Wild changes in reception

[jason-a-alexander]

Hello,

I have AIS-Catcher running on a Raspberry PI 5 with a RTL-SDR v3 connected to a Shakespeare 8db offshore antenna mounted on my deck. Here are my stats from VesselFinder

https://stations.vesselfinder.com/stations/6948

I get these wild swings in the number of ships that I can see. There are sailboats in my marina about 5 miles away that when things are good I can pick up and then I have these drop off's that just wipe out my reception. Input settings are as follows:

"rtlsdr": {
"tuner": "auto",
"bandwidth": "192K",
"sample_rate": "1536K",
"biastee": false,
"rtlagc": true,
"freqoffset": 0

Thoughts?

[jvde-github]

It is a good set up.

A nearby AIS station (Hampton in VF) does not seem to have these dips, so likely not vessel traffic or weather.

Anyway, it can be weather where reception to certain pockets shifts in and out causing substantial swings in vessel count. Can also be interference from a station broadcasting on a nearby frequency, a lamp in the house, or your Raspberry Pi doing some heavy computation causing noise on the USB, or the dongle. These are tricky things to correlate.

Do you see anything in the signal level chart during these periods of high and low reception? This could point at interference.

Depending on the root cause you can take some measures (like ferrites, adding a filter etc). You already run with -a 192k. You could try as an experiment to run at a lower sample rate for a while (288K). This will negatively impact reception but will be less sensitive to nearby broadcasting.

Your weekly chart and monthly chart do not suggest a strong pattern, except that these dips typically are for a few hours?

Am afraid not very concrete....

[jvde-github]

Perhaps running at fixed tuner gain, and tracking the signal level on one of the base stations in your area gives an idea? Needs some programming though as it cannot be done easily with AIS-catcher UI.

[jvde-github]

The story of the lamp: #106

[jason-a-alexander]

Based on that article it has to be interference. Like in the lamp example now that understand better I see that my minimum signal level does increase. I have a couple of thoughts and the first I'm exploring is the fan on the active cooler on the Pi5. My current setup has a Pi 5, AirSpy R2 and RTL-SDR all in an external enclosure with a WiFi power switch. It can't be any of those individually as they are all in the box running now and my reception has returned. But the fan comes on as the box warms up. So I'll be testing that today. But I'm also within 10 miles of two large US military bases which I was thinking could be the issues but do see that other station near me is not having the same drop out so that should exclude some sort powerful transmission from the bases.

I also have an AIS notch filter coming from GPIO labs to help knock down any interference.

[jason-a-alexander]

So clearly I have an interference issue.

[jason-a-alexander]

So I tried every combination of hardware in the box today and the interfernce doesn't seem to be coming from the radio box itself. I set the fan to full power and then off and the reception didn't change. I the the CPU temp down to almost nothing to make sure it wasn't a throttling issues but honeslty AIS catcher uses only about 1% cpu and the rest of the PI is idle. I tried switching USB ports trying all four USB 2 and 3 ports. Lastly I removed the smart plug just incase the wifi on that was an issue. Unless it's the power supply it has to be external to that setup.

I have switched to trying the Airspy vs the RTL-SDR and am getting better results. The airspy is supposed to have better rejection of broadband signals so maybe I just need to wait for the GPIO filter to arrive.

[jvde-github]

I have better results with the RTL-SDR V3 as well, not sure why that is. On the RTL-SDR V3 you might test with running at -s 288K to see if a lower bandwidth helps. But from your signal level chart it looks indeed there is some noise that drowns the weaker signals from time toi time.

[jason-a-alexander]

So I'm thinking that it has to be external to the house. The noise went away this morning at 1:30. We had been asleep for couple of hours and all everything would have been stable from a configuration standpoint. All the LED lights in the house would have been off two hours before the signal came back. If GPIO labs ever sends the filter I ordered I guess that is the next step. Also thinking about ordering a 162MHZ LNA.

[jason-a-alexander]

So I'm very confused. I decided to turn of the AGC and set a fixed tuner gain for 25. Based on this image I seem to get a much wider range of reception, but yet I'm getting fewer ships. Maybe I don't fully understand now the AGC works in this instance.

[jvde-github]

Question, can the interference be caused by the antenna cable or connector close to the power and adaptor cables? The power adapters btw can also be a source of interference. Lots of cables crossing in your setup also ethernet.

[jason-a-alexander]

It's possible but I would expect the interference to be more general if it was the power adapter and not these step functions tht come and go. But the next time it happens and it's not raining I'll pull all the power outside the box an bring in a single USB C to power the PI. I have several 100PD USB adapters that should allow the PI to pull all the power it needs for testing. The adapter I'm using now is the official PI USB C adapter. I have a filter and uptronics filters LNA coming as well so if it's inbox interference I'm hoping some filters will help.

[Gazman429]

OK interference issues can be tricky to find, it can be external radio RF or local electronic RF.
From some of my experience with dealing with interference if you can rule out RF transmitter/s and RF frequency intermodulation interference, the other cause is local electronic interference, this harder to find and deal with.
It could be anything some already mentioned, switch mode power supplies ie wired and wireless chargers, laptop supplies, then there is other devices, Air Con / Heat pump control panels, TV's (Plasma TV's are bad ones for this when CAPs go dry). also not knowing how the antenna is setup even an earth fault leak when certain electrical devices turn on or in use (ground loop noise).
Electrical RF interference could even come from nearby properties and nothing in your home.
RF Transmitter issues can be detected using SDR Console or SDR Sharp and watching the water fall and you can usually see if there is RF Transmitter interference, RF electronic interference can sometimes be seen as well but it usually is fairly broad across a few megahertz and can be difficult to interpret.
I know the pain of interference, I have these problems too.
Below 3 screen shots, 1st without RF cavity filter, 2nd with RF cavity filter, 3rd a medium age electronic device with dried out CAPs, on the 1st one you can see currently mild for my current time of day here but how I'm suffering RF interference, on bad days it really can be bad.

EDIT, also forgot to add, coax cable run between receiver and antenna - how long is the run, what type of cable, this does affect both performance and resistance to interference, wrong run length with thinner cable, double or triple shielding cable too long run, can introduce increased interference problems.

[jvde-github]

I was in contact with someone that managed to remove interference from solar panels with the option -a 25K. Theoretically it does not make sense, but it is worth a try if nothing else works.... I tried it on my station and the SNR seems to improve somewhat.

[jason-a-alexander]

So I cleaned up the box and tired to separate power and signal to each side of the box. I have an uptronics filter/LNA installed that has helped as well.

I have also been running without RTLAGC and using the following settings.

{"tuner":"6","bandwidth":"0","sample_rate":"1536K","biastee":true,"rtlagc":false,"freqoffset":0}

Yesterday on a nice clear morning I was getting reception at 25+ miles which feels like the limit without skip.

[jvde-github]

"bandwidth":"0" is essentially using no bandwidth filter, you could try "25K" or "192K", but it sounds you have it under control now.

Setup look great btw.

[Gazman429]

Nice clean setup, Jasper there could be something in the -a option, think I'm going to tackle a side by side test on this, did a few quick dirty tests, on my C/D channel receiver I set BW to -a 50k from the previous 192k my vessel count went from 8 to 14 then 15 back to 14 vessels, getting C/D channel messages from vessels I didn't before.

[jvde-github]

That is cool! Not sure there is more I can do from my side, I simply call; rtlsdr_set_tuner_bandwidth but I don't have a lot of documentation how this works and for some people it seems to do miracles.

I suspect it is linked to this: https://www.rtl-sdr.com/investigating-the-adjustable-if-bandwidth-on-the-r820t-chip/

But, if you have more insights from your experiments, I am interested.

[jason-a-alexander]

The only issue I might still have is that I may be over driving the receiver. But this setting after a lot of playing with it seem to be the best for reception that I've found over the the past few weeks.

[jvde-github]

I don't think you are overdriving based on this and this min/max looks quite healthy actually.

[jason-a-alexander]

Good to know. I felt like driving the signal over zero db was a bad thing. I have to say this LNA is a nice addition. Yesterday we had some ducting going on I had a hit at 140+ miles out in the Atlantic.

https://shop.wegmatt.com/products/uputronics-filtered-preamplifier-for-ais

[sledgemhammer]

Hi,
I'm new to AIS, but got my stuff working pretty well and have plenty RF experience.
Compliments on this nice setup...

From reading this discussion :

• Theoretic do not always practical apply on RF interference experimenting as you did will help finding the best setup
• typical the noise floor drops at night that's normal, consumer electronic has less radiation at those times and atmospheric conditions are typical better.
• Make sure your antenna is resonant to the correct freq. and does not move to much. (+/- 300000/freq and some shortening factor , there are some great online calculation tools) + the higher the antenna the better : VHF = line of sight reception
• The antenna cable, especially the cheaper ones can be of influence. Shielding is what makes the difference. (And correct cable impedance !)
• In my experience the RP does influence the RTL-SDR. try using a usb extension cable and move the RTL about couple of ft away and see what the result is. (Airspy is less impacted.)
• I have a couple of ferrite bead installed on powercords and it helped a bit (not huge)
• Heating up can cause you RTL to drift a bit, watch ppm in the graph during daytime (at night there will be a cool down).
• Ducting/ atmospheric changes can be off influence, especially at night for range, also antenne location (a couple of inches can do a lot)
• Nearby pagers, cell towers or commercial FM stations can saturate the RTL, but i see a bandpass installed, so that should be handled
• In my experience keeping the dB in the graph just below zero has the best results.
• Chinese solar inverters are know for their broad radiation patterns and can cause a lot of RF problems, keep them away as far as possible.
• option : If you are able to run a tool like gqrx on your RPI (on a different image?) see whats going on the freq. to find if there is interference.

Hope the above will help to find a setup with less variation. Also watch the global site and see how others perform, it can be an indication of atmospheric changes, that can reduce and increase range. (earlier this week i had >100NM but nearby bad reception (while typical 10-15NM range)