So close! (RTL-SDR & Raspberry Pi 4B)

[allen lutins]

I'm a newbie when it comes to my Raspberry Pi 4B and the Linux environment (been using SDRs on a Windows PC for years). I've installed RTL_433 (seemingly successfully), but when I execute it (with an RTL-SDR V3 Blog plugged in), it doesn't appear to fully function. I get:

Found Rafael Micro R820T tuner
Exact sample rate is: 250000.000414 Hz
[R82XX] PLL not locked!

And then it stops. According to the documentation I read, I think it should be followed by:

Sample rate set to 250000 S/s.
Tuner gain set to Auto.
Tuned to 433.952MHz.
baseband_demod_FM: low pass filter for 250000 Hz at cutoff 25000 Hz, 40.0 us

I couldn't find any kind of error file that might have been generated. How would I go about troubleshooting this? Thanks!

[allen lutins]

Guess I should have provided the command line I''m using:

rtl_433 -M level -f 433952500

[Christian Z.]

Is that a recent version? It should start with a line about the version and options, then conf files, protocols. Only then will the device be started.

Add   -M level  to see level messages at short intervals.

[allen lutins]

I've made a lot of progress, but I'm not getting any output. I can see clear pulses at 433.945 MHz (and nearby) using SDRConsole, so I'm running rtl_433 using  rtl_433_64bit_static.exe -f 433952000 -A -Y level=0. When I upload a CU8 file to https://triq.org/pdv/ I see the pulses, but they are clearer if I adjust the Gain Range from +30 (default) to +24 (see below). I suspect I need to use -g to change the gain, but neither kalilinuxtutorials.com/rtl_433 nor triq.org provides recommended values. I tried every value from 0 through 5, and then some higher values (15, 20) but none of the resultant CU8 files yielded a clearer spectrogram than leaving it off. Perhaps I need to employ some other options?

Thanks for any guidance you can provide!

[Christian Z.]

That's really faint. You need a much stronger signal. Is that a E4000 RTL-SDR? They have auto-gain problems. Maximum gains are 49dB for the R820, 42dB for the E4000, so -g 50 should be a safe choice.

[allen lutins]

I'm using an RTL-SDR V3 blog. You may be right about the signal being too faint. It's a lot stronger on my RSP1A, but (correct me if I'm wrong) rtl_433 does not appear to support the SDRPlay devices. (Btw, I'm picking these up using a roof-mounted discone; I suspect some are TPM's, but I'd really like to know for sure what they all are!)

[Christian Z.]

rtl_433 works with RTL-SDR and/or SoapySDR, which support the SDRplay devices!

The RTL-SDR V3 blog should be R820T2, the auto-gain should work then.

[allen lutins]

OK, Let's try using the RSP1A. I installed the latest version of Pothos (which includes SoapySDR). I ran Zadig, which identifies the RSP1A on USB ID "1DF7 3000", then soapysdrutil -f, which identifies the RSP1A as device 3. Next, I ran soapysdrutil --make="sdrplay", which apparently (based on the generated output) completed successfully.

At this point, I think I'm missing a step, because running rtl_433_64bit_static.exe -d sdrplay -f 433952000 -A -Y autolevel instantly returns me to a command prompt. Do I need to put something together in Pothos Flow to make this work?

Once again, I feel like I'm on the verge of success! Any idea what step(s) I'm missing? Thanks again to any and all (and especially Christian Z!) who can afford the time to come to my assistance :)

[Christian Z.]

Looks good so far. The "static" version does not include the dynamic SoapySDR loading. The first line of output will list all available inputs.

[allen lutins]

The only other executable in the folder is rtl_sdr.exe - when I try running that (with the same arguments) I get a Windows pop-up error dialog box reading, The procedure entry point rtlsdr_cancel_async could not be located in the dynamic link library. Not sure which file that refers to (SoapySDR.dll, librtlsdr.dll & libusb-1.0.dll reside in that folder); maybe there's a missing dll specific to the RSP1A?

[Christian Z.]

It's difficult with path and DLLs on Windows and I don't know a lot about that.

The error is related to librtlsdr.dll -- no idea why rtlsdr_cancel_async() would be missing?

Best to copy rtl_433.exe to where all the Pothos exe's are, and don't copy the DLL's. You must use the Pothos SoapySDR.dll. The one with rtl_433.exe does not support the SoapySDR modules.

[allen lutins]

Well Christian, I owe you a beer or a coffee or a juice - thanks for all your help - but I'm about to throw in the towel here. I copied rtl_433.exe to the PothosSDR "bin" folder where all the executables are, tried running from there (command prompt window with admin privs) and received a Windows error complaining about librtlsdr.dll not being found; so I also copied librtlsdr.dll, but then when I ran rtl_433 again I got the same "The procedure entry point..." error as before.

I have plans to get WSJT-X running on my Raspberry Pi, so I'll probably bail on this project, and start concentrating my efforts on the more important task - unless someone has something substantially more to add...

[Christian Z.]

I've setup a Win 10 VM, then installed Pothos 2021.07-25 (check set path for user), then SDRplay API Drivers 3.11, then unpacked the rtl_433-win-x64 zip, deleted the SoapySDR.dll an run rtl_433.exe -d driver=sdrplay

Works fine here. Sorry I can't be more help, I don't use Windows. I guess you have other stuff installed that interferes, e.g. a different librtlsdr.dll somewhere?

[allen lutins]

Actually, I'm not sure that librtlsdr.dll is implicated - the exact error I'm getting is "The procedure entry point rtlsdr_cancel_async could not be located in the dynamic link library C:\Program Files\Pothos\bin\rtl_433.exe".

I sure appreciate all the time you've spent helping me, Christian, and wish someone else would chime in with some advice on how to further troubleshoot this.

[Christian Z.]

The function rtlsdr_cancel_async is in librtlsdr.dll or rtlsdr.dll, you seem to load an old dll without that function. Search your system for other, old/incompatible dlls.