Help newbie get on the right path

the closer the LNA is to the antenna the lower the induced noise at the receiver. i do exactly what Lawrence Hill suggests and it works very well. the Pi is in the house., the LNA is in a small weather proof box on the roof with an 18 inch (approx.) cable from the FA antenna to the filtered LNA, i use the RTL-SDR Blog 1090 LNA. it is filtered so you get the advantage of the LNA close to the antenna AND filtering at the antenna as well. then you can use any wideband receiver (i use Airspy Mini…great dynamic range).
i don’t know if RTL-SDR make a filtered LNA for 978 but there is so little traffic compared to 1090 that i haven’t been able to justify the 978 setup. you may have more UAT traffic in your area???

Yes, there is a big advantage in using an LNA at the antenna. The problem with an LNA in the dongle is that it will amplify any noise in the system in front of it. Look up “cascaded noise figure”.
The noise figure of a system shows how well it performs in terms of sensitivity. The lower the noise figure the weaker the signal that can be received. The lower the noise figure, the more sensitive the system, so the lower the better. The problem is, any losses before the first stage directly add to the noise figure of the overall system. A typical dongle will have a noise figure of about 3dB. To show the effect of an LNA at the antenna let’s see what happens if you have 10dB loss in your co-ax. I know this is pretty extreme but it illustrates the point. Without an LNA, the system noise figure will be 13dB. 10dB cable loss plus the 3dB dongle noise figure. Now, if we put a 20dB gain LNA with a typical 2dB noise figure at the antenna then the system noise figure reduces to 2.25dB, over 10dB better.
If you want to play with a noise figure calculator try:
https://www.everythingrf.com/rf-calculators/cascaded-noise-figure-gain-calculator

1 Like

Thank you all for helping me learn… Which dongle do you use with your LNA setup?

I use the RTL-SDR Blog V3 dongle. It has a bias-t to power the RTL-SDR LNA. It’s still $32.95 at Amazon.

I don’t seem to be getting much UAT traffic. I live less than 10 miles away from a small airport that recently expanded its runway. When I’ve been outside, I’ve seen very expensive-looking private jets flying in and out of that airport. Perhaps those jets are using 1090 Mhz or my gain is set too high to pick up planes flying low and close to me that are using 978 Mhz. I have to do more research to figure that out.

I have 2 sites so that I can make comparisons between them if I make changes.
One uses a FlightAware blue dongle, the ProStick Plus. The other uses an Airspy mini.

1 Like

UAT is only for aircraft that operate below 18,000 ft (i’m old, that altitude from memory). so any aircraft that operates above 18,000 ft must use 1090. in my area even small single engine trainers tend to use 1090…its very rare that i hear an aircraft overhead and don’t see on my screen using 1090 only. tweeking gain is part of the fun of this hobby

1 Like

If I can get another Pi, I might make two sites. I ordered the blue dongle a couple of days ago. I’m currently using the orange one. Just to be clear, having an external LNA at the antenna paired with an LNA built in the blue dongle doesn’t cause problems?

That’s great that you have two sites and can compare their performance. Which hardware and settings work best for you?

The potential problem with having an external LNA as well as the LNA in the blue dongle is that of overloading on nearby signals. My setup has a 27dB gain 1dB noise figure LNA at the antenna. The blue dongle is fed from the coupled port of a 20dB directional coupler so it effectively only has 7dB extra gain. The through line of the coupler goes to the AirSpy mini.
In terms of performance the AirSpy has the edge but whether it is enough to justify the extra cost is up to individual users to decide for themselves.
You can compare them here if you like.
This one (site 72716) is the ProStick:

This one (site 97430) is the AirSpy mini:

So you amplify the signal at the antenna with an external LNA then you attenuate the signal with something like this right before the dongle?

Something like that would do fine. I’m actually using something like this:
https://www.mini-circuits.com/WebStore/dashboard.html?model=ZUDC20-5R23-S%2B
Just because I had it lying around.

At this point, I’m most interested in the RTL-SDR Blog V3 dongle and the RTL-SDR LNA. I have a few more questions:

  • What cable and connectors will I need? Are the connections: ADSBexchange antenna > short N to SMA coax > RTL-SDR LNA > Longer SMA to N coax > some kind of N to SMA adapter > RTL-SDR Blog V3 dongle > Pi?

  • Does anyone know if the ADSBexchange antenna is a DC short circuited antenna? The RTL-SDR website give the following warning about using their bias tee:

Feature 2: Software Selectable Bias Tee

V.1. and V.2. of our dongles included a bias tee which could manually be enabled by opening the case and soldering two pads on the PCB together. V.3. introduces a 4.5V bias tee that can be toggled entirely in software. The bias tee can continuously pull up to 180 mA of current.

WARNING: Before using the bias tee please ensure that you understand that you should not use this option when the dongle is connected directly to a DC short circuited antenna unless you are using an LNA. Although the bias tee circuit is dual protected against accidental shorts with a thermal self-resetting fuse and overcurrent protection on the LDO, short circuiting the bias tee for an extended period of time (days) could damage the LDO or fuse permanently. Only use it while connected to an actual powered device, like an LNA, active antenna or the SpyVerter.

To make things clearer: DC Short Antenna → LNA → Coax → V3(bias tee on) is absolutely fine. What’s not good and makes no sense anyway is DC Short Antenna → Coax → V3(bias tee on). DC Short Antenna → Coax → V3(bias tee off) is fine.

Note that the legacy DVB-T TV drivers will activate the bias tee by default. On Linux ensure that you have properly blacklisted the DVB-T drivers. More info on how to blacklist on the Linux section on the quickstart guide.

  • The cleanest setup to me seems to be one that uses the USB powered bias tee of the dongle to power the LNA. It seems some people use an external bias tee injectors. Why not just use the bias tee in the dongle?

That’s what I have and use the bias tee to power the LNA. You don’t have to worry about the antenna being a short circuit, since the bias tee provides the voltage at the output side of the LNA.

By the way you don’t need lengths of coax to connect antenna → LNA → dongle, just use adapters. That is, N to SMA from antenna to LNA and SMA to SMA from LNA to dongle.

One issue about using the bias tee in the dongle, is how do you turn it on after your system has started up. After dump1090-fa starts up, it has control of the dongle, so you can’t turn on the bias tee. There are various solutions to this that are documented in other threads. I just have a simple script that I run that stops dump1090-fa, turns on the bias tee, and then starts dump1090-fa again. There are other ways of doing it that are more eloquent, but this works.

I just counted 14 potentential sources of loss.

What setup do you suggest?

I use raspi b3+ and some open source construct. I haven’t changed my sd card for a long time. She still seems to work pretty well for my Etihad stamp-collecting hobby.

… blue dongle and a very short run of coax.

Every other device (splitters, preamps, amps) you introduce, the more potential loss, if that makes sense.

Maybe @bigclivedotcom or some ge engineers will correct me. I welcome it.

Sorry, that is wrong. Adding an LNA at the antenna does not introduce loss. As long as it is done correctly it will improve the performance of the system.
Just as a comparison this is my equivalent graph with an LNA at the antenna.
I know it’s probably not fair to try and compare my site with yours as we are in very different locations but mine works much better with an LNA at the antenna.

Nice signals.

I didn’t mean guaranteed a loss. I just meant a potential loss.

Sorry, I’m not an electrical engineer. I guess that’s why I was looking for help.

Tnx!