Hardening the V4 for 915MHz:Solving Urban De-Sensing with a >50dB Saw Shield

General
1

Hi Everyone,I’ve been testing the RTL-SDR V4 for dedicated 915MHz monitoring (Meshtastic/LoRaWAN) in a high-density urban area.Whilst the triplexer is a huge step forward, I noticed that the R828D tuner was still being ‘De-Sensed’ by local 800MHz cellular towers,forcing the LNA to go into gain compression (lower LNA gain) so producing a loss of sensitivity for LoRa packets.

To help overcome this problem,I’ve developed a very tiny connectorized (u.fl) SAW filter designed to sit inline with the V4’s internal pigtail.It provides a >50dB rejection cliff at 800MHz.Because its internal,it keeps the setup compact without the bulk of external SMA dongles.

See the plot below of filter performance measured on a VNA:

PassBand_915_Span(200MHz)_2
PassBand_915_Span(200MHz)_21755×950 220 KB

As you can see from the VNA plot,the rejection is very high at 821MHz which is the highest Cell Tower Transmitter frequency for the n20 band (commonly referred to as 800MHz_5G.It drops more than 50dB just outside of our wanted passband,effectively silencing the Cellular Noise before it hits the tuner’s LNA.

I’m curious if anyone else on the forum has experimented with internal SAW Filter mods for the V4.I’m currently testing the thermal impact of the board sitting near the R828D-so far,it looks stable.Happy to share my findings if anyone else is fighting similar noise/blocking issues.

2

In addition,the V4 can suffer a loss of Sensitivity due to high power Cell Towers Transmitting in the band (925-960)MHz.They can transmit at up to 80W rms.Eighty watts is 100 billion times more powerful than a weak LoRa signal (1 nanoWatt).However,with this filter modification, the rejection at 960MHz is typically <-45dB and this could make the difference between receiving few or even no packets and receiving 100% of all packets.