FUNcube Dongle V2.0 (Composite Parent), install WinUSB.
- macOS: streams via Core Audio. Allow Microphone for SDRmm (or terminal) in System Settings, or samples read as silence.
Tested on hardware provided by FUNcube Dongle.
ESP-SDR
Receives via ESP32 Wi-Fi radio flashed with ESP-SDR (browser flasher). Reliable on 2.4 GHz (5 GHz on C5); 80/40/16 MS/s, 8 or 10 bits.
- Serial: captures bursts (4096 samples default, ~0.1 s gap at 921.6 kBd). Auto-probes 2 MBd, 1 MBd, 921.6 kBd (CP2102 needs
CONFIG_ESP_SDR_UART_BAUD=921600). Linux requiresdialoutgroup. - Ethernet (ESP32-S31
esp32s31-stream): gapless 4–40 MS/s at 8 bits (2150–2800 MHz); add via Network → ESP-SDR (Gigabit needed for 40 MS/s).
AntSDR
Talks directly to Pluto firmware iiod (no libiio; UHD images unsupported, and e310v2 does not run on E310). E310 RX1/RX2 share one synthesizer (phase coherent) and lock antenna/TX ports in firmware. Auto-discovers ant.local / 192.168.1.10. USB works out of the box (Windows needs PlutoSDR drivers).
Direct Ethernet setup and faster UDP streaming (sdrmm-iqlinkd)
Direct Ethernet: board defaults to 192.168.1.10. Assign your Ethernet port static IPv4 192.168.1.100/24 with no gateway (so internet stays on Wi-Fi). On Linux:
nmcli connection add type ethernet ifname <port> con-name antsdr ipv4.method manual ipv4.addresses 192.168.1.100/24
For router LANs outside 192.168.1.x, connect over USB and edit ipaddr_eth/netmask_eth in config.txt (SSH login: root / analog).
Faster UDP (sdrmm-iqlinkd, ~20 MS/s vs ~15 MS/s iiod): auto-detected on UDP port 30432:
rustup target add armv7-unknown-linux-musleabihf
cargo build --release --target armv7-unknown-linux-musleabihf -p sdrmm-iqlinkd
scp -O target/armv7-unknown-linux-musleabihf/release/sdrmm-iqlinkd root@192.168.1.10:/mnt/jffs2/
ssh root@192.168.1.10 'echo "/mnt/jffs2/sdrmm-iqlinkd > /tmp/iqlinkd.log 2>&1 &" > /mnt/jffs2/autorun.sh; reboot'
Tested on hardware provided by MicroPhase.
PlutoSDR and other AD936x boards
Direct iiod over USB or Ethernet (auto-discovers pluto.local, 192.168.2.1, ant.local, 192.168.1.10). Rates ~260 kS/s–61.44 MS/s (FPGA decimates below 2.08 MS/s); 2×2 boards are phase coherent.
- 8 fast-lock profiles cut repeat retunes from ~10 ms to ~2 ms; TX synth sleeps when idle.
- Extra chips expose per-lane AGC thresholds, quadrature/DC tracking, dual DDS TX Tone, chip state/calibration, external LO, and 1 Hz temperature/RSSI.
- Linux requires
libiioudev rules (sdrmm --doctorchecks them).
SDRplay
Install SDRplay API 3.15+ and run sdrplay_apiService (no SoapySDR needed; no vendor code bundled). RF (LNA) and IF (0–39 dB) sliders both increase gain upward; AGC runs IF at 5/50/100 Hz.
- RSPduo: Tuner 1, Tuner 2, Dual Tuner (≤ 2 MS/s each), Master, and Slave appear as separate entries (busy modes hidden).
- Containers: see Docker SDRplay. Remote RSPs: see SDRconnect.
Aaronia SPECTRAN V6
Supports SPECTRAN V6 PLUS and SPECTRAN V6 ECO via RTSA-Suite PRO (Windows/Linux only; close RTSA-Suite before opening in SDRmm; run setup_udev.sh on Linux). Set SDRMM_AARONIA_DIR for custom paths.
- V6 PLUS: selectable
Rx1/Rx2inputs, host-chosen receiver clocks up to 491.52 MS/s (80–245 MHz RTBW per model). - V6 ECO: fixed 61.44 MS/s clock (40 MHz RTBW), single RX input.
- GAIN is inverted reference level (
20 dB=-20 dBmfull scale). Sample rate is1.5 × span(two-thirds clean); TX is half-duplex (0 dBm full scale before TX gain).
Dragon Labs CR-8
8 phase-coherent lanes (iq1–iq8), 24 MHz–1.766 GHz, fixed 12.5 MS/s, internal/external clock. Excluded from packaged builds: compile with --features cr8, install the vendor SDK (override path with SDRMM_DLCR_LIBRARY), and verify via sdrmm --doctor.
SoapySDR
Covers bladeRF (SoapyBladeRF), LimeSDR (SoapyLMS7), USRP (SoapyUHD), and remote servers (SoapyRemote). Loads at runtime (0.8 or 0.9); built-in drivers take precedence. Probed out-of-process for crash isolation (SDRMM_SOAPY_PROBE=in-process for debugging).
Install commands and custom library environment variables
| System | Core | Example module |
|---|---|---|
| Debian / Ubuntu / RPi OS | sudo apt install libsoapysdr0.8 |
soapysdr-module-bladerf |
| Fedora | sudo dnf install SoapySDR |
soapy-uhd |
| Arch | sudo pacman -S soapysdr |
soapybladerf |
| macOS | brew install soapysdr |
pothosware/pothos/soapybladerf |
| Windows | PothosSDR | Included |
| NixOS | soapyPlugins |
Included via override |
Custom paths: SDRMM_SOAPY_LIBRARY, SOAPY_SDR_ROOT, SOAPY_SDR_PLUGIN_PATH. Check detection with SoapySDRUtil --find.