Guidance on mobile SDR (large acreage site)

[john21443]

Environment

I am running a mobile SDR rig over 600 hilly acres (spectator at an auto race track).  I only receive with focus on 650-670Mhz narrow band fm.  The transmitters being listened to are both fixed (facility/racing teams) and mobile (drivers in race cars).

I am utilizing a gnuradio based application to demodulate multiple transmissions simultaneously (ham2mon).

Problem

With an RTL dongle I was unable to both receive weak signals (car somewhere on race track) and strong signals (fixed transmitter very close).  What happens:  If I turn down gain I missed the weaker signals.  If I turned it up the gain then the signals overload receiver and spurious signals are generated.  These spurious signals  are offset from intended frequency and picked up as “duplicate” transmissions.

The rig is on a golf cart and moves around on a whim so need something that handles wide range dynamic without manual tweaking.

I do not have easy access to environment so cannot readily test out options to improve reception.  Therefore I am looking for things I can do to prepare the set up for when I use it next.    I am no looking for perfection.

For the mobile rig I am using an antenna optimized (according to the manufacturer) for 450-470Mhz.

What has been done in preparation for next time I am at track:

• Switched from RTL dongle to Airspy Mini.  Compared to the RTL dongle this has  lowered the noise floor and improved linearity.
• Attempted to simulate dynamic range:
• Connected to TV antenna mounted on my home’s roof
• Alternately keyed up FRS radio and changed center frequency to pick up analog police 15+ miles away
• Adjusted gains so both scenarios are received clearly at same levels.    Seemed like with further tuning I can leave gains consistent between two cases and get reasonable reception for cases

Possible options I would like feedback on:

• Should I consider blocking fm radio frequencies with a notch filter or use bandpass for 450-470Mhz?  The idea would be to further reduce the noise floor.  I don’t see a downside to this.
• Add an attenuater to reduce signal strength.  This would reduce the potential to overload front end and support higher gains.  Wouldn't it also reduce weaker signals in a way that would offset the benefits?
• Anything else?

Yes, I can play around with things when at the venue but I am also drinking a lot of beer at that time so my debugging judgement will be questionable.    As it stands it is suspect.

Thanks,

John

[DaJMasta]

Are the strong signals overloading your frontend in your band of interest?  Basically, are the fixed transmitters the team's signals going out to the drivers?  Cause if they are, a bandpass filter isn't going to do much.  If you have other frequencies outside of that 650-670MHz band that are too strong and are being picked up and overloading your frontend, a bandpass filter for the band of interest will reduce overloading.  In either case, if it's properly tuned it can't really hurt, and the bandpass filter over the region you want will lower the noise floor of the system as a whole.

It does sound more like a problem of dynamic range, though, and without active notching of nearby signals or a tight bandpass filter around the frequency you're currently tuned to, your best bet is probably an SDR with more effective bits on its ADC.  If you can get another few bits (with a higher resolution converter, software decimation, or whatever), then you can put an attenuator in (or just reduce programmable gain) and still be able to discern the weak signals.


A potential third option could be a directional antenna.  If you know that the home stations for the racing teams are in a certain direction and are the source of the strongest signals, you could have an antenna with some gain and then just keep it pointed away from those higher power transmitters.

[john21443]

Thanks for your informative reply.

In my case the strong signals that are overloading are in the band of interest.   And yes, they typically are the fixed stations.   So as you indicate a bandpass filter won't directly help.  However, I will take a look at this for reducing the noise floor in general.

The Airspy SDR has a higher bit rate than the RTL I was using.  Maybe this is part of why it seems to work better in my "simulation".   For the first time out I will work with the gains and go from there.

The directional antenna is something I haven't thought of before.   It might not be practical in my situation but I will think about it some more.

[DaJMasta]

I think for your setting, a directional antenna would help sometimes and not others - I assume the home bases for the cars are all in the same area, so most of the time you'd be able to keep them out of the antenna's focus, but there would be times when you couldn't avoid pointing at them, and you'd probably have to fish around for the car you were looking for sometimes - could be a lot more work.

A bandpass filter will keep high level stations on other bands the receiver can pickup out, but it doesn't sound like they're contributing significantly to your overload.  Maybe next time you're out, take a broader look at the spectrum before the race and see if there's anything else strong getting close - if there are, then maybe the bandpass filter would help.

You could try bandstop/notch filters to remove specific strong stations, but if they change frequency or if there are several, it could be a real pain to try and chase down.  At least theoretically, you could have an adjustable notch filter that you have inline with your antenna where you look at the spectrogram of the bandwidth coming in, and then tune up or down to pick out the strongest signal, you could largely remove it from being a problem.... but if there's more than one, you start needing more notch filters and the whole thing gets complicated fast.

You still have the attenuator-on-the-input option if you can differentiate the weak signals in the noise.  It sounds like the airspy mini may do the trick for that, getting a couple extra bits worth of usable data beyond the more standard 8 bit converters, so if you haven't already tried it, bring a modest attenuator to check inline with that - maybe the new hardware will give you that low signal performance.  If you have adjustable input stage gain, you can also just turn it down, which spares you from needing a physical attenuator.