Airpsy R2 Gain Distribution and Filtering


Will
 

I have 2 Airspy R2, both running RC10-6-g4008185. Frequency calibration looks fine. One is from around 2018, and the other I believe is from 2020.

The 2020 unit receives a weak signal at about 406 MHz with gain at 19, signal is -65 and noise floor is -75. The 2018 device receives the same signal at a gain of 14, signal is -75 and noise floor is -85. If I move the gain up on the older device, it quickly overloads and the noise floor is very high. I'm testing with no decimation.

If I put in an FM bandstop filter for 88-108, both devices perform essentially the same, with signal at -80 and noise floor of -90 with gain 14. I can increase gain to 19 with no overload on either device. Is there a difference in overload response between older and newer devices, or is this indicating a problem with one of them?

Thanks.


prog
 

On Fri, Apr 23, 2021 at 10:21 AM, Will wrote:
I have 2 Airspy R2, both running RC10-6-g4008185. Frequency calibration looks fine. One is from around 2018, and the other I believe is from 2020.

The 2020 unit receives a weak signal at about 406 MHz with gain at 19, signal is -65 and noise floor is -75. The 2018 device receives the same signal at a gain of 14, signal is -75 and noise floor is -85. If I move the gain up on the older device, it quickly overloads and the noise floor is very high. I'm testing with no decimation.

If I put in an FM bandstop filter for 88-108, both devices perform essentially the same, with signal at -80 and noise floor of -90 with gain 14. I can increase gain to 19 with no overload on either device. Is there a difference in overload response between older and newer devices, or is this indicating a problem with one of them?
The very reason there is a gain setting is to allow you to adjust the NF of the receiver according to your signal environment. The variability of the silicon may give different optimal settings with different chips. In either case, you still need to find that optimal operating point.


Will
 

Excellent - thank you.