Do you have noise floor abt S9+30 when turning gain to full, GR=0?
I tested more again and found 24Mhz and other spurs are on screen even the adjacent input pins are shorted. Level measured at ant. connector then decreases, as it should. So, these must come from chip itself? not direct from xtal etc. Maybe some components are broken inside?
I measured some dynamic range. (All DR values about)
Gr=59, 6Mhz, IF bw 1.536Mhz, LO 6.2Mhz (so dc not affect). IMD 75dB, Blocking 80dB.
Gr=10 IMD abt. 45dB, Blocking 50dB.
Single signal 1Khz from dc-spike, "IMD" 20dB.
Gr=59, 6Mhz, IF bw 1.536Mhz. IMD 65dB, Blocking 80dB. Image rejection (LO 2.048 above) 6Mhz -42db, 125Mhz -45dB.
Comparison to rtl v3, G=30, 125Mhz, Image (LO 3.57 above) -53dB, IMD 70dB (6Mhz+upconv.), blocking 70-78dB at all gain settings.
Talk about other SDR products like the FunCube, HackRF, BladeRF etc.