|
Reader Question · Raspberry Pi 5 · RTL-SDR Blog V4L
RTL-SDR Blog V4L (Lite) on a Raspberry Pi 5: Drivers Installed, Now What?
What the V4L changes, the one test that tells you whether the new driver is really the one in use, and a clean rebuild for when it is not.
|
A reader who previously got an RTL-SDR Blog V4 working on a Raspberry Pi 5 has now moved to the newer RTL-SDR Blog V4L (Lite). He has installed the drivers and updated his software following the vendor’s page, but the dongle is not behaving. His note arrived without a description of the symptom, so this post covers the faults that account for nearly every “I followed the instructions and it still does not work” report with this dongle, in order of likelihood, plus the output to send if none of them fits.
|
Short version
The V4L needs a V4L-aware librtlsdr. The vendor states that without updated drivers it “will simply not function correctly, and you will receive no signal.”[1] On a Pi the usual failure is not that the new driver was never built — it is that the old library is still the one being loaded, by the OS package, by a second copy in Run |
1. What is different about the V4L
The V4L keeps the V4 architecture (RTL2832U, 1 ppm TCXO, built-in HF upconverter, software-switchable bias tee) but swaps the tuner. Per the retailer’s comparison it uses a Rafael Micro R828S with two RF inputs instead of the V4’s R828D with three.[2] That matters for two reasons:
- Detection is by EEPROM string, not by tuner ID. The driver author notes the R828S presents itself on I²C as an R820T/R860, so the code looks for the manufacturer string
RTLSDRBlogand the product stringV4Lto know what it is dealing with.[3] An old library therefore sees a generic R820T-family tuner — it will open, it will stream samples, and the spectrum will be wrong or empty. - Less front-end filtering. With two inputs the V4L is diplexed rather than triplexed, so out-of-band rejection is reduced; the retailer suggests external band-pass or notch filters in strong RF environments.[2]
| Item | RTL-SDR Blog V4 | RTL-SDR Blog V4L |
|---|---|---|
| Tuner | Rafael Micro R828D | Rafael Micro R828S |
| RF inputs | 3 (triplexed front end) | 2 (diplexed front end) |
| Driver needed | V4-aware librtlsdr | V4L-aware librtlsdr (newer than V4-only builds) |
| EEPROM product string | Blog V4 |
Blog V4L |
| HF reception | Direct tune; no direct sampling | Direct tune; no direct sampling |
| Bias tee | 4.5 V, software-switched | 4.5 V, up to 180 mA, software-switched |
The vendor says the updated drivers are backward compatible with the V4, V3 and generic dongles, so there is no downside to moving your whole system to the new library.[1]
2. The first test: is the new driver the one in use?
rtl_test -t |
A working V4L should report the dongle and then a line ending in V4 Lite Detected. The Osmocom tree that carries the V4L patch prints that string in the same way it prints “V4 Detected” for the older dongle.[4][3] The signature of the old library is a tuner line with no “Detected” line after it. I have not been able to confirm the exact tuner line the old library prints for an R828S, so judge by the missing “V4 Lite Detected” line rather than by the tuner name.
|
Why this test is reliable
|
3. Find out which library is actually loaded
ldconfig -p | grep rtlsdr ldd $(which rtl_test) | grep rtlsdr dpkg -l | grep -i rtlsdr ls -l /usr/local/lib | grep rtlsdr ls -l /usr/local/bin | grep '^.* rtl_' |
You want exactly one librtlsdr.so.0 resolving, and it should be the one you built. Typical bad states:
- Two copies. The Debian package puts the library in
/usr/lib/aarch64-linux-gnu; a sourcemake installputs another in/usr/local/lib. Which one wins depends on loader order, and different programs can pick different copies. The retailer’s troubleshooting table lists the same fault and the sameldconfig -p | grep rtlsdrcheck.[2] - Only the distribution package. Raspberry Pi OS Bookworm ships
librtlsdr0 0.6.0-4, which predates the V4 family altogether.[5] Trixie ships 2.0.2-2.[6] Treat any distribution package as suspect for a V4L unlessrtl_testproves otherwise. - An application that bundles its own copy (some Docker images, AppImages and pre-built decoders). Rebuilding the system library does not help these; they need an updated image or build.
4. Clean rebuild on a Pi (package method)
The vendor recommends building .deb packages on Debian-package-based images such as OpenWebRX+, FlightAware and ADS-B feeders, so that package-dependent software uses the new library.[1] Take it in this order.
Step 1 — remove the old library and any hand-installed copies
sudo apt purge '^librtlsdr' sudo rm -rvf /usr/lib/librtlsdr* /usr/include/rtl-sdr* /usr/local/lib/librtlsdr* \ /usr/local/include/rtl-sdr* /usr/local/include/rtl_* /usr/local/bin/rtl_* |
|
Read the apt removal list
On a system running OpenWebRX+ or a feeder, |
Step 2 — build dependencies
sudo apt update sudo apt install libusb-1.0-0-dev git cmake pkg-config build-essential debhelper |
Step 3 — build and install the packages from the Osmocom tree
git clone https://github.com/osmocom/rtl-sdr cd rtl-sdr sudo dpkg-buildpackage -b --no-sign cd .. sudo dpkg -i librtlsdr0_*.deb librtlsdr-dev_*.deb rtl-sdr_*.deb sudo apt-mark hold librtlsdr0 librtlsdr-dev rtl-sdr |
The vendor’s guide clones the Osmocom repository for Linux.[1][4] The apt-mark hold line is my addition: without it, a routine apt upgrade can quietly put the older distribution library back, which is also the cause of the “worked until the last update” reports on ADS-B feeders.[2] Package names can change between Debian releases; install whatever the build actually produced in the parent directory.
Step 4 — stop the kernel DVB-T driver grabbing the dongle
echo 'blacklist dvb_usb_rtl28xxu' | sudo tee --append /etc/modprobe.d/blacklist-dvb_usb_rtl28xxu.conf sudo reboot |
This matters more on the V4L than a plain dongle: the vendor warns the legacy DVB-T driver turns the bias tee on by default on Linux.[1]
Step 5 — verify
rtl_test -t |
|
Success looks like
The dongle is found, a V4 Lite Detected line appears, and the test finishes without |
5. Checks that bite even with the right driver
- Do not touch the EEPROM strings. The driver identifies the dongle from manufacturer
RTLSDRBlogand productBlog V4L. If they have been changed, restore them withrtl_eeprom -m "RTLSDRBlog"andrtl_eeprom -p "Blog V4L", then replug.[1] Runningrtl_eepromwith no options only reads and prints them. - HF needs no direct sampling. Tune to the HF frequency directly. Enabling direct sampling on a V4-family dongle yields nothing.[1] Carry-over settings from a V3 profile are a common cause.
- Disable “IF AGC” in SDR#, HDSDR and SDRuno. The vendor says it does not work correctly on the V4L.[1]
- Bias tee only into a powered device. 4.5 V at up to 180 mA. The vendor warns against connecting a DC-shorted antenna with the bias tee on; days of that can permanently damage the LDO or fuse. Control it with
rtl_biast -b 1/rtl_biast -b 0.rtl_eeprom -b 1forces it always on and only works with the RTL-SDR Blog drivers; I would not use it unless the LNA is rated for it.[1] - Pi 5 USB power. Without a recognised 27 W supply the Pi 5 caps USB current, and a bias-tee LNA on top of the dongle’s own draw can brown it out. Check
vcgencmd get_config usb_max_current_enableand the Raspberry Pi documentation before blaming the dongle.[7] Plug directly into a Pi port or a powered hub. - Strong local signals. Reduced filtering means FM broadcast and pager transmitters overload the V4L more easily than the V4. Reduce gain before adding gear; if that is not enough, add an FM-broadcast notch or band-pass filter.[2]
6. Software status
The vendor lists SDR++ (latest builds), SDR#, CubicSDR and the Android SDR Driver app as updated for the V4L, and gives Debian-package update steps for OpenWebRX and similar Pi images. Arch’s AUR rtl-sdr-git package had not been updated when the guide was written, and Osmocom/SoapySDR users may need to reinstall that software after updating the library.[1] OpenWebRX+ on a Pi uses the system library, so the package rebuild in section 4 is the fix there.[8]
7. Symptom table
| Symptom | Most likely cause | What to do |
|---|---|---|
| No signal anywhere, dongle opens fine | Old librtlsdr in use; no “V4 Lite Detected” | Section 3, then rebuild per section 4 |
“Device busy” or usb_open error |
DVB-T kernel driver or missing udev permissions | Blacklist dvb_usb_rtl28xxu, reboot; confirm udev rules and plugdev membership |
Works in rtl_test, not in the app |
App loads a second or bundled library | ldd the app binary; remove duplicates; restart service |
| HF dead, VHF fine | Direct sampling enabled | Switch it off; tune the real HF frequency |
| Worked, then died after an update | Package update restored the older library | Reinstall built .debs, hold them, reboot |
| Random USB drops, bias tee in use | Pi 5 current cap or noisy supply | Proper 27 W supply, powered hub, check current limit |
| Spectrum full of junk, strong FM | Overload (reduced V4L filtering) | Lower gain; add notch or band-pass filter |
8. If it still fails: what to send
Without the symptom I can only guess, so collect this and reply with it:
uname -a; cat /etc/os-release | head -3 rtl_test -t 2>&1 | head -20 rtl_eeprom 2>&1 | head -20 ldconfig -p | grep rtlsdr dpkg -l | grep -i rtlsdr lsmod | grep -i dvb lsusb | grep -i -E 'rtl|0bda' vcgencmd get_throttled |
Also say which application you are using, what you see (nothing, wrong frequency, noise, crash), and whether the dongle is on a Pi port directly or via a hub.
Sources
- [1] RTL-SDR Blog, RTL-SDR Blog V4L (Lite) Users Guide — www.rtl-sdr.com/v4l/. Driver requirement, Linux build steps, Debian-package method, DVB-T blacklist, bias-tee behaviour and warning, EEPROM strings, HF mode (no direct sampling), IF AGC note.
- [2] SDRstore.eu, RTL-SDR V4L Setup Guide — www.sdrstore.eu/rtl-sdr-v4l-setup-guide-windows-linux-raspberry-pi-sdrpp-sdrsharp-adsb/. Third-party (retailer) guide: V4L vs V4 comparison, R828S tuner, specifications, Raspberry Pi package-build sequence and troubleshooting table.
- [3] Carl Laufer, [PATCH] Add RTL-SDR Blog V4L Support, osmocom-sdr mailing list (Aug 2026) — lists.osmocom.org/hyperkitty/list/[email protected]/2026/5/. R828S presents itself on I2C as an R820T/R860, so detection relies on the EEPROM strings “RTLSDRBlog” and “V4L”; two-input switching; tracking filter disabled in HF upconversion mode.
- [4] Osmocom project, rtl-sdr source repository — gitea.osmocom.org/sdr/rtl-sdr. Canonical upstream (GitHub mirror at github.com/osmocom/rtl-sdr). The V4L guide builds from this tree.
- [5] Debian package tracker, rtl-sdr in bookworm — packages.debian.org/source/bookworm/rtl-sdr. Bookworm ships librtlsdr0 / rtl-sdr 0.6.0-4.
- [6] Debian package tracker, librtlsdr0 in trixie — packages.debian.org/trixie/librtlsdr0. Trixie ships 2.0.2-2 (packaging by the Debian Hamradio Maintainers).
- [7] Raspberry Pi Ltd, Raspberry Pi documentation — www.raspberrypi.com/documentation/. Pi 5 USB current limits and the
usb_max_current_enableoption. - [8] Marat Fayzullin, OpenWebRX+ Home Page — fms.komkon.org/OWRX/. OpenWebRX+ installation channels and images.
Credits
- The RTL-SDR Blog V4L hardware, drivers and user guide are by RTL-SDR Blog (Carl Laufer); the V4L driver patch was submitted to the Osmocom list by him.
- librtlsdr and the
rtl_*tools are the work of the Osmocom project (Steve Markgraf and contributors). - The Raspberry Pi package-build sequence and troubleshooting matrix draw on the guide published by SDRstore.eu; independent verification on a V4L unit is welcome.
- OpenWebRX+ is Marat Fayzullin’s fork (fms.komkon.org/OWRX). Debian packaging of rtl-sdr is by the Debian Hamradio Maintainers. Raspberry Pi OS and documentation are by Raspberry Pi Ltd.
|
Mike Peace VK6ADA
Last reviewed 22 November 2026. Corrections and test results from real V4L units are welcome.
|