Gain staging, overload and noise: getting the most out of RTL-SDR, SDRplay, Airspy and HackRF
- Replies 2
- Views 14
- Created
- Last Reply
Top Posters In This Topic
-
Admin_Terry 1 post
-
Mod_Chad 1 post
A better way to browse. Learn more.
A full-screen app on your home screen with push notifications, badges and more.
A lot of SDR troubleshooting threads come down to the same cause: the front end is being overdriven, or the receive system is limited by local noise rather than by the dongle. A common pattern is a waterfall full of phantom signals that move around when you touch the gain slider, with weak signals that never seem to improve however much gain you add. It is worth sorting out which problem you actually have, because the fixes are different.
Start with dynamic range, since it explains why different hardware behaves so differently. An ADC gives roughly 6 dB of theoretical range per bit, so an 8-bit device such as an RTL-SDR or HackRF has far less headroom than a 12-bit Airspy or LimeSDR, or a 14-bit SDRplay RSP. Real figures are lower than the theory because of the tuner, mixer and filtering ahead of the ADC. The practical result is that one strong broadcast FM or pager signal inside the wide capture bandwidth can eat the whole range of a cheap dongle and generate images and intermodulation products across the band. Narrowing the sample rate and decimating can help a little, but it does not remove strong out-of-band signals that have already reached the mixer.
The software side matters for diagnosis. SDR# and SDRuno give you good waterfall and spectrum controls. HDSDR relies on ExtIO drivers for many devices. GQRX and CubicSDR are popular on Linux and macOS and typically sit on top of SoapySDR or gr-osmosdr. The tool you choose changes how convenient the gain controls are, but not the physics. Whatever program you use, check whether the RF gain, LNA state and IF or AGC controls are separate, because it is easy to run too much gain at one stage and too little at another.
My question for the thread is how people approach this in a repeatable way. What is your method for deciding when gain is right, and when do you add filtering (FM or broadcast band-stop, bandpass for the band of interest) instead of just turning things down? Also, how do upconverters, bias-tee LNAs and long coax runs fit in? An upconverter lets a cheap dongle cover HF, but it adds its own mixer, local oscillator and potential spurs, and direct sampling modes on some dongles are a different compromise again. For decoding digital modes like FT8, ADS-B, DMR or pager traffic, I would also like to hear which audio or IQ routing methods hold up best.
Link to comment
https://www.hamradiobase.com/forums/topic/11830-gain-staging-overload-and-noise-getting-the-most-out-of-rtl-sdr-sdrplay-airspy-and-hackrf/Share on other sites