RTL-SDR dongle seems less sensitive.

Main forum to discuss RTL-SDR related topics.
Post Reply
gerryk
Posts: 3
Joined: Fri Feb 12, 2016 11:05 am

RTL-SDR dongle seems less sensitive.

Post by gerryk » Fri Feb 12, 2016 11:11 am

I recently bough the rtl-sdr R820T2 based dongle. It seems less sensitive though.
On a Nooelec unit I have (R820T) I can use kal to pick up a nearby GSM base station and use it to calibrate.
Using the rtl-sdr unit though, it doesn't pick up the celltower, despite using the same antenna and feedline.

Code: Select all

[email protected]:~/Development/multimon-ng/build $ kal -s GSM900
Found 1 device(s):
  0:  Generic RTL2832U OEM

Using device 0: Generic RTL2832U OEM
Found Rafael Micro R820T tuner
Exact sample rate is: 270833.002142 Hz
kal: Scanning for GSM-900 base stations.
GSM-900:
I also get a lot of dropped bytes on any sample rates above 2.4Ms/s

Code: Select all

[email protected]:~/Development/multimon-ng/build $ rtl_test -s 3.2e6
Found 1 device(s):
  0:  Generic RTL2832U OEM

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 

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...
lost at least 12 bytes
lost at least 188 bytes
lost at least 68 bytes
lost at least 68 bytes
lost at least 188 bytes
lost at least 188 bytes
.
.
.

Code: Select all

[email protected]:~/Development/multimon-ng/build $ rtl_test -s 2.8e6
Found 1 device(s):
  0:  Generic RTL2832U OEM

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 
Exact sample rate is: 2800000.037087 Hz

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...
lost at least 148 bytes
lost at least 68 bytes
lost at least 68 bytes
.
.
.

Code: Select all

[email protected]:~/Development/multimon-ng/build $ rtl_test -s 2.4e6
Found 1 device(s):
  0:  Generic RTL2832U OEM

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 

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...
lost at least 188 bytes
^T^CSignal caught, exiting!

User cancel, exiting...

kentro
Posts: 3
Joined: Wed Mar 02, 2016 6:30 am

Re: RTL-SDR dongle seems less sensitive.

Post by kentro » Wed Mar 02, 2016 6:42 am

I've just had the same experience. My new R820T2 has 30km less range than the old R820T with dump1090 measured over the last week.

I'm not sure if this is a bad batch for the R820T2 or if all of the T vs T2 reviews were completely qualitative and produced an incorrect conclusion. The TXCO seems much more stable but I did notice the gsm bands were different for my kal test.

Have you done any more tests on your setup?

rtlsdrblog
Site Admin
Posts: 2743
Joined: Mon Nov 19, 2012 11:54 pm

Re: RTL-SDR dongle seems less sensitive.

Post by rtlsdrblog » Thu Mar 03, 2016 12:11 pm

The R820T2 has better sensitivity in single tone SNR tests.

There could be a possibility that the different filters in the R820T2 could cause interference to show up at different places than on the R820T, maybe you are experiencing this?

kentro
Posts: 3
Joined: Wed Mar 02, 2016 6:30 am

Re: RTL-SDR dongle seems less sensitive.

Post by kentro » Sun Mar 06, 2016 5:15 pm

I've done some pretty extensive simultaneous tests at 1090MHz from the same machine with similar antennas sharing the same ground plane. This is essentially a single tone SNR test and the R820T always has a better noise floor and SNR.

1090MHz results:

Code: Select all

     r820t           r820t2	
-27.4897381431   -26.7697898363   mean noise dB
-24.0856451845   -23.9171988461   mean signal dB
  3.4040929586     2.8525909902   SNR dB
In case someone wants to try to reproduce this here's the rtl_power code. This runs for 5 hours from 1060MHz to 1120MHz, you'll also need to calibrate the ppm for the R820T. Run both SDRs from the same machine and use similar antenna heights next to each other on a metal plate.

You have to run the benchmark on both of the SDRs at the same time and location because ADS-B traffic can vary greatly from minute to minute.

Code: Select all

rtl_power -f 1.060M:1.120M:1K -d 1 -p 163 -i 5m -e 300m -c .2 r820t-4.csv
rtl_power -f 1.060M:1.120M:1K -d 0 -p 0 -i 5m -e 300m -c .2 r820t2-4.csv
After that you can use flatten.py to make the results a little easier to digest. The signal will be measured on rows 53 and 54 and the noise is measured everywhere else.

rtlsdrblog
Site Admin
Posts: 2743
Joined: Mon Nov 19, 2012 11:54 pm

Re: RTL-SDR dongle seems less sensitive.

Post by rtlsdrblog » Mon Mar 07, 2016 5:45 am

I'll recheck later this week too.

But RTL-SDR's have a tolerance variation in sensitivity between different units by maybe 1-2 dB's, even with the same tuner. It's likely that the R820T2 and R820T are more or less identical up at 1090 MHz, and you're just measuring the variations between different units.

kentro
Posts: 3
Joined: Wed Mar 02, 2016 6:30 am

Re: RTL-SDR dongle seems less sensitive.

Post by kentro » Mon Mar 21, 2016 1:50 am

I've done a few more tests with a couple of old R820T and also added a Nano R820T2/.5ppm txco to the mix.

Short story, the nano T2 still seems to have less range than the standard R820T with dump1090 and the 8 element co-coax. The noise tests are somewhat nebulous since the nanoT2 vs R820T looks like it has a much better SNR. I need to re-test both of the T2s side by side and probably 3 dongles at once. The noise floor seemed much lower than my previous test for all of the dongles.

Old dongles, small difference possibly because the B dongle was crossing the laptop power supply.

Code: Select all

r820ta          r820tb	
-32.518479299	-32.2348957446	mean noise
-25.8787516549	-25.8325028244	mean signal
6.6397276441	6.4023929202	SNR

Code: Select all

r820t          r820t2 nano	
-32.0066932056	-36.0740015451	mean noise
-25.7541080785	-26.4982864223	mean signal
6.2525851271	9.5757151228	SNR

rtlsdrblog
Site Admin
Posts: 2743
Joined: Mon Nov 19, 2012 11:54 pm

Re: RTL-SDR dongle seems less sensitive.

Post by rtlsdrblog » Wed Mar 30, 2016 10:55 pm

I did some tests with 10x R820T's and 10x R820T2 dongles. At max gain on 1090 MHz the R820T is on average about 0.5dB more sensitive which seems to fit well with your results. But at lower gain values the R820T2 is on average 2-5dB more sensitive.

Both tuners are the same inside, but according to the factory they are just made at different locations, with the T2 being made at the factory with higher quality standards and having a very slightly different IF filter. So it seems that the manufacturing variability for the R820T pushes the average sensitivity up a bit at max gain, but at lower gain levels the better manufacturing on the R820T2 improves things much more.

Post Reply