Using AIS-catcher graphs to fine tune an RTL Dongle

[g7ruh]

I just want to share what happened and how I improved reception using AIS-catcher Signal Level and Frequency Shift graphs. This may help inform other discussions about how to go about automatically setting gain.

My equipment:

antenna ---- VHF/UHF splitter ---- 162MHz notch filter ---- RTL Dongle ---- Raspberry Pi 4 (1)
                              |
                              |---- Airspy Mini ---- Raspberry Pi 4 (2)

so both receivers get the same signal from the antenna.

I will be removing the notch filter at some point to test the RTL Dongle with wide band receive to see if there is any difference. If there is I will update this post.

The goal: I wanted to get best performance out of AIS-catcher.

I used the Airspy Mini as an experiment 'control'. I gradually increased -gm linearity in the Mini until I did not see any further increase in number of vessels received when compared with the RTL Dongle.

While I was doing this I noticed 3 Tugs alongside a local jetty on the Mini but only one showing on the RTL. They are about 2.5NMi from the receiver. Signal overload, frequency shift in the receiver, bad transmitter on the vessel were possibilities amongst others.

I had been experimenting with rtlagc on and determined that it kept the upper signal level below 0db. I turned the AGC off and changed the -gr tuner nn.n setting downwards until a drop off in vessels received (compared with the Mini) was noticed. I then gradually (over several days) increased the gain until it was above 0db when I noticed a drop off in vessels received. This was expected as I was most likely overloading the RTL Dongle front end.

I then lowered the gain setting until it was mostly below 0db

At which point the number of vessels received was about the same as the Mini (tended to be 10% or so less before I started), so: a little improvement.

The missing Tugs were still an issue: Why?

Looking at the graphs below:

the PPM per hour is annotated:

PPM25 was the setting when I started adjusting gain: it was determined from the RTL-SDR calibration process, waiting for the PPM setting to stabilise where I arrived at a value of 25 which I had been using for some months. You will notice that this shows on the AIS-catcher graph as a frequency shift of 2-3 PPM per hour.

This got me thinking about how to reduce the shift if the graph was showing a significant deviation from expected frequency.

I modified the PPM setting to 23 (as in the graph above) and immediately got a huge drop of in number of vessels received. I had found an upper boundary for the PPM setting for this dongle.

I then changed it to 27 (the reverse of the last step) and noticed a significant reduction in the PPM / hour value and a small increase in the number of vessels received.

I decided to increase PPM by 1 to 28 as it looked like the trend was good.

Referring to the right hand graph above, you can see AIS-catcher reporting a reasonably stable PPM equally above and below zero.

Below are the graphs at the time of writing this post five days after the previous graphs. Ambient temperature over the last 24 hours has been between -2 and +8 Deg C. Since the earlier RTL Dongles like mine are known to be less stable, temperature-wise, the results are reasonable, in that the PPM difference is greater, but still shows similar deviation above and below 0 PPM over 24 Hours

GRAPHS FROM RTL DONGLE:

By comparison here are the graphs from the Airspy Mini for the same 24 hour period as of the time of this post.

GRAPHS FROM AIRSPY MINI as a comparison:

The deviation is much less, as expected of the higher spec Airspy Mini.

The overall shape of the 24 Hr graph is similar in the RTL Dongle and Airspy Mini, just the amount of variation in the Mini is much less.

The two receivers are located next to each other.

What about the Tugs? Well after the PPM settings changes, the RLT Dongle shows three tugs, just like the Mini! Whatever the root cause, geting the RTL Dongle working as close to the correct frequency is important for AIS.

I now get at least 10% more messages received on the RTL Dongle than the Airspy Mini but not a noticable increase in number of vessels.

Conclusions:

The early RTL Dongles are very good, but very sensitive to proper calibration for AIS reception. My original calibration method did not produce anything like the accuracy that using AIS-catcher graphs and a little thinking did.

I did not expect to get many more vessels, as the vessels are in range for a long while, so both receivers are likely to see most vessels. It is just that weak signals will make it through to decode if there is an optimal receive path.

I will be interested to see what happens in summer as The Solent is known for its small boats and is very busy. It is likely that the calibrated RTL Dongle will dig a lot more of them out of the noise, pleasure craft use lower power transmitters.

The Mini is more stable, but costs a lot more than the old RTL Dongle did. Given that the RTL Dongle is doing such a good job, I will repurpose the Mini for other use

I hope this may be useful information in improving AIS-catcher. Thanks to Jasper for a great bit of software and excellent support too.

Roger

[Taxom]

It is worth considering that the frequency shift of the AIS-сatcher calculates as the average arithmetic from the receiving signals, and not from any reference signal. Accordingly, it depends on the error of each generator on the received vessels. So if you have the AIS base station in the receiving zone, it is better to calculate the frequency shift of their signals. This is clearly visible to me, since I receive several base stations, the frequency of my receiver relative to the base stations approximately +-0.2PPM, but the total error including the signals of the vessels are +-0.7PPM. The reason is that there are sometimes ships with a strong frequency shift in the receiving zone.

[g7ruh]

Yes as a base line the base stations show that well. In addition to that, I was just looking at my post and the diurnal variation for the last 24 hours as your post arrived.

Looking at that, my supposition that the original calibration was not very accurate could be incorrect. I recall I did that calibration when it was considerably warmer, so in summer the PPM calibration may well be closer to my original of 25 PPM and the overall swing of PPM correction values may be 24-28 PPM.

The equipment is located in an unheated outbuilding so the annual variation in temperature will be about 30 deg C. About freezing to plus 30 would be a reasonable average.

Given that I found a small decrease in PPM caused a huge drop-off in messges, I am wondering if the calibration is set at one end of the range it will cause problems at the other.

Of course all the different low-end dongles will most likely have different temperature related profiles. All the more reason for seeing if we can get a bit more control over the PPM figure, maybe by using the diurnal variation as a measure.

I will wait for a day or two when the diurnal variation is very small and see if that shows a lower PPM change over 24 hours. I am expecting it will.

Roger

[jvde-github]

Thanks for sharing these observations. Very useful!

The program is quite sensitive to frequency shifts as it only corrects for shifts in a narrow window (+/- 5 ppm). There is a setting to allow for wider corrections (-o AFC_WIDE on) but I noticed that a smaller window worked well on modern sticks and commercial vessels. So more messages and longer distances but perhaps less ideal for Class B equipment. The other consequence is that it is more sensitive to drift as you show. In principle the graph that you shared could be used for the model to auto calibrate itself but my impression is that you need quite a few messages to get a stable estimate. Also does not solve the problem of Class B senders which are off key.

Look forward to your further investigations.

[g7ruh]

Thanks Jasper,

The least I can do for all your efforts in providing a great bit of software, having only recently got into AIS, last October.

I will provide a few more graphs from the plot page where diurnal variation is greater or lesser than today's graph to help get some idea how much change there is in ambient temperature impacting devices with higher than expected PPM variations. Typically, during Feb and April we get some high diurnal variations e.g. 0 Deg night and 10-15 Deg day.

I suspect quite a lot of temperature related variation, however that also means that a gradual change in PPM by AIS-catcher will keep a Dongle within useful bounds WRT maximising performance. A gradual change in PPM compensation may well give a lot more dongles a very good performance. Of course, at present, I am basing this on a 'small sample of one'!

I agree with your comment on -o AFC_WIDE, I have limited myself to working on the current setup to get the best out of the dongle. This will provide some data to test PPM variation and lower but more reliable data / number of messages received. When there is enough data, I will see if using that setting provides any more stable benefit without impacting the maximum message / vessel numbers too much.

If you think of anything I can help idea-wise, just ask. I may well ask you to help collect more or different data to help analyse what is happening too!

Roger

PS, there may be a few new feature requests coming in as a 'quid pro quo'!

[g7ruh]

An update on the main post:

For some reason the Pi 4 I am using seems to hang up. It has done it a few times. Given that it is the live feed to tracking sites, I decided to swap it for another Pi 4. I am investigating the reason. The cause could be hardware, software or external e.g. temperature or mains voltage issues. It is running AIS-catcher from the RTL Dongle, a script to feed a dummy GPS message to OpenCPN and Spyserver connected to an Airspy Mini for Marine voice channels.

There is another Pi 4 running my test system nearby. I have not had any problems with that one. It stays as it is for the moment running my test system.

As a first elimination step I decided to change the Pi 4 unit. I had one being used as a media server - Yes, these were obtained before "The Great Pi Shortage" happened.

Having powered it off and replaced the unit and reconnected the USB devices, I noticed the AIS-catcher plot of messages per second: it was very low (as in one or two per second). I looked at the Frequency Shift per minute and it was 5 PPM (graph below), too far from optimum of about zero.

After about 50 mins, I captured the graph and the corresponding Messages Per Minute. Careful inspection shows a lag before the number of messages increases and this is when the PPM shift goes down a little.

The level of about 1 PPM is similar to the graph in the main post when the ambient temperature is about 0 deg C in both cases. The RTL Dongle was not moved from its position at any time.

Another bit of evidence about ambient temperature change impacting low-end dongle performance. It also shows that some form of software compensation may help these devices work year-round without much loss of performance or manual adjustment of the PPM value.

Roger

[g7ruh]

Update while waiting for a stable day / night temperature: still waiting!

Last night the temperature dropped 6 deg C in about 30 minutes at about 04:00. This provides a useful variation to inspect the impact on the hourly PPM change graph. Using a black RTL dongle.

Annotated graph below:

I have also noticed that +5 or -5 PPM or more will change the receive frequency enough that message decoding is significantly impacted. Using rtl_test -p60 (for example) to arrive at about the right place is a vital starting point: then use the PPM per minute graph to adjust to near the zero line.

I now suspect that a very temperature sensitive dongle (like my one under test) may benefit from some automatic PPM adjustment. This is because it will drift far more than the +/- 5 PPM range AIS-catcher allows. For example: a dongle calibrated in winter may not work in summer and AIS performance will degrade over time as the PPM limit is reached with temperature variation.

Roger

[g7ruh]

Another interesting comparision.
Temperature range +15 to +7 C over about 20 hours.
Two RTL black cased dongles SN 300 and SN 111
SN111 ppm=28 SN300 ppm=79

The graph of ave PPM per hour are the same shape (more or less) but opposite in direction. The similarity is interesting as I suspect SN300 to be more sensitive to temperature changes than SN111. These graphs suggest similar ppm variation with temperature chnage.

The range is identical zero to minus 1.4 in both graphs.

Roger

[g7ruh]

Observations on PPM change with increasing daytime temperature.

To recap:

Using 2 RTL black dongles SN111 and SN300. Also using Orange FlightAware dongle rated at 0.5 PPM, SN1000. All fed via a VHF splitter so from the same RF source.

The PPM was set early Feb this year. The folowing observations are using RTL SN111 and Orange FA dongles.

Normal conditions when PPM is within 0.5 of its settting in AIS-catcher startup, the Orange FA dongle does not hear as many messaages as the RTL SN111 dongle. It may be slightly deaf by comparision. The original post explains how I set gain and initial PPM.

Very subjective observation, but results as expected. When the RTL dongle Fqy Shift (avg PPM / min) is greater than plus / minus 1.5 ppm, the Orange FA dongle starts to perform better than the RTL SN111 dongle. At 2ppm difference there is a noticable fall off in the RTL messages received.

See graphs below:

If AIS-catcher provides auto-calibration based on PPM change, it will need to kick in with a change of PPM of about +/- 1.0
During initial dongle calibration PPM of 3 / -3 from the zero PPM setting caused a significant drop-off in messages received. It looks like that starts to kick in at about 1.5 and is noticable at 2.0.

A starting point to automate calibration may be to change the PPM offset when it is greater than +/- 1 and consider a change every 5 hours. Obvioulsy this will vary by dongle type and sensitivity to temperature, but this would enable testing on my two test subjects, SN111 and SN300.

Another option is to include a temperature sensor nearby the dongle and monitor change of temperature and of PPM value for data collection. It would be good to record the temperature and show it on the Fqy Shift (avg PPM / min) graph.

Roger

[g7ruh]

Update following testing of the new -m 4 -go AFC_WIDE on option.

Enabled on the live system RTL dongle V2 SN111 the messages per hour graph shows slightly more than the control Orange FlightAware dongle. This is the same as before enabling the option. Conclusion: with little PPM variation (less than 1.5) there is no real impact on messages received.

Further testing is required on the RTL Dongle serial 300 which has a greater PPM change with ambient temperature to simulate a wider temperature variation.

Further update when that testing is complete.

Roger

[jvde-github]

Thanks Roger. I think if PPM for dongle is well under control there is no benefit for the new engine, unless there are some senders that are a bit out of the spec. Your test gives confidence that there is also nothing lost if we use the more robust engine as default....

Cheers and thanks for sharing and testing, Jasper

[g7ruh]

Update after testing "-go AFC_WIDE on" setting having changed PPM setting to simulate temperature induced PPM change.

Normal PPM setting giving +/- 1 PPM change in 24 hrs is 82

Key:
A = -go AFC_WIDE on PPM = 78
B = -go AFC_WIDE off PPM = 78
C = -go AFC_WIDE on PPM = 82 (normal setting for this dongle)

Observations:

The PPM setting did not change when -go AFC_WIDE is changed, but the reported PPM in the Fqy Shift per min graph was about 5 with it on and about 3 with it off. Should this happen or should it be about the same as the PPM has not changed? Program issue or just variation in how the data is calculated?

The SN300 dongle is set to maximum gain and still hears slightly fewer messages than the live system SN111. For reference the same graph from SN111 is below, the reference / comparison is the shape of the Messages per Minute graph in each picture. Both devices fed from the same antenna via a splitter.

Testing continued, gradually changing the PPM further away from the best setting (82). Subjective observations: the -go AFC_WIDE on option works with little loss of messages up to +/- 6PPM deviation and would seem to be OK up to at least +/- 10 PPM deviation from a calibrated zero PPM change point.

This should mean that a non-temperature compensated dongle which suffers from a wide PPM change with changes in ambient temperature will perform acceptably over a much wider temperature range when -go AFC_WIDE on is used with little impact on reception at least for a PPM variation of up to +/- 10 PPM.

Note that the above has been tested with one rather temperature-sensitive RTL Version 2 Dongle, the above findings may not apply to all dongles but the setting is well worth trying.

Huge thanks to Jasper for developing this option as a simple way to help improve reception for some (temperature sensitive) dongles.

Roger

[Gazman429]

In a perfect world it would be 0, one thing with PPM is it will vary a bit due to transmitted signals being received are not on spot frequency, all transmitting devices can vary a few kHz and that would be seen on PPM jumps, I see the same thing on ADS-B and satellite ADS-C, RTL dongles while being a great versatile device does have flaws as their design was for TV and broadcast radio reception so 20 / 30 kHz drift was not an issue with temperature changes, the after design to add a TXCO to help stabilise temperature drift is about the best it will get as the chip itself has internal designs that temperature change cannot be compensated for, the RTL dongles were never designed for what we use it for, there will always be temperature drift the addition of a TXCO helps reduce drift,

[Taxom]

In a perfect world it would be 0, one thing with PPM is it will vary a bit due to transmitted signals being received are not on spot frequency, all transmitting devices can vary a few kHz and that would be seen on PPM jumps, I see the same thing on ADS-B and satellite ADS-C, RTL dongles while being a great versatile device does have flaws as their design was for TV and broadcast radio reception so 20 / 30 kHz drift was not an issue with temperature changes, the after design to add a TXCO to help stabilise temperature drift is about the best it will get as the chip itself has internal designs that temperature change cannot be compensated for, the RTL dongles were never designed for what we use it for, there will always be temperature drift the addition of a TXCO helps reduce drift,

No, the drift depends only on the accuracy (stability) of the crystal oscillator, not on the chip's temperature.

And no, what you see as "drift" in AIS-catcher is not always the drift of your own device's frequency.
It is calculated as the average of all received signals, so if one of the AIS transmitters has a strong drift and is transmitting frequently (like a high-speed vessel), you will see it reflected on the graph.

[jvde-github]

Thank you Roger for extensive testing. This seems to be in line with my theoretical tests. I have made the default for the switch "on" as per version 0.48 as it does not negatively impact the performance too much for stable dongles but adds additional stability for less accurately tuned dongles.

"The PPM setting did not change when -go AFC_WIDE is changed, but the reported PPM in the Fqy Shift per min graph was about 5 with it on and about 3 with it off. Should this happen or should it be about the same as the PPM has not changed? Program issue or just variation in how the data is calculated?"

I think this behaviour can be explained. With WIDE on, the decoder will pick up all messages and therefore the ppm correction average is around 5. When WIDE is off, we will not pick up the messages with a PPM correction of >5. So, the average over the subset will be ~3 and biased.

Thanks again, your support is much appreciated!

[g7ruh]

Jasper,

Glad to help make a better AIS solution. Keep up the good work.

Thanks for explaining the PPM difference, makes sense to me.

Roger

[NR857]

I have been using plots to fine tune my Nooelec Smart SDR and report very good results but definí there is a pattern of PPM for various zones, obviously dependent on conditions. But I would also comment that as always with antennae, "height is might". This evening I changed tge aerial at my home location to 2m higher and immediately recorded an increase to a maximum distance of 60 nautical miles!!

Screenshot_20240424-204447_Chrome.jpg

[Phil353556]

@g7ruh A huge thanks for yours tests and explanations,
@jvde-github a ton for you, for the program aiscatcher and your availability, that's great!

My feedback I should like to share.
HW :
Host: Raspberry Pi 4 Model B Rev 1.4
OS: Debian GNU/Linux 12 (bookworm) aarch64
Kernel: 6.6.20+rpt-rpi-v8
Dongle: a blue one with USB RTL2832 with tune R860 (reference https://www.passion-radio.fr/recepteurs-sdr/rtl-sdr-r820t2-248.html)
Antenna: a very basic one (reference https://www.passion-radio.fr/antennes-sdr/telescopique-mcx-325.html)
Total for Dongle + Antenna: 36€ delivered to my home, so I perfectly now this is not a professional configuration.

@NR857 you're right, one week ago, I moved the antenna 2meters up and the resultt is incredible

My goal now is to tune the SDR with the (in bold) parameters in file /usr/share/aiscatcher/aiscatcher.conf
The current graphics are:

And the default parameters used:
-v 10
-M DT
-gr TUNER 38.6 RTLAGC off
-s 2304k
-p 3
-o 4
-u 5.9.207.224 xxxxx
-u 31.172.90.245 xxxxx
-u 195.201.71.220 xxxxx
-N 8383 FILE stat.bin BACKUP 10
-N PLUGIN_DIR /usr/share/aiscatcher/my-plugins
-N STATION ZZZ LAT YYY LON -XXX SHARE_LOC on
-X xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Reading you @g7ruh and looking at the graphics, Does that means -p 30 ou -p 33 should be used?
Reading @jvde-github github webpage, the parameter for -gr should be -gr RTLAGC on TUNER auto -a 192K?

Thank you for time and answers

[g7ruh]

Hi, thank you for your kind words, I am glad my tests have helped you.

Several observations, no particular order of importance:

As I understand it, the RTL-SDR dongles and clones AGC do not handle short data packets very well, it is designed for longer length signals, e.g. voice (ssb or am radio). My experience with both AIS and ADSB (1090MHz aircraft data) follows that, messages per second (for example) is lower when the dongle AGC is used.

As per posts above, any signals above 0db will not be decoded very well. There is a trade-off here: given normal reception conditions, a high power signal (very close) may overload the receiver but if they are less frequent, a higher gain setting will miss those but get the more distant signals. I have that at my location: strong (close to overload the dongle) from vessels about 2NM away but mostly the more distant Class A (commercial, using high power) or local sailing craft (Class B) using lower power work well most of the time. It is a gradual modification of the gain that guides you here: it takes time. I take a few above 0db as they are the nearby (2-3NM) vessels.

Your question about what PPM to set: yes about 30-33 would be what I would use based on the graph you provided. It does raise another question: with a setting of -p 3 and you are seeing a fqy shift of about 33, I am surprised you are seeing so many vessels. Not knowing your location (roughly) I cannot check using another site. Maybe you will see more when you tune your system!

From my experience a deviation of more than + / - 3ppm is enough to stop AIS-catcher doing a good job. I would suggest changing the -p setting by e.g. 6 ppm every hour or so (while the number of vessels is reasonably stable over a few hours) and look at the messages per hour graph and see if there is a notable difference. If it increases, keep going, and stop when it starts declining from the peak. Then leave it there for a while to get a longer term view.

Of course you could try changing to e.g. -p 30 and see what happens and vary from there up and down.

You should check the day / night variation by looking at the frequency shift per hour to see if there is a large variation over night as well. That will help inform you about the temperature stability of the dongle. I am now using a rtl-sdr V3 dongle which is temperature compensated. Over the winter (UK south coast), it has been within the capability of AIS-catcher to handle it.

You can see the current graphs here https://aiscatcher.org/blackfield01/?lat=50.8200&lon=-1.3700&zoom=10.00&tab=plots

EDIT: for some reason the link did not work as a insertion and it does not give the graphs until refreshed, click on it and refresh and you should see the graphs.

The ppm is a little below zero but still well within AIS-catcher reception capability.

On a busy day I get over 1000 messages per minute here.

hope that helps

kind regards
Roger G7RUH

[NR857]

Very useful Roger

73s

[Phil353556]

Hi Roger,
thanks again!

Some modifications today, will see tomorrow let's time to time
Clearly more AIS messages, will see in the next days about the number of vessels.

Your location is excellent, many many vessels seeing, Solent is a real motorway for vessels.
Location of this receiver, heading around 190° from yours: Saint-Malo, France.

I noticed something interesting, on https://aiscatcher.org chosse a vessel close to Saint-Malo, go to details
and then for example: https://aiscatcher.org/ship/details/241784000
the AIS Source is a link (my receiver)

do the same, close to your location, choose a vesell and so on, the AIS sources is not links (222,489)
Do you know why?

Kind regards
Philippe

[jvde-github]

The -u feed is only needed if you want to keep the separate dashboard (aiscatcher.org/blackfield01) If you are happy with the direction of the new dashboard (will improve and expand) we can stop that if you want. Saves some bandwidth.

[jvde-github]

Hi Roger, maybe to clarify, these are two different systems. One is the online variant of the local dashboards and the other is part of a new community feed system. The first auto feeds in the other. So it is a choice what you want. There is a reason for both

[g7ruh]

Thanks Jasper, that makes sense.
That is what I thought but I was not sure.
Roger

[g7ruh]

@Phil353556

I had a look at your coverage this morning and compared it with MarineTraffic. You can see that your coverage is not bad and the 25-30NM range is what I would expect for VHF without any enhanced reception. Class B vessels typically use 2 watt transmitters and Class A 25 watts. That would explain the lack of Class B in a line from Jersey to Granville in your coverage. MT is fed by quite a few stations along the coast and on Jersey.

Your coverage:

MarineTraffic

So not bad coverage, but don't stop trying to improve it. I gradually squeeze my system to see how I can make it better. Jasper has helped too by continuing to improve AIS-catcher.

Regards
Roger

[Phil353556]

Hi Roger,
Thank you so much, I appreciate. You took time for me.

I looked at this morning too.
Here is some facts and remarks.

The data from aiscatcher is fed to aiscatcher of course, and also to others (see #75 (comment)) : Marine Traffic, VesselFinder and ShipXplorer.
I plan to make a comparison slide between all four, in a few days.

As the data is the same for all, I really don't understand why the number of vessels is so different. For MT, I exchanged with them and it is on the first in first counted, so Ok but strange. For vessel finder the number is significatly lower than aiscatcher, but it is easy to see which receivers are used for one vessel (see screenshot in this case two : n°6783 me and n°5832 Jersey). For shipXplorer the number is quite the same as aiscatcher.

ex number of vessels seeing at 04:23PM local time:

• AISCatcher: 79
• MT: 12
• VesselFinder: 44
• ShipXplorer: 74
  The interface of ShipXplorer is nice too and the station range cleary indicated see screeeshot
  

In terms of number of messages, it increases, thank you!
It is possible to clearly see two steps:
-"as always with antennae, "height is might" as written by @NR857
-fine tuning (thank again)

I think it is possible to add some PPM to the -p option, currently -p 32, I will test -p 35 on monday.
It is a good idea to have temperature sensor close to the dongle, I will think about it. However the antenna is under the root of a house (slate roof) but the dongle and the pi are just under the ceiling, very close to the top. The house is normaly well isolated so the temperature should move a lot. It is possible to monitor the temperature of the pi itself with command vcgencmd measure_temp

Let see how it will be next week.
THANKS to @jvde-github for this nice piece of software with updates and his availabiliy that's great.

kind regards
Philippe

[g7ruh]

Windows test.

I have not tried the Windows version of AIS-catcher before, so a useful thing to do.

2 V3 dongles purchased at the same time so assume they may be from the same manfacture batch.

The only difference is that the Linux one goes through a 162 MHz notch filter, the Windows one uses the same signal path apart from that. (See first post for detail). Therefore the gain settings are different. Windows system only been running 18 hours and gain may need fine tuning. Other settings are the same.

The screen grabs are taken about the same time but an update or more between them. The data shows the same number of vessels (or very nearly) and the graphs look similar.

Windows gain may not be perfect, but close enough to show that both systems give very similar results. Not that I had any reason to think otherwise! However, comparing different architectures without confirming similarity is risky.

Conclusion: I see the same results from both the Pi and Windows based AIS-catcher systems in this test.

Linux count

Windows count

Linux plots

Windows plots

[jvde-github]

Good to see. Have the same experience but with cheaper hardware (like early Raspberries) the quality and noise generated by the USB ports could impact reception. Also poor shielding. But the OS and CPU should not have an impact. Great to see confirmation.

[g7ruh]

@markdj57

Are you a member of the shipplotter group? If so did you see post number 24251 which discusses expereince with different receivers. TL;DR: V3 RTL dongles are as good or better than more expensive ones, and V4 looks good but not all AIS related software supports it yet.

With reference to your station performance graph on aiscatcher.org looking at 'Day' plot, there is a very nice shape of graph as you have gradually tuned your setup. Much progress, well done.

Roger

[markdj57]

I'm not on that group, I'll have a look.

I did have an idea about a more effective means of tuning, if I could listen to the incoming stream and adjust the gain to have the best signal, would that work? is it possible to do it by listening? I'm still playing with gain settings a bit to see if I can stretch the maximum distance reliably as unlike the Solent, the ships here are 20+ miles away with the eastern Irish Sea towards England and the Isle of Man going further.

I found this calculator very useful...

https://www.omnicalculator.com/physics/distance-to-horizon

So I can see that if I put the antenna up by 5 metres, which is huge for an unsupported stick, distance to the horizon only goes up by 1.4 nautical miles, not worth doing imho. I think I will change my antenna for a larger one though.

[g7ruh]

Mark,

For VHF transmissions (including 162 MHz) the 'rule of thumb' for distance is the horizon plus a bit as the signal is refracted somewhat following the earth's curvature. That is assuming that both TX and RX stations are on the surface of the earth. In most cases that is modified by the receiving station (you) and the antenna height above ASL as a 'reference' point. That is further modified by the transmitting vessel and it's aerial height above ASL. In the case of a small Class B vessel it would be not too much above sea level. If it is a England - NI ferry the aerial will be quite a lot above sea level: at the top of the superstructure, typically above the bridge. In the case of a container vessel: the large ones 300m length overall or larger, then they are even higher, so the aerial will be higher still. They are quite tall! So at distance (as you say above 20 NM miles or more away) you are at the end of the line of sight plus curvature effects. Then it is down to the height above ASL for the RX and TX station. A good example fo this effect is when there are not too many vessels about and looking at your aiscatcher.org Distance chart, I see a high point slowly going down in a gradually decreasing graph height as the vessel gets closer to you (and the reverse as a vessel gets closer to England). This is because you do not have too many vessels so the max distance is set by that single ferry gradually increasing (or decreasing) its distance from your station.

Listening to the signal is not going to be easy to determine best gain setting. You do not know the signal origin so you do not know if it is a Class B transmission (2 watts) or a Class A one (12.5 watts) or how far they are away from you. The best way to do this is from the signal strength graphs as that tries to set a level playing field signal strength-wise for your station.

Also, the data transmission is so short it is hard to compare one with the other. It is not like SWL (short wave listening) where you have a long duration transmission (broadcast radio) which allows a better comparison signal-wise in you head.

Keep in view that atmospheric enhancement can change things dramatically while you are testing. I use the number of fixed (base and AtoN) as a way to see if that is an issue (determine the 'normal number') or as a way to see better conditions. You need to discount that before you start changing things.

Aerial height: rule of thumb higher is better. but after so much it does not make so much difference. Other things to consider are:

using good quality coax (expensive stuff) from aerial to dongle (limits signal loss), putting the dongle very close to the aerial (so long coax run of expensive (low loss) cable is not needed. Using a Raspberry Pi attached to the dongle and short link to aerial then sending that data from the 'Pi to the local network. Using a filter / preamp near the aerial (needs voltage to power it) and the signal goes further without so much signal loss when it gets to the receiver. To do that you need Bias-T, that combines a DC voltage to power the pre-amp / filter and the signal from the aerial.

Some dongles provide a limited Bias-T voltage but typically cannot power the Uputronics Filtered Preamps as an example, the Bias-T on the device cannot supply enough current to that device. That filter pre-amp is what I use, powered by a separate USB feed to the filter / pre-amp.

Aerials with more gain may work better (or a pre-amp like above) may do just as well. All depend on the aerial quality or preamp quality.

All of the above are enhancements / trade-offs depending on what fits your set-up. Testing and trying is what makes it fun!

I hope that helps answer your questions above.

regards
Roger

[NR857]

Mark, If you want to try a different antenna and you have a geographic limitation i.e. hills or mountains, why not try a two or three element Yagi? Failing that a good marine antenna works wonders. I use a Banten 6dBi 2.7m whip. I notice that it doesn't give me that much extra range, because of my local mountain landscape, but it does pick up more class B vessels. I've also used an old high band VHF TV antenna tuned down to 156/162, with a mast head preamp. That was also effective. To be honest I think a lot is down to topography. I'm in Dublin at the moment near the airport, 80m ASL and I'm getting everything in the port plus a bit further out to about 15 nm. I understand the point about using fixed stations as a guide but I think they sometimes can be misleading as they are usually on hills etc. For example I'm picking up the Irish Coastguard remote station at Carlingford, 39 nm, but it's at 500 as far as I know. My advice is to look closely at the hill shadows etc. Best regards, Kieran Sedella AIS mobile

…

On Thu, 17 Apr 2025, 14:44 markdj57, ***@***.***> wrote: I'm not on that group, I'll have a look. I did have an idea about a more effective means of tuning, if I could listen to the incoming stream and adjust the gain to have the best signal, would that work? is it possible to do it by listening? I'm still playing with gain settings a bit to see if I can stretch the maximum distance reliably as unlike the Solent, the ships here are 20+ miles away with the eastern Irish Sea towards England and the Isle of Man going further. I found this calculator very useful... https://www.omnicalculator.com/physics/distance-to-horizon So I can see that if I put the antenna up by 5 metres, which is huge for an unsupported stick, distance to the horizon only goes up by 1.4 nautical miles, not worth doing imho. I think I will change my antenna for a larger one though. — Reply to this email directly, view it on GitHub <#75 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BH3OCY7PJYLUDND6Z7PDED32Z6O2HAVCNFSM6AAAAAB2PYEI6KVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTEOBWG4YDOMQ> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[markdj57]

So I've left things alone and just don the odd adjustment of gain. I do have a pre-amp/filter installed now which runs with the biastee on but it just means I can lower the gain, I'm not clear if it has improved anything or made reception of ais b more reliable (there is a yacht at 5 miles that still goes in and out of reception at times that is constantly on). According to the spec on the preamp/filter website and the rtl-sdr spec, it can run with biastee on and that is confirmed with the green light on the unit itself.

Even with the filter and magnetic chokes, I am still getting odd things happening with large drops in the number of ships received at times which coincides with the ppm going negative and this can last for 40 minutes and then the SNR bottom graph change dramatically too.

I put magnetic chokes on the usb and 12v power supply for the miniPC but that didn't seem to do anything, so much like in the other post in this group discussing wild changes in reception, it doesn't seem to have an obvious source or cause.

I'm hoping to upgrade to a longer antenna in the next few weeks, still deciding on what to get and discussing with a local radio comms firm who seem experienced that field.

[g7ruh]

Hi @markdj57
please can you post another set of graphs like above as soon as possible. I am interested if the lack of vessels is related to your problem or just lack of vessels. I have been looking at aiscatcher.org graphs from your station today.

thanks,
Roger

[markdj57]

Hi @markdj57 please can you post another set of graphs like above as soon as possible. I am interested if the lack of vessels is related to your problem or just lack of vessels. I have been looking at aiscatcher.org graphs from your station today.

thanks, Roger

Thank you for continuing your persistence into looking at this. I had looked a couple of times and I think it looked like just a lack of vessels. I have seen the dramatic drop in vessels received yesterday that lasted about 20 minutes each time, which is definitely a rtl-sdr thing, though I haven't seen that happen today. I could try putting a heatsink on the dongle??

So far today, there hasn't been any issues with drops in reception.

And the map....

[g7ruh]

Thank you. What was the weather like yesterday between you and Isle of Man? In particular low cloud / fog / drizzle / rain.? I noticed the number of vessels gradually declining during the afternoon 1300-1700 then slowly increasing. That could be a weather related thing as I see it here when fog / drizzle is about. Of course it could just be less vessels at the weekend too.

Pictures below from aiscatcher.org Station 1852

I also noticed short (30 minutes or so events like we have been discussing here) at 05:56-06:26; 09:00-09:21 and (shown above): 12:29-12:47.

After about 18:30 there is a huge increase in distance (from about 30 NM to 60 NM):

Maybe you could track those short events for a few days to see if they are always roughly the same duration (which you have mentioned before as being about 40 minutes) and if they occur about the same time of day repeatedly. That may indicate some kind of interference.

Also, but not at the same time as above, you could consider turning autogain on for a few days to see if it makes any difference to the times when the noise floor is much higher during those short events. It is somewhat counter-intuitive to my experience, but then so is the condition we are trying to track down (local radio transmissions are not, as a radio amateur, you soon learn what your own transmissions cause issues with!). We do need to track them before to determine the regularity, otherwise they might have gone away, rather than autogain fixing it!

The above thoughts were triggered by looking at the noise floor being higher (-25db) rather than lower (-35db) during the 30-40 minute times we are looking at. In my experience that is often caused by a strong signal overloading the receiver. Of course, if that is the case, using an aerial with higher gain may make things worse!

[markdj57]

The weather was fairly settled yesterday. I did notice there wasn't many boats about but more appeared in the evening. I think it was just picking up individual boats that was changing the results of the max. distance but I definitely do want to get to the bottom of the big drops in range/messages/vessels which appear at times.

When there are only a dozen or so boats getting picked up, it does make changes to reception/distance more based on their locations rather than anything with the receiver.

I have ordered a larger antenna and hope to get it up next weekend if the weather is suitable.

Just an off topic thing, I have a Synology NAS but limited knowledge on hosting a web server for direct outside access. I do use wordpress but I'm not a programmer, so any direction/pointers in helping to set up a page where I can see my aiscatcher stuff remotely would be most helpful.

[g7ruh]

Off-topic: I have a Synology NAS but no experience of web hosting on it. With web pages, security and illigal access is a risk, and needs to be considered. I use my own VPN on a 'Pi. PiVPN (just an installer and Wireguard VPN). It Just Works! It allows me to access my network via an encrypted link, so by definition, more secure. Also has the advantage that, when in a public place, you can use the VPN link encrypted back to your home and access internet stuff from there. Avoids the risk of a person monitoring your comms from the next table.

Having said all that, Synology offers this help information: https://kb.synology.com/en-uk/DSM/tutorial/How_to_host_a_website_on_Synology_NAS

I suggest we take this OT subject onto direct mail, if neeeded.

[markdj57]

Concerning? PPM shift and increase in vessel count? or is it just that there are more vessels?

Map.

[markdj57]

The good reception lasted 45 minutes, then an 8 minute "blip" of drop in reception followed by back to reasonable reception.

The rtl-sdr v4 is just warm to touch, nothing that would concern me.

[markdj57]

I received my larger antenna today and have it installed already :-)

There is a marked difference in reception of vessels. There aren't that many vessels on the water to begin with but if the numbers go from about 16 up to 28 vessels per minute and a marked increase in range, I think that is a noticeable improvement.

I'm receiving vessels in harbours that I never have before, in Dublin harbour 66NM and in the York dock 15NM in Belfast, as well as the Larne - Cairnryan ferries 33NM that were more elusive than the Belfast to Cairnryan ferries which I believe I am picking up more consistently at between 15-45NM.

Considering my visual line of sight is about 10NM, I'm definitely stretching out the reach of AIS.

With regards the PPM and gain adjustments, I've removed the entries from the bat file so ppm and gain are on auto. I did some testing through the week, and of course this may change with the new antenna, but I have seen no advantage or even a slight drop when switching to manual gain. The signal dB looks awful but it's working with no difference in messages per minute.

One think I have found out is I do like the graphs being saved after I close the program, and now I know how to close the program without losing any data. To close AIS Catcher, bring up the cmd screen that it is running in and hit CTRL+c, it will ask if you wish to stop and type y and hit enter. This way I don't lose any historical data in the graphs, even if I have the graphing saving at 5 minute intervals.

AIS-catcher -m 4 -a 192K -gr BIASTEE true -X [MYCODE] -u 127.0.0.1 12345 -N 8100 LAT 54.4111 LON -5.6651 SHARE_LOC on PLUGIN_DIR plugins FILE stat.bin BACKUP 5
PAUSE

Oh, and the antenna I went with, a Moonraker Marine 500 Base Station 2.25m VHF antenna, a bargain compared to a Shakespeare that had the same 3 x 5/8 wavelength arrays.

Good Watch

Mark

[g7ruh]

Mark,
excellent progress!

With the new aerial and auto gain on, we need to see if you still get the 30-40 minute downturns in reception. Please keep an eye out and let us know.

Try for a few weeks or so to get some idea how well the new config works then you could try using manual gain to see if you can get any better performance. Given your location and the smaller number of vessels, it would be interesting to see if there is a difference. Of course, if you are happy with it 'as is', that is fine too. I do subscribe to the school of "if it is not broken, don't fix it", but sometimes curiosity wins out :)

Looks like you may well have reached the knee of the diminishing returns curve where extra <cost | time | effort> will not give much in return!! That is a good place to be.

Roger

[markdj57]

I'm still getting occasional large drops in reception, sometimes for a short time of 7 minutes, sometimes for 40 minutes.

https://aiscatcher.org/stations/details/1852

You can see the drop by the number of vessels received per minute.

I have tried ferrite cores, I have tried using a usb extension to move the rtl-sdr dongle away from the mini PC. I have attempted to listen with the squelch turned off using a handheld radio when the drop occurs to see if there is anything being received, I have now used another PC to run the software and plugged in the rtl-sdr to it and have so far run it overnight, but there have been 2 dips in the past 24 hours that I have noticed. I have monitored the usb voltage to look for any power supply issues, I'm going to get an inline dongle to check this physically rather than relying on software but I don't think it is this. I will try with a different dongle but, going from previous experience when I had tested a month or so ago, this will not be the issue either. I have monitored the PC resources, processing, ram use etc to look for bottlenecks. I've turned off led lights, I've run the microwave, hob, oven etc to look for correlation but found none.

I am a tech guy, so I do have a lot of smart devices, power supplies, Phillips hue with ZigBee lights, smart monitoring for the heating oil tank, smart garage door openers, and a mesh WiFi 7 network, but I have not been able to correlate anything that occurs with the brief drops in reception.

So I am at a complete loss as to tracking down why or how I'm getting these random drops, last night about 3am, the previous evening when everyone was out of the house, and today again when no one was here at about 12:35pm.

I do have 3 WiFi cameras, and one went offline briefly when the reception dipped so even though 2.4Ghz is no where near 161Mhz, I am going to turn it off just to be sure that it isn't related either.

Throughout the past week, propagation has been really great with vessels up to 100nm away received, but then this morning that has been cut in half to 50nm which is still high for vhf with a horizon at 10nm.

Just my frustrated thoughts on reliability from the past week.

Good watch

Mark

[NR857]

It's a harmonic from the mesh network. For example, a good strong WiFi signal obliterates channel 28 marine. The RX (ship) side of the channel 28 duplex pair is 162.000, which is perilously close to the two AIS frequencies of 161.975 and 162.050. If you have a handheld marine set, switch to channel 28 and bring around the WiFi AP's and extenders etc. Change the polarisation of the handheld to test for spurious emissions. There might be a reflector somewhere in the house like a metal tank, which is contributing. Best of luck. Kieran NR857

…

On Sun 11 May 2025, 14:56 markdj57, ***@***.***> wrote: I'm still getting occasional large drops in reception, sometimes for a short time of 7 minutes, sometimes for 40 minutes. https://aiscatcher.org/stations/details/1852 You can see the drop by the number of vessels received per minute. I have tried ferrite cores, I have tried using a usb extension to move the rtl-sdr dongle away from the mini PC. I have attempted to listen with the squelch turned off using a handheld radio when the drop occurs to see if there is anything being received, I have now used another PC to run the software and plugged in the rtl-sdr to it and have so far run it overnight, but there have been 2 dips in the past 24 hours that I have noticed. I have monitored the usb voltage to look for any power supply issues, I'm going to get an inline dongle to check this physically rather than relying on software but I don't think it is this. I will try with a different dongle but, going from previous experience when I had tested a month or so ago, this will not be the issue either. I have monitored the PC resources, processing, ram use etc to look for bottlenecks. I've turned off led lights, I've run the microwave, hob, oven etc to look for correlation but found none. I am a tech guy, so I do have a lot of smart devices, power supplies, Phillips hue with ZigBee lights, smart monitoring for the heating oil tank, smart garage door openers, and a mesh WiFi 7 network, but I have not been able to correlate anything that occurs with the brief drops in reception. So I am at a complete loss as to tracking down why or how I'm getting these random drops, last night about 3am, the previous evening when everyone was out of the house, and today again when no one was here at about 12:35pm. I do have 3 WiFi cameras, and one went offline briefly when the reception dipped so even though 2.4Ghz is no where near 161Mhz, I am going to turn it off just to be sure that it isn't related either. Throughout the past week, propagation has been really great with vessels up to 100nm away received, but then this morning that has been cut in half to 50nm which is still high for vhf with a horizon at 10nm. Just my frustrated thoughts on reliability from the past week. Good watch Mark — Reply to this email directly, view it on GitHub <#75 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BH3OCY2P6S7HONOZGB6WRXL255JJ5AVCNFSM6AAAAAB2PYEI6KVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTGMJQG4ZTGOI> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[NR857]

Also, I may have said this before, you could try a short Yagi, say three or four elements, point it towards the sea - it will help block out anything from inside the house by adding a bit of directivity.

…

On Sun 11 May 2025, 16:51 Kieran Higgins, ***@***.***> wrote: It's a harmonic from the mesh network. For example, a good strong WiFi signal obliterates channel 28 marine. The RX (ship) side of the channel 28 duplex pair is 162.000, which is perilously close to the two AIS frequencies of 161.975 and 162.050. If you have a handheld marine set, switch to channel 28 and bring around the WiFi AP's and extenders etc. Change the polarisation of the handheld to test for spurious emissions. There might be a reflector somewhere in the house like a metal tank, which is contributing. Best of luck. Kieran NR857 On Sun 11 May 2025, 14:56 markdj57, ***@***.***> wrote: > I'm still getting occasional large drops in reception, sometimes for a > short time of 7 minutes, sometimes for 40 minutes. > > https://aiscatcher.org/stations/details/1852 > > You can see the drop by the number of vessels received per minute. > > I have tried ferrite cores, I have tried using a usb extension to move > the rtl-sdr dongle away from the mini PC. I have attempted to listen with > the squelch turned off using a handheld radio when the drop occurs to see > if there is anything being received, I have now used another PC to run the > software and plugged in the rtl-sdr to it and have so far run it overnight, > but there have been 2 dips in the past 24 hours that I have noticed. I have > monitored the usb voltage to look for any power supply issues, I'm going to > get an inline dongle to check this physically rather than relying on > software but I don't think it is this. I will try with a different dongle > but, going from previous experience when I had tested a month or so ago, > this will not be the issue either. I have monitored the PC resources, > processing, ram use etc to look for bottlenecks. I've turned off led > lights, I've run the microwave, hob, oven etc to look for correlation but > found none. > > I am a tech guy, so I do have a lot of smart devices, power supplies, > Phillips hue with ZigBee lights, smart monitoring for the heating oil tank, > smart garage door openers, and a mesh WiFi 7 network, but I have not been > able to correlate anything that occurs with the brief drops in reception. > > So I am at a complete loss as to tracking down why or how I'm getting > these random drops, last night about 3am, the previous evening when > everyone was out of the house, and today again when no one was here at > about 12:35pm. > > I do have 3 WiFi cameras, and one went offline briefly when the reception > dipped so even though 2.4Ghz is no where near 161Mhz, I am going to turn it > off just to be sure that it isn't related either. > > Throughout the past week, propagation has been really great with vessels > up to 100nm away received, but then this morning that has been cut in half > to 50nm which is still high for vhf with a horizon at 10nm. > > Just my frustrated thoughts on reliability from the past week. > > Good watch > > Mark > > — > Reply to this email directly, view it on GitHub > <#75 (comment)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/BH3OCY2P6S7HONOZGB6WRXL255JJ5AVCNFSM6AAAAAB2PYEI6KVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTGMJQG4ZTGOI> > . > You are receiving this because you were mentioned.Message ID: > ***@***.*** > com> >

[NR857]

All, I'm not too sure if any of you have seen the attached. It ay be relevant to Mark's signal loss issues

…

On Sun, 11 May 2025, 16:54 Kieran Higgins, ***@***.***> wrote: Also, I may have said this before, you could try a short Yagi, say three or four elements, point it towards the sea - it will help block out anything from inside the house by adding a bit of directivity. On Sun 11 May 2025, 16:51 Kieran Higgins, ***@***.***> wrote: > It's a harmonic from the mesh network. For example, a good strong WiFi > signal obliterates channel 28 marine. The RX (ship) side of the channel 28 > duplex pair is 162.000, which is perilously close to the two AIS > frequencies of 161.975 and 162.050. If you have a handheld marine set, > switch to channel 28 and bring around the WiFi AP's and extenders etc. > Change the polarisation of the handheld to test for spurious emissions. > There might be a reflector somewhere in the house like a metal tank, which > is contributing. > > Best of luck. > > Kieran > > NR857 > > On Sun 11 May 2025, 14:56 markdj57, ***@***.***> wrote: > >> I'm still getting occasional large drops in reception, sometimes for a >> short time of 7 minutes, sometimes for 40 minutes. >> >> https://aiscatcher.org/stations/details/1852 >> >> You can see the drop by the number of vessels received per minute. >> >> I have tried ferrite cores, I have tried using a usb extension to move >> the rtl-sdr dongle away from the mini PC. I have attempted to listen with >> the squelch turned off using a handheld radio when the drop occurs to see >> if there is anything being received, I have now used another PC to run the >> software and plugged in the rtl-sdr to it and have so far run it overnight, >> but there have been 2 dips in the past 24 hours that I have noticed. I have >> monitored the usb voltage to look for any power supply issues, I'm going to >> get an inline dongle to check this physically rather than relying on >> software but I don't think it is this. I will try with a different dongle >> but, going from previous experience when I had tested a month or so ago, >> this will not be the issue either. I have monitored the PC resources, >> processing, ram use etc to look for bottlenecks. I've turned off led >> lights, I've run the microwave, hob, oven etc to look for correlation but >> found none. >> >> I am a tech guy, so I do have a lot of smart devices, power supplies, >> Phillips hue with ZigBee lights, smart monitoring for the heating oil tank, >> smart garage door openers, and a mesh WiFi 7 network, but I have not been >> able to correlate anything that occurs with the brief drops in reception. >> >> So I am at a complete loss as to tracking down why or how I'm getting >> these random drops, last night about 3am, the previous evening when >> everyone was out of the house, and today again when no one was here at >> about 12:35pm. >> >> I do have 3 WiFi cameras, and one went offline briefly when the >> reception dipped so even though 2.4Ghz is no where near 161Mhz, I am going >> to turn it off just to be sure that it isn't related either. >> >> Throughout the past week, propagation has been really great with vessels >> up to 100nm away received, but then this morning that has been cut in half >> to 50nm which is still high for vhf with a horizon at 10nm. >> >> Just my frustrated thoughts on reliability from the past week. >> >> Good watch >> >> Mark >> >> — >> Reply to this email directly, view it on GitHub >> <#75 (comment)>, >> or unsubscribe >> <https://github.com/notifications/unsubscribe-auth/BH3OCY2P6S7HONOZGB6WRXL255JJ5AVCNFSM6AAAAAB2PYEI6KVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTGMJQG4ZTGOI> >> . >> You are receiving this because you were mentioned.Message ID: >> ***@***.*** >> com> >> >

[g7ruh]

@NR857
Hi Kieran

Sorry for delay, "Real Life" has got in the way.

Interesting thought. Worth @markdj57 discounting it. I am not sure though as the loss occurs for 10-40 minutes at a time. It does seem to to be random, as far as I can see from the web data. Switching off all local WiFi when the issue is present will indicate if it is a problem.

Yes yagi could help, but it looks like fairly local interference of some kind, so if it was from in front or to the sides (lobes) it could make it worse, but at least we would know where is located! Mark, how is your yagi construction skill? I have made a temporary yagi out of wood bar or dowel and drilled holes for stiff copper wire (thick mains cable will do, or steel fence wire), yes I know it is not too weather-proof! Plastic bag helps that :) Use an online yagi calculator to get the element sizes.

Mark,

Thank you for the update, I have see a few times on the web page. At least we know it is not the dongle or aerial as they have been swapped.

Do you still receive the base station on the Isle of Man at Snaefell summit when you have the problem? May take you a while to confirm though.

Since you have a spare aerial and a spare dongle, maybe you could set up a receiver at a different location (but within 1/2 mile or so) to test. Even the bottom of the garden would work with a laptop or Raspberry Pi.

You need to turn everything off in your house while the problem is present, all power supplies, computers, multimedia stuff: even turn off ring-main breakers (except for your AIS system!) will help discount it. Maybe it is a neighbour with something causing the issue.

The slightly better news is that it is infrequent, harder to track down, but (marginally) less annoying than if it were there all the time. I had a problem on 50MHz many years ago. Garage door opener receiver was the cause.

I have a nasty spike at 162.100 MHz but it does not cause problems with AIS with or without a 161 MHz filter.

Regards,
Roger

[markdj57]

I've been quiet for a while, not much more to report. I got the lend of a mobile scanner to look around the 162Mhz and while there are electronics in the house that seem to produce signals around that frequency, nothing has changed.

I changed the power supply of the miniPC to see if it was that

I moved my weather station and AIS over from the miniPC to the main PC and disconnected the miniPC from the power to see if that made a difference, but none, previously I had still both going one with ais and one with weather station.

I have mass turned off a lot of stuff when the reception drop occurs but nothing found yet. I did use the scanner around the wifi router but nothing found.

I've disconnected DECt phones, power supplies, smart speakers, tvs, all when the drop occurs, which is random so I'm limited as to when I see the drop and to when I am free to start pulling things out of the wall.

I haven't yet tried the laptop on the mobile vhf on the roof of the van yet, when the drop occurs, so that's something else to test.

I am sticking with auto gain, auto ppm, etc. and with some propagation I do get out to 110 nautical miles, I've received signals from rescue helicopters that were at 4000 metres right into middle England, about 250NM away. It's very interesting watching the range increase with changing weather or time of day.

I noticed today on the top of a factory that is about 1/4 mile away, a yagi and not sure what it's for, it doesn't look like it would be for a TV, that might be something to look into further.

[NR857]

Well the test will be when you get mobile. If it's completely random, it could be a system somewhere near you that's programmed to send data when a buffer is full or something like that. The Yagi on top of the factory is most likely a telemetry fixed link. You'll know by the length of the elements - the most common ones are UHF and will have at least 3 director elements, then the live dipole, then a reflector. I don't really want to know your home location, but if you've checked everything in the house and it's not consistent, then it's a high power transmitter somewhere behind you that's drowning out the very weak AIS signals. Have a look around (virtually) and see if there's anything.

…

On Sat 24 May 2025, 16:52 markdj57, ***@***.***> wrote: I've been quiet for a while, not much more to report. I got the lend of a mobile scanner to look around the 162Mhz and while there are electronics in the house that seem to produce signals around that frequency, nothing has changed. I changed the power supply of the miniPC to see if it was that I moved my weather station and AIS over from the miniPC to the main PC and disconnected the miniPC from the power to see if that made a difference, but none, previously I had still both going one with ais and one with weather station. I have mass turned off a lot of stuff when the reception drop occurs but nothing found yet. I did use the scanner around the wifi router but nothing found. I've disconnected DECt phones, power supplies, smart speakers, tvs, all when the drop occurs, which is random so I'm limited as to when I see the drop and to when I am free to start pulling things out of the wall. I haven't yet tried the laptop on the mobile vhf on the roof of the van yet, when the drop occurs, so that's something else to test. I am sticking with auto gain, auto ppm, etc. and with some propagation I do get out to 110 nautical miles, I've received signals from rescue helicopters that were at 4000 metres right into middle England, about 250NM away. It's very interesting watching the range increase with changing weather or time of day. I noticed today on the top of a factory that is about 1/4 mile away, a yagi and not sure what it's for, it doesn't look like it would be for a TV, that might be something to look into further. — Reply to this email directly, view it on GitHub <#75 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BH3OCY75PMVQETZCF2GGUI33ACIVHAVCNFSM6AAAAAB2PYEI6KVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTGMRVG43TKMA> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[g7ruh]

Mark,

thanks for the update. I have been watching the data at a distance using aiscatcher.org performance graphs (minute) to see if I can spot anything.

This is an update to say I am still 'on the case'. I notice yesterday and today several long 'outages' reception-wise.
I will provide a few thoughts in a day or so.

Roger

[g7ruh]

Mark,

I have been watching your station via aiscatcher.org and every so often check out the graphs (not constantly!).

I have noticed that over the last month or so, at the times I have checked (a few times a day / evening) that most "outages" seem to be for 10 - 30 minutes, sometimes a bit longer.

All seem to be of random duration and at random intervals. The graph below is not typical in that there are lots of events in a short time but shows the issue. Today for example there have been a few short duration events (when I have been checking).

Then, last weekend, there were a few longer duration events lasting for an hour or more.

Nothing above I suspect you do not already know.

The long duration events at the weekend suggest the cause could be residential rather than commercial.

My first stop in interference investigation is to use Airspy SDR# to look at the frequency in question. On the top window (the spectrum available for your SDR bandwidth) has a maximum hold feature. With this, any big signals will raise the floor and the yellow trace is a hold function.

The following is a Airspy Mini trace locally and you can see the peak hold for the channels in use. Peak hold is a right mouse click to toggle the function.

If there is a signal present which is de-sensitising your receiver, I would expect the yellow line to be a higher line covering the frequency in question and likely higher and lower frequencies too. In my example, the line would be (say) halfway towards the strongest peaks, drowning out the weaker signals, but still getting some signals.

I know your RTL dongle does not have the same bandwidth but it could help identify if a radio signal on the same frequency range is present.

It could be someone locally using a radio transmitter but for communications, they tend to be two-way, so the interference would be shorter duration, no interference while the other party is transmitting. There are some methods of communcation which are constant, so may still be a local transmitter. Ideal use for your spare aerial and dongle.

Another suggestion is to chat with a Radio Amateur locally, or if you don't know one, contact the secretary of a local Radio Club. You will find a broad range of RF-related skils among the members. Many will offer help to track down RF related problems for folks. They typically have a wide range of kit to deploy and skills to direction find, as well as a lot of knowledge.

We also have a lot of resource to tap (internet, of course) where some web sites are dedicated to digital signals and identifying them based on signal sound and shape.

If all else fails, contact the Radio Society of Great Britain and they can help put you in touch with someone.

When you feel that you have investigated locally for things even in nearby houses (with the resident being willing to help you, of course) and have run out of ideas, the above is another way to investigate. Many issues have been found locally and most often the person who is causing the issue does not know that they are. If it is e.g. a power supply and you offer to replace it, they will accept the offer.

I hope that helps give you a few ideas. If you want to discuss in detail, I suggest we take this into private email.

Roger

[markdj57]

Thank you for the suggestions. Unfortunately I have not been able to spend the time to track down the source of the interference due to work and other commitments where I have not been home to do a further search. I did bring up some interesting info here on radio licenses in the area.

https://radioreferenceuk.co.uk/frequencies.php?freq=156-180&town=Downpatrick&range=10

I will try to make contact with a radio amateur group and see if anyone is in the area.

[NR857]

All. T NI Fishery frequency is Channel 14, Maritime Mobile service. The Phennick Cove frequency 157.025/161,625 is Channel 80, Maritime Mobile Service, (UK marina channel) Phennick Cove is a local development vehicle for Ardglass Marina. Neither of these is likely to cause that kind of interference unless they have rigged an illegal repeater. KFH

…

On Fri, 6 Jun 2025 at 14:46, markdj57 ***@***.***> wrote: Thank you for the suggestions. Unfortunately I have not been able to spend the time to track down the source of the interference due to work and other commitments where I have not been home to do a further search. I did bring up some interesting info here on radio licenses in the area. https://radioreferenceuk.co.uk/frequencies.php?freq=156-180&town=Downpatrick&range=10 I will try to make contact with a radio amateur group and see if anyone is in the area. — Reply to this email directly, view it on GitHub <#75 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BH3OCYYCNPZAXQGBI6KH4G33CGLTVAVCNFSM6AAAAAB2PYEI6KVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTGMZZGA4DAMI> . You are receiving this because you were mentioned.Message ID: ***@***.***>