Hi!
I've just noticed one issue about E4000 driver that bothers me - IF bandwidth at the chip is set to 8 MHz, where maximum bandwidth of RTL2832U sampling is 3.2 MHz.
Wouldn't reducing IF bandwidth at tuner chip improve overall sensitivity (and maybe noise level)?
best regards
Michal Rogala
Hi Michal, Thank you for asking the question - I was thinking about this too last night (actually this morning) while I was trying to catch up on some sleep, but my brain wouldn't stop ticking over :)
Can't be sure of the effect as datasheets aren't public, but I will experiment with narrowing the bandwidth. The 'IFfilter' function in the E4000 tuner code is made up of a massive if/else block that sets various register values by checking the input IF-BW values from <= 2150 to > 5500, so it appears to support a wide range (despite the convenience 'E4000_BANDWIDTH_HZ' enum in the header).
Will post findings a little later, unless someone beats me to it :)
Kind regards, Balint
PS: "Software-Defined Radio For $11" on Slashdot! http://hardware.slashdot.org/story/12/03/31/1914217/software-defined-radio-f or-11
PPS: I've been working to improve the GNU Radio Source block and Windows plugin with generous feedback from the community. Grab them here if you wish: http://wiki.spench.net/wiki/gr-baz#rtl_source_c http://wiki.spench.net/wiki/USRP_Interfaces
-----Original Message----- From: Michał Rogala [mailto:michal.rogala@gmail.com] Sent: Sunday, 1 April 2012 4:36 AM To: balint256@gmail.com; osmocom-sdr@lists.osmocom.org Subject: RTL-SDR: E4000 IF bandwidth
Hi!
I've just noticed one issue about E4000 driver that bothers me - IF bandwidth at the chip is set to 8 MHz, where maximum bandwidth of RTL2832U sampling is 3.2 MHz.
Wouldn't reducing IF bandwidth at tuner chip improve overall sensitivity (and maybe noise level)?
best regards
Michal Rogala
Hi!
Thanks, would be great if it turns out that those changes improve reception :). What I've seen in source code FC0013 also has regulated IF bandwidth filter, although in very limited steps - 8 MHz, 7 MHz, 6 MHz.
best regards and thank you for great work :)
Michal Rogala
W dniu 1 kwietnia 2012 04:53 użytkownik Balint Seeber balint256@gmail.com napisał:
Hi Michal, Thank you for asking the question - I was thinking about this too last night (actually this morning) while I was trying to catch up on some sleep, but my brain wouldn't stop ticking over :)
Can't be sure of the effect as datasheets aren't public, but I will experiment with narrowing the bandwidth. The 'IFfilter' function in the E4000 tuner code is made up of a massive if/else block that sets various register values by checking the input IF-BW values from <= 2150 to > 5500, so it appears to support a wide range (despite the convenience 'E4000_BANDWIDTH_HZ' enum in the header).
Will post findings a little later, unless someone beats me to it :)
Kind regards, Balint
PS: "Software-Defined Radio For $11" on Slashdot! http://hardware.slashdot.org/story/12/03/31/1914217/software-defined-radio-f or-11
PPS: I've been working to improve the GNU Radio Source block and Windows plugin with generous feedback from the community. Grab them here if you wish: http://wiki.spench.net/wiki/gr-baz#rtl_source_c http://wiki.spench.net/wiki/USRP_Interfaces
-----Original Message----- From: Michał Rogala [mailto:michal.rogala@gmail.com] Sent: Sunday, 1 April 2012 4:36 AM To: balint256@gmail.com; osmocom-sdr@lists.osmocom.org Subject: RTL-SDR: E4000 IF bandwidth
Hi!
I've just noticed one issue about E4000 driver that bothers me - IF bandwidth at the chip is set to 8 MHz, where maximum bandwidth of RTL2832U sampling is 3.2 MHz.
Wouldn't reducing IF bandwidth at tuner chip improve overall sensitivity (and maybe noise level)?
best regards
Michal Rogala
-
Balint Seeber -
Michał Rogala