If you place a good band-pass filter like Flightaware Light Blue or Dark Blue Filter or a Cavity Filter beween Antenna and Receiver (or Antenna & LNA if you are using a LNA without builtin filter), then all the FM and VHF UHF out of band frequencies will be eliminated by filter and wont reach the Receiver or LNA. This will prevent:
(1) Overloading & de-sensing of receiver or of LNA
(2) Creation of harmonics by mixing & cross-modulation of various out of band frequencies in the amplifier stages of receiver or LNA.
Please see “Scan 3 of 3” below which shows how adding a Filter Between Antenna and Receiver eliminates unwanted RF signals picked by Antenna.
The scan 1 shows RF noise picked by antenna and processed without any filter , The scans 2 & 3 were done with filters, and show how filters remove this noise.
Scan 1 of 3 - FA Antenna + Generic DVB-T (no internal or external filter)
Thanks for the replies regarding the filter, at the moment the plan is: 1090MHz antenna (reference the picture at the start of this thread), then as short as possible coax (3D-FB) run into a waterproof box with the 1090MHz filter before the LNA which would be both in the box directly bellow the antenna, then as short as possible LMR-400 cable run indoors - which would go through a bias tee then into the FlightAware Orange SDR. Any further advice would be appreciated if anything seems in the wrong position or anything else what would improve it further. Thanks.
There is no need of that cable to be low loss. After the LNA you can have attenuation, the resulting chain SNR will not be affected significantly.
That “better” cable expense will not yield any real benefits.
Another questionable decision. People keep being scared of “interferences” in front of a LNA. I think that’s not needed in most of the cases, and placing a filter in front of an LNA will just decrease the SNR of the chain. You can try to see what results you get.
Thanks for the advice, guess something like RG58 would do the trick. Would I still see a relative improvement in messages received and the amount of aircrafts seen even using a low quality coax after the LNA - guess that would be roughly 2/3 dB amount of loss.
Not likely at 1GHz. I used RG6. It’s an impedance miss-match, but it won’t affect too much the end result.
Note that you might need adapters at the end of the cable: F-type to SMA.
I haven’t ordered any coax yet, so should be able to get some RG6. I’ve got all the adapters needed for all the different male / female connections. Not sure if you have seen the scan I did of RF noise posted earlier, but here it is attached regarding the amount of noise being received by the antenna. I’m planning on getting a RTL-SDR 1090MHz LNA (what has no filter, unlike the uputronics LNA.)
That scan looks “normal”, except the end portion. Also I think the antenna you used is not appropriate for higher frequencies. Your receiver stops working at about 1.5GHz, what you see after that is a mirror image created in software.
The LTE signals look impressive on the scans but remember, all the cellphones work perfectly fine tuning on those close frequencies, without interfering one with another. The LNA chips inside the ADS-B LNA filtered amps are similar in their technical capabilities and will not be overloaded. Unless you are really at less than 30’ of a cell tower. A good ADS-B antenna is tuned for 1090MHz, so it will favor that signal.
Thanks for the advice, regarding the antenna - its a AirNav ADS-B 1090 MHz (not sure if its specifically tuned at 1090MHz - anyone’s knowledge on this antenna would be useful.) I do live in a rural environment, but with very strong LTE coverage what doesn’t help. I do have a LTE tower (I think) at 533m what might be why I’m receiving the strong signals on the RF graph.
I think you are far enough from that tower so you don’t need any filters ahead of the LNA.
I am at about 300 m of several towers plus 1000 meters from an airport (SSR emitter at 1030MHz) and I don’t use a pre-filter.
Raise the antenna as much as you can above surrounding obstacles (houses).
Thanks for the advice, don’t know if its a LTE transmitter or something else, don’t have any knowledge in that area. I’m planning on raising it more before I’ve finished with the upgrade, only issue is the location isn’t ideal to the northeast - I have a line of large trees what make it impossible to get above them. Although I have line of sight out towards the North Atlantic what helps with range, I’ve attached a reference photo if you want to have a look.
My trees are surrounding me. After fall I see a bit of improvement in long range (200-250 miles) position numbers, from <50 to about 555 a day. All in all not a big deal, I have 800k positions a day.
I’ll attempt at raising it further and see if it improves my range by a couple of miles, but I think the main improvement will be the shortened coax length from the fixed connection on the antenna - also with the LNA which should boost the number of packets received with positions.
Hello Keith, thanks for your help. At the moment I’m deciding if I want to reuse the coax what I’ll be cutting from the antenna what’s connected as standard. This would mean Antenna > very short run of 3D-FB into a weatherproof box > RTL-SDR LNA > then around 14 feet of the 3D-FB (1.48db loss at 1GHz) > bias tee > SDR > Pi 4 2GB. Other than the coax, I think the plan is nearly sorted. I was going to use a pigtail from the end of the coax to the SDR - but I’ve realised its RG316 with 0.88 attenuation at 1GHz! So I’ll just use a SMA Male - SMA Male coupler. As the LNA specs say it is expected to be used with roughly 3db loss from the LNA to the SDR, it seems that using 3D-FB will do the job - which makes it pointless paying £40+ for good coax just for playing around with probably small numbers of loss what wouldn’t make a noticeable difference. Although your opinion is very welcome, this upgrade has been helped a lot from others already!
Well, I have to say that I’m impressed even with the first data coming through about my upgrade what I did today.
I’ve changed from:
10m / 33ft of 3D-FB (roughly 3.7dB loss over 10m at 1GHz), which has been shortened into a very small run into a waterproof box which houses a RTL-SDR LNA powered via a bias tee (external). This then heads inside, with roughly 18ft of 3D-FB taking it into the bias tee then the FA Pro Stick.
Another upgrade has been the Pi 4 2GB, this has made a huge difference. With the processor around 7.5% usage at peak. This has directly impacted graphs1090, with a higher message rate (peaked at 950 earlier) with range maxing out at 230 (cross-checked with heywhatsthat range outline added onto tar1090 and it matches my predicated max range.)
Although, I’m planning in the future a couple more upgrades before its complete. This includes getting a new SDR (airspy mini) and higher quality coax for the run from the LNA down into the SDR.) I was planning on getting better coax now, but as I have a FA Pro Stick with the integrated amp - I wasn’t that bothered on loss as I didn’t want any front end issues on the SDR as it would have double amplification.
I’ll post a few photos of my graphs in a couple of days, as they are probably not accurate enough after 8 hours.
I made a change from 3B to 4 a year ago without significant improvement. That was expected as the USB port is not impacted that much by an RTL-SDR, only an Airspy is keeping the newer USB slots more busy
Did you start feeding new on Flightaware? Your site is up and running only since yesterday
The URL to adsbexchange is not working. Map loads but stating “problem fetching data…”
I’ve changed from a Pi Zero, so that’s probably why the difference is noticeable. I had issues recently with FA on the Pi Zero, so since I’ve got the Pi 4 2GB - I’m now sharing my feed with FA and a variety of different companies. I also had issues with sorting out all the feeds yesterday, so that’s why the adsbx link is broken: but here is the new one: ADS-B Exchange - 1/43 (adsbexchange.com)
After nearly two weeks, I’ve gathered enough data to give a reasonable performance level of my upgrades.
From the data collected with graphs1090, the LNA has directly impacted the number of messages received (with positions) and the number of aircrafts tracked. Although this hasn’t directly impacted my max range as I had already reached the maximum range without any upgrades, but the LNA has certainly helped with the signal level (lowest received is now around -32.2 dbfs with the previous lowest being -26.2 dbfs.)
Another positive is the number of aircrafts being tracked over a 24-hour period, this has changed from around 1,000 at peak in the summer to a lowest of 525 (Christmas day). But on average its 1,000+ at the moment in Winter. This has impacted the positions reported, with a peak of 179,539 seen on the 27th. But this heavily depends on the North Atlantic tracks what change daily, as on some days all the traffic heading for AMS and LHR (only naming a few) could be directly over me - what really boosts my aircraft tracked and message count. An extra benefit is the traffic towards Spain and Portugal, which is at peak in the summer - which will probably give me 1,500+ aircrafts in 24h and 200,000+ positions which will be better than the winter traffic which is low.
I thought it would be worthwhile adding a 24h period of tracks from tar1090 to give a reasonable idea of which directions my traffic comes from which changes daily. In simple terms:
From Ireland heading for Europe.
From the north Atlantic (most of the time.)
From Scotland, BHX and MAN heading for Spain and Portugal. This is the same for LGW and LHR, which is shown in the tracks clearly.
From LHR heading for the North Atlantic tracks (mostly.)
Overall, I’m very happy with the impact the upgrades which I’ve done. They have hugely changed the stats, which is now maxing out my location for the aircrafts seen and positions reported. But I cannot wait till summer, as my setup should be at peak then which will give me a good idea of the max I can achieve.
Out of curiosity, does the FA Pro Stick not have a bias tee or was this external bias tee required because of your location? I’m wanting to improve my system (I can learn IT stuff – and my ADS-B has taught me a lot about Raspberry Pi – but RF is a whole new thing for me). If adding LNA can help, I’ll look into it., but I don’t want to buy a part and have it not work solely because I don’t know what I’m doing!
