search

AlbrechtL / welle.io

A DAB/DAB+ Software Defined Radio (SDR)
https://www.welle.io
Other
629 stars 117 forks source link

welle-cli - gain option has no effect #812

Open Hoerli1337 opened 2 months ago

Hoerli1337 commented 2 months ago

Hi! I've been playing around with the project for a week now. I really like it so far.

I have an RTL-SDR v5 stick from nooelec. This is connected to a Raspberry Pi 3B+. It runs RaspberryOS 64bit as normal.

When trying out the best settings, I noticed that the gain parameter (-g) has absolutely no effect. I live in a place where DAB+ is only possible with an external antenna and optimal positioning. I wanted to fine-tune everything on the software side. Bildschirmfoto vom 2024-08-17 23-19-58

I can use manual amplification via the GUI.

Is this a bug or deliberately not possible?

andimik commented 2 months ago

I have the same stick, 

rtl_test
Found 1 device(s):
  0:  Realtek, RTL2838UHIDIR, SN: 00000001

Using device 0: Generic RTL2832U OEM
Found Rafael Micro R860 or 820T/2 tuner
Supported gain values in dB (29): 0.0 0.9 1.4 2.7 3.7 7.7 8.7 12.5 14.4 15.7 16.6 19.7 20.7 22.9 25.4 28.0 29.7 32.8 33.8 36.4 37.2 38.6 40.2 42.1 43.4 43.9 44.5 48.0 49.6 
Supported bandwidth values in kHz: 290 375 420 470 600 860 950 1100 1200 1300 1503 1600 1753 1800 1953 2200 3000 5000 6000 7000 8000
Sampling at 2048000 S/s.

but if you use a too high manual gain for local channels, there is definitely a warning:

RTL_SDR: ADC overload. Maybe you are using a too high gain. Info: ADC overload. Maybe you are using a too high gain.

But it seems that a too high gain will set gain to 0, as the value is "unknown"

see https://github.com/AlbrechtL/welle.io/blob/5481d57e728e03441c83b5586c2017f858f1dc31/src/input/rtl_sdr.cpp#L195-L197

I have tested with some gain settings:

local mux (8 km away)

gain 0.7

Ensemble ID ECC SNR RX gain Freq corr Date LTO FIC CRC Errors Tuned at FCT0 frame received at
0xa304 0xe0 16.0 0.0 135 2024-8-18 7:46 UTC 2 51 2024-08-18T07:46:15.616Z 2024-08-18T07:46:24.700Z

gain 9.7

Ensemble ID ECC SNR RX gain Freq corr Date LTO FIC CRC Errors Tuned at FCT0 frame received at
0xa304 0xe0 11.0 15.7 150 2024-8-18 7:51 UTC 2 233 2024-08-18T07:51:32.592Z 2024-08-18T07:51:36.700Z

gain 19.7 (overload, no lock)

Ensemble ID ECC SNR RX gain Freq corr Date LTO FIC CRC Errors Tuned at FCT0 frame received at
0x0000 0x00 7.0 36.4 -1 0-0-0 0:0 UTC 0 276 2024-08-18T07:47:53.299Z 2024-08-18T07:47:53.299Z

gain 29.7 (will be set to 0, see SNR)

Ensemble ID ECC SNR RX gain Freq corr Date LTO FIC CRC Errors Tuned at FCT0 frame received at
0xa304 0xe0 16.0 0.0 144 2024-8-18 7:49 UTC 2 61 2024-08-18T07:49:33.463Z 2024-08-18T07:49:48.699Z

gain 39.7 (will be set to 0)

RTL_SDR: Unknown gain count39

Ensemble ID ECC SNR RX gain Freq corr Date LTO FIC CRC Errors Tuned at FCT0 frame received at
0xa304 0xe0 16.0 0.0 151 2024-8-18 7:54 UTC 2 69 2024-08-18T07:53:56.875Z 2024-08-18T07:54:00.700Z

no manual gain

Ensemble ID ECC SNR RX gain Freq corr Date LTO FIC CRC Errors Tuned at FCT0 frame received at
0xa304 0xe0 21.0 1.4 148 2024-8-18 7:54 UTC 2 69 2024-08-18T07:54:43.626Z 2024-08-18T07:54:48.699Z

weak mux (67 km away, amplifier needed)

no manual gain

Ensemble ID ECC SNR RX gain Freq corr Date LTO FIC CRC Errors Tuned at FCT0 frame received at
0x9200 0xe4 17.0 25.4 181 2024-8-18 7:57 UTC 2 267 2024-08-18T07:55:54.842Z 2024-08-18T07:57:09.798Z

gain 0.7 (no lock, signal is too weak)

Ensemble ID ECC SNR RX gain Freq corr Date LTO FIC CRC Errors Tuned at FCT0 frame received at
0x0000 0x00 0.0 0.0 13018 0-0-0 0:0 UTC 0 120 2024-08-18T07:57:50.973Z 2024-08-18T07:57:50.973Z

gain 9.7 (lock)

Ensemble ID ECC SNR RX gain Freq corr Date LTO FIC CRC Errors Tuned at FCT0 frame received at
0x9200 0xe4 11.0 15.7 188 2024-8-18 7:58 UTC 2 94 2024-08-18T07:58:43.253Z 2024-08-18T07:58:45.798Z

gain 19.7 (overload, error)

RTL_SDR: ADC overload. Maybe you are using a too high gain. Info: ADC overload. Maybe you are using a too high gain.

Ensemble ID ECC SNR RX gain Freq corr Date LTO FIC CRC Errors Tuned at FCT0 frame received at
0x9200 0xe4 15.0 36.4 191 2024-8-18 7:59 UTC 2 92 2024-08-18T07:59:14.383Z 2024-08-18T07:59:14.383Z

gain 29.7, will be set to 0

see the following message

RTL_SDR: Unknown gain count29

Ensemble ID ECC SNR RX gain Freq corr Date LTO FIC CRC Errors Tuned at FCT0 frame received at
0x0000 0x00 0.0 0.0 0 0-0-0 0:0 UTC 0 0 2024-08-18T08:00:15.373Z 2024-08-18T08:00:15.373Z

workaround

you can try with rtl_tcp. In a first console you type

rtl_tcp

and in a second console 

./welle-cli -c 8D -Dw 7979 -F rtl_tcp -g 29.7

RTL_TCP_CLIENT: Tuner gain count: 29

BTW: the total amount of values is 29, but 29.7 is not the 29th value of course, see

https://github.com/AlbrechtL/welle.io/blob/5481d57e728e03441c83b5586c2017f858f1dc31/src/input/rtl_tcp.h#L128-L130

As I am not a programmer I cannot tell you if also this code makes sense:

https://github.com/AlbrechtL/welle.io/blob/5481d57e728e03441c83b5586c2017f858f1dc31/src/input/rtl_sdr.cpp#L188-L191

Hoerli1337 commented 2 months ago

Thanks for the quick reply. I have tested your commands, but they don't seem to work.

First I had to install rtl-sdr. That wasn't on the Pi.

pi@raspberrypi3bplus:~ $ rtl_test
Found 1 device(s):
  0:  Nooelec, NESDR SMArt v5, SN: 00000001

Using device 0: Generic RTL2832U OEM
Found Rafael Micro R820T tuner
Supported gain values (29): 0.0 0.9 1.4 2.7 3.7 7.7 8.7 12.5 14.4 15.7 16.6 19.7 20.7 22.9 25.4 28.0 29.7 32.8 33.8 36.4 37.2 38.6 40.2 42.1 43.4 43.9 44.5 48.0 49.6
[R82XX] PLL not locked!
Sampling at 2048000 S/s.

Info: This tool will continuously read from the device, and report if
samples get lost. If you observe no further output, everything is fine.

Reading samples in async mode...
Allocating 15 zero-copy buffers
lost at least 68 bytes

I was then able to start rtl_tcp:

pi@raspberrypi3bplus:~ $ rtl_tcp
Found 1 device(s):
  0:  Nooelec, NESDR SMArt v5, SN: 00000001

Using device 0: Generic RTL2832U OEM
Found Rafael Micro R820T tuner
[R82XX] PLL not locked!
Tuned to 100000000 Hz.
listening...
Use the device argument 'rtl_tcp=127.0.0.1:1234' in OsmoSDR (gr-osmosdr) source
to receive samples in GRC and control rtl_tcp parameters (frequency, gain, ...).
client accepted! localhost 32796
Allocating 15 zero-copy buffers
set gain mode 1
set gain 0
set sample rate 2048000
set freq 201072000

welle-cli with different gain levels did not work. According to the log, the gain remains at 29, regardless of what I enter via -g.

pi@raspberrypi3bplus:~ $ welle-cli -c 8D -w 7777 -F rtl_tcp -g 49.6
welle-cli 2.4+ds
InputFactory:Input device:rtl_tcp
OFDM-processor:restart
RTL_TCP_CLIENT: Try to connect to server 127.0.0.1:1234
Try to connect to: 127.0.0.1
RTL_TCP_CLIENT: Successful connected to server
RTL_TCP_CLIENT: Tuner type: 5 R820T
RTL_TCP_CLIENT: Tuner gain count: 29
ofdm-processor: SyncOnEndNull failed
ofdm-processor: Lost coarse sync (coarseCorrector: 0; fineCorrector: 0)
ofdm-processor: SyncOnPhase failed
Could not understand GET request /favicon.ico

I live really badly for DAB+ / certain companies don't want to expand. The signal has to get over two mountains. So I need a lot of input gain to get a usable signal or an extremely large antenna. I don't have the last of these at the moment and didn't necessarily want to buy one. I currently use a 60cm rod antenna, which was supplied with the stick.

Hoerli1337 commented 2 months ago

Okay, I'm stupid ... I misunderstood the instructions on the net for the gain parameter. Instead of the db values, the number must be specified. I always specified the supported values.

After testing with the numbers, I realized that I can only receive anything at all at levels 0-3. From level 4, nothing is received.

pi@raspberrypi3bplus:~ $ rtl_tcp
Found 1 device(s):
  0:  Nooelec, NESDR SMArt v5, SN: 00000001

Using device 0: Generic RTL2832U OEM
Found Rafael Micro R820T tuner
[R82XX] PLL not locked!
Tuned to 100000000 Hz.
listening...
Use the device argument 'rtl_tcp=127.0.0.1:1234' in OsmoSDR (gr-osmosdr) source
to receive samples in GRC and control rtl_tcp parameters (frequency, gain, ...).
client accepted! localhost 42526
Allocating 15 zero-copy buffers
set gain mode 1
set gain 0
set sample rate 2048000
set freq 201072000
set gain 14
set gain 27
set gain 37
set gain 77
set gain 87
set gain 125
set gain 144
set gain 157
set gain 166
set gain 197
set gain 207
set gain 229
set gain 254
set gain 280
set gain 297
set gain 328
set gain 338
set gain 328
set gain 297

What I still find a bit strange is the log itself. Regardless of whether I enter gain 0 or 3, these two lines can always be found in the log:

set gain mode 1
set gain 0