New rtlsdr Plugin

[seandepagnier]

Another way to perform autocalibration with much less cpu would be to just look for bursts to get the initial error, then fine tune it from there. Maybe this isn't even that difficult to implement, and if we had it, we wouldn't need to pass an error correction to rtl_ais anymore as it would always just find it automatically.

[duichan]

Originally Posted by boat_alexandra

Another way to perform autocalibration with much less cpu would be to just look for bursts to get the initial error, then fine tune it from there. Maybe this isn't even that difficult to implement, and if we had it, we wouldn't need to pass an error correction to rtl_ais anymore as it would always just find it automatically.

This I like. At the moment it spends too much time and effort scanning all possible error values. It would be good to quickly establish a rough setting and then fine tune around that. Incidentally I think the final error value should also be passed to rtl_fm for VHF radiotelephony - just in case the error is particularly high.

Andrew

[piluso]

i have test the software (https://github.com/dgiardini/rtl-ais/releases/tag/v0.1) and it works, have you make a performance test between rtl_ais, AISRec and PNAIS?

No yet. Keeping in mind that the rtl_fm base program was coded for performance using integer arithmetics and many hacks for getting a low CPU usage, I don't expect a longer range than other programs, but low resources consumption.

When i use rtl_ais i have to enable the RTL chip AGC, without it not works

I start rtl_ais using -R -g28; this is a fixed gain and the RTL agc ON for a better performance. With RTL agc off, still get (less) messsages.
May be we need to start a survey to get good defaults values.

I still want to build it directly into the plugin if possible because people using windows will always have problems until this is done

I think a library could be built using MinGW and used into VC. I'll try this as a next step.

Another way to perform autocalibration with much less cpu would be to just look for bursts to get the initial error, then fine tune it from there. Maybe this isn't even that difficult to implement, and if we had it, we wouldn't need to pass an error correction to rtl_ais anymore as it would always just find it automatically.

This is like to automate a graphical method, seems to be feasible, would be great if we can code it.

[bluegoose57]

Hi,
I have a old dongle RTL2832U + E4000 tuner, and commercial MCX pigtail, which I've used with SDRConsole, GNUradio, for several years, using AisMon and AIS decoder.

My main question is, there any way to increase the audio gain to around the 65% mark, I'm getting 50% when I log the levels. I found 65% gave the best results on this dongle

Also when using -R the message rate drops off sharply.

My string currently is rtl_ais -g 42 -p -102 -n -L

Regards
Goose

[piluso]

Thanks your feedback Goose

My main question is, there any way to increase the audio gain to around the 65% mark, I'm getting 50% when I log the levels. I found 65% gave the best results on this dongle

At this moment there is not way to change the audio gain. Certainly would be a useful feature, will put this in the todo list.

Regards, David

[rgleason]

I would like to thank everybody who helped me through this process. It has not been easy. The dAISy https://www.tindie.com/products/astu...-ais-receiver/ by Adrian Studer was the answer for me, along with a very few basic facts. This receiver is quite small, almost the same size as the BNC adapter which you do need.
1. I connected it to about 20' of coax with the end modified as an AIS antenna as the Muck showed above. I did eventually solder a coax connector on the end for better signal, but the base coil and stripped end is quite simple to make and it does work. Mine was pretty rough and not at all fancy, but the measurements were right. Everything was held together with duct tape and I used a thin garden stick to keep it straight. Thanks the Muck!!
2. dAISy does not require any calibration and the driver creates a Com port which you can see in Device Manager - Com Ports shows as " dAISy AIS Reciever (Com8)" for example.
3. Installing the driver into Win 8.1 is a challenge and at first it seems difficult, but just persevere and follow the instructions, it will work. It just takes an effort the one time to get it installed. I had to do the process 3x's as I learned how to do it. Eventually I made it all the way through. Just follow the instructions to look up your Bitlocker Release Code. The installation of WinXP and 7 is very easy. Windows 10 drivers will install automatically.
4. For this test the antennas are about 200' above Mystic CT but fairly near the village but will no line of sight. The results show a number of boats AIS about 1 mile out. I am hoping that things will improve on the water and as I get this second antenna located on the boat's backstay.
5. After installing the drivers and getting the Com Port number from Device Manager > Com ports go ahead and configure the Opencpn v4.0 Options > Connections > Add Serial Port for the AIS Com Port using 38400 baud. Try the show NMEA Debug window and look for the AIS com port.
6. OpenCPN v4.1.718 does not work. AIS is not working. Don't bother trying it.
7. OpenCPN v4.0 does work. Make sure AIS icon is not crossed out.
8. If you are looking for a trouble free installation at a somewhat greater cost, I suggest you get the dAISy, it is dedicated to only AIS signals.
9. I am going to be modifying this for sending a Bluetooth signal and will report that setup.
10. I hope the range gets better with a better antenna and I will report on my experience with that too.
11. Adrian has been very helpful and prompt in responding and the mail delivery of the dAISy occurred precisely as he predicted. With the BNC connector, the proper cable was included and there was an alternative cover with a tab for mounting the device on a wall. The device is quite small.

[rgleason]

Went up to the attic / cupola and tried AIS. This location has much better direct line of sight out to Fisher's Island (5nm), Watch Hill Point (7nm), Montauk Point (17nm)

I can now pick up AIS 3nm away in Stonington Harbor. It is night. I need to make this work better. Don't know what the problem is, maybe my antenna.

[the_muck]

Hey,
with AISRec i have a range up to 35nm, when i change my antenna the min. range was ~27nm (Dipol), with my FPA ~35nm. But the antenna need a position with a free look to the targets . So i think the difficult part is the filter algorithm in the software. The interesting thing is that sometime AISRec find ships and rtl_ais & PNAIS find it not. It looks like that the software do not like some signals , this is not a range problem . Sometime i miss signals 2nm away from me, big ships, a ferry or cargo ship.

So when i try new software/hardware AISRec is my reference, with out a reference it is not easy to look for problems, antenna, connectors...

[rgleason]

Will make new better antenna and try AISrec

[Maelruan]

Hi, I reckon AISRec to give the best performance with RTLSDR dongles, even though I've managed to download only the trial version so far. It is also far more user friendly than the range of requirements with SDRSharp+VBCable+AISMon/ShipPlotter that I've used up to now, but which I found very useful finding my way in very thick fog thru' a fleet of trawlers last season!
The images attached are arranged to demonstrate finding 31 targets this afternoon at ranges up to 47.4 nm (many behind intervening landmasses). The antenna used is a homebrew Slim Jim from 300 Ohm Ribbon cable which has been hoisted as shown for nearly 2 yrs now, and without the PVC tubing called for in the build instructions.

[piluso]

The images attached are arranged to demonstrate finding 31 targets this afternoon at ranges up to 47.4 nm

While sailing I saw AIS targets at distances > 30 nmi using the one channel rtl_sdr plugin, with an antenna 2 mts over the water surface.
I was very surprised, because that was beyond the range of any VHF and the fact that is a single channel receiver. After a little googling, I could confirm that some base stations are also repeaters; so one can see AIS targets very far, but in the range of the base station.

I don't know if there is a way to see if a message comes from a repeater or a target; but is easy to get confused about the actual range of your AIS device :-)

Personally, I think that a reliable behavior in a range of 8 nmi is good enough for sailing (I adjust the AIS radar plugin range at 2nmi) and the rtl_sdr plugin works fine for this purpose.

Vessel tracking however, needs an increased range, probably there are better solutions for this.

[Maelruan]

Tks for the heads up Piluso.
I can confirm however, that there are no confusing repeaters this end.
Also, there are large high speed ferries doing 40 kts in these waters which I prefer to keep at about 2 nm distant (See quoted warnings below). And as my max speed is only 5 kts I like to be aware of them before they are 24 mins away, which is at a range of 16 nm, or more.

"Beware ferries which follow a 'parallel' course! Two HSS and two displacement ferries ply the route on a regular basis (about 12 crossings per day in each direction). Instructions in Notices to Mariners say that yachts should NOT alter course if they suspect a collision situation is developing because the ferry's navigation computers will assess the most appropriate action for the ferry to take in order to miss you (sic!). No further comment is needed or possible!"

[rgleason]

This morning I decided to try sticking my very rough duct taped AIS collinear antenna a la "the Muck" out the window horizontally in downtown Boston to see what would happen. I was pleasantly surprised by the 20 ships within about 2.5 nm

Today I found a need for the built in Connections >Show Nmea data stream. I would like to be able to "filter" certain connections so that they are NOT shown. That way I can see the AIS data stream.

-Re: AISrec I found that software to be from China and I wonder if it is "clean" considering the source. Does anyone have enough experience with it or know the fellow? I did not download it. Glad it works so well.

With a better antenna I think this will work. I hope the dAISy will go further than 2.5nm with a better antenna and not in the 60'-600' canyons of the city.

I have started to build a 162 mhz High Gain Collinear based on Neal Arundale's
High Gain Collinear AIS (162MHz) Receiving Aerial and I don't know if I should make 4 sections 611mm or do the larger one of 8 sections 611mm. Does anyone know the answer? Is the 4 section adequate?

Here is a picture. Does anyone know how to get the AIS List text smaller, I have tried setting all of the UI choices I think.

[LeoA]

I get as many or more targets reported with rtl_ais as I do with aisrec

Sent from my XT1031 using Cruisers Sailing Forum mobile app

[the_muck]

Originally Posted by LeoA

I get as many or more targets reported with rtl_ais as I do with aisrec

Sent from my XT1031 using Cruisers Sailing Forum mobile app

That is interesting... on this weekend i can test a bit more . can you tell something about the stick and start oparators?