Hello Francesco,
As far as I can see there is no difference when I set 0x05 or 0xd5. The bits are no longer unknown - the posting from Steve Markgraf tells us that they mean nothing for a e4000 in "serial interface gain control mode" so the appropriate value should be 0x05.
When I first implemented code for rtlsdr in Linrad I did not know how to disable the AGC. As a resuilt the system performance was at the "toy" level. (Not too bad for USD 20)
Steve Markgraf wrote 2012-06-22:
The remaining AGC that's active is not in the E4K, but it's the Digital AGC (DAGC) of the RTL2832. Unfortunately we don't know how to disable it, since the way it's supposed to be disabled does not work.
That is somehow in contrast to Steves latest posting....
Anyway, today the performance is well above the "toy level" now that we know how to disable the AGC.
Regards
Leif
Hi Leif! Nice work, and I'm glad to hear my post inspired you, even if my code didn't work at all! I'm now setting 0x05 (0000 0101) instead of the original 0x25 (0010 0101) or your 0xd5 (1101 0101), to avoid setting unknown bits, and it seems to have the same effect (I see an about 10 dB noise floor drop when on minimum gain with no signal).
Could you test the new value and check if you get the expected results?
Thank you!
-- Francesco Gugliuzza HackLabProject.org Administrator Linux user #374630 Tel (VoIP geographic number): +39 0921440446 Tel (Libera il VoIP number): 5125320 E-mail: f.gugliuzza@hacklabproject.org