Is this interference i see at 137MHz?

Get help identifying an unknown signal.
LAH
Posts: 13
Joined: Tue May 26, 2020 5:27 pm

Re: Is this interference i see at 137MHz?

Post by LAH » Thu Jun 18, 2020 7:36 pm

I wish the only source of noise was the desktop computer that would be easier to solve. But i doubt EMI from the computer is the issue in my case since i managed to replicate the interference running SDRTouch on an android device. This also means that changing the active USB cable will also have little to no effect to fix anyhing in my case. Unless the EMI is due to the routing. If not, it is deffinitely something the antenna is picking up. My plan is to first sort out my setup to isolate the source.

I’ve done some testing today with the new 50cm passive patch cable connected with the actives. No antenna connected, max gain, active USB cables are rolled and at my desk in close proximity to the compture. I wanted to see what is actually coming from the USB cable side alone.

First i checked 480 Mhz to see how each cable transmits the USB noise. The patch cable alone and the old 10m active USB cable with one small ferrite at the computer end showed similar results. Only the acitve cable seems to amplify the noise a little. The 15m cable on the otherhand shows more of a flooding in that frequency and no spikes. Could this be the 2 large ferrites at each end?
Looking at the frequency of interest, 137Mhz, the 15m cable showed no spikes while the 10m still had the spikes i was familiar with. No drifting interference, which means it should be coming from the antenna (or the coax). When i lay out the USB cable to its new route i will run the dongle with no antenna and check again to confirm.

Then I did some continuity checks and found out that on the 50cm patch cable the shield (metal casing of the usb connectors) and USB ground (one of the 4 wires inside the cable) are not connected. On the active USB cables and the dongle itself the ground and the shield was connected.
I also found out that the shield on the 15m cable was an open circuit but the 10m had continuity at it’s shield.

This means if i will have to disconnect the shield from the ground i will have to do it within the dongle itself.

The chimney i’m using as elevation has a flat top about 50x40cm, i don’t have much wiggle room as you can see but i have a couple of ideas/positions to try. If everything fails, I also have some TV coax laying around, i just need to check the length and get some SMA to F connectors to try out. As you said you will never know if you never try :)

It finally rained today(!) and the weekend looks wide open :P , hopefully i will have some time and give it a go.
Attachments
new 480Mhz_.png
480Mhz 15m active + 50cm patch at the dongle end.
new 480Mhz_.png (175.13 KiB) Viewed 1192 times
old 480mhz_.png
480Mhz 10m active + 50cm patch at the dongle end.
old 480mhz_.png (183.08 KiB) Viewed 1192 times
old 137mhz_.png
137Mhz 10m active + 50cm patch at the dongle end.
old 137mhz_.png (137.82 KiB) Viewed 1192 times

JEL
Posts: 177
Joined: Wed Feb 07, 2018 2:08 am

Re: Is this interference i see at 137MHz?

Post by JEL » Fri Jun 19, 2020 3:17 am

Just a few quick links that touches this subject (Maybe you have already seen these, but I post them just in case you or others might find them useful. If not for anything else, then perhaps simply to compare notes :) ) :

https://sdr-radio.groups.io/g/main/topi ... 0,33030109

https://www.reddit.com/r/RTLSDR/comment ... _and_your/

http://www.radioforeveryone.com/p/reduc ... noise.html

https://github.com/mossmann/hackrf/issues/544

LAH
Posts: 13
Joined: Tue May 26, 2020 5:27 pm

Re: Is this interference i see at 137MHz?

Post by LAH » Sat Jun 20, 2020 6:49 pm

Thanks fo the links, great read! I've seen couple of them before but the rest i didn't.
I followed a somewhat similar method to the article on radioforeveryone.com.
I layed out the cable to the new route and tried different configurations with and without the antenna.

I started by disconnecting the shield, i did not remove the metal case as sugested but wrapped the male USB tip with isolation tape effectively preventing coupling. Since i already confirmed the ground and shield was connected internally on both the dongle and the cable i was expectin no changes with this mod. Which is exactly what i got.

Then introduced the 50cm patch cable which gave mixed results. Without the antenna i got 1.5dB improved noise floor but with the antenna attached i got 11dB added to the noise floor which was quite shocking. I then wrapped aluminum tape to the connection points making sure they are connected to the shielding with no significant effect. Connected the negative terminal of a 12v battery, got 2 to 3 dB improvement. This was dissapointing since i was hoping the battery would give the best improvement as this was the case in the article. As a last effort i covered the patch cable completely with aluminum tape leaving the battery negative connected, grounded from dongle side, got 2dB additional noise, grounded from USB side got 2dB again. Grounded from both sides and go the same result as with no patch cable or battery.

So i decided to ditch the battery and the patch cable and go directly to the active USB. I noticed that the addition of the patch cable added 11dB to the noise floor but the erratic EMI was not visible. Removing the patch improved noisefloor instantly but now i could see the erratic EMI. I assume the increased noise floor prevented the EMI to get through completely. I'm not sure which one is worse though.

My final configuration for today was PC to 15m active USB to dongle to antenna. Since the new cable is longer i was able to remove the 3m kit coax from the line. I did not see any change at the frequency of interest with the dongle being closer or further from the antenna so i will keep it this way for now.

Coming to my initial problem, with the results of the no antenna tests, i concluded that the erratic EMI is something the antenna picks up. I am also pretty confident it is not originated from my residence. I guess from now on i will see what i can do with the LNA, FM trap and the band pass filter. I'm expecting the filter to arrive early this week but the LNA and the FM trap will take more time.

Meanwhile I will be experimenting with the antenna position starting from the extremes and also the cable route. I noticed I can improve the noise floor by simply pulling the USB cable away from the chimney at a certain angle (radio magic?). At one point I might try to route through the chimney just to see if it changes anything and also because i have the possibility to do so :D apart from the EMI i think my noise floor with this cable is not very good either. Today, with max gain, I got -23dB at worst with the patch cable and -42dB at best without the patch cable.

JEL
Posts: 177
Joined: Wed Feb 07, 2018 2:08 am

Re: Is this interference i see at 137MHz?

Post by JEL » Sun Jun 21, 2020 12:11 pm

I was just about to say "Welcome to the hobby" :lol:

Maybe somebody else with more knowledge and experience than myself can chip in, because I'm not really sure what you are up against.

I want to show you one thing though, which I battled for a while some years ago: A neighbors guitar-amp :shock:

https://www.youtube.com/watch?v=BFZ1Vf5qy7o

Whenever he turned it ON and started his riffs... the entire airband completely died for me.

Now, as you can see in the video his guitar-playing can actually be heard, so that's probably not what you're facing, but the carrier-spikes you face could perhaps be caused by something similar to the guitar-amp; ie a disturbance from some device out-of-house (As you have suspected yourself)

Not that this helps you identify your problem-source in any way (It's obviously much easier to guess what's going on when you can actually hear a guitar, rather than just facing 'faceless' carriers polluting all the frequencies), but perhaps you really do need to look around for some device that could be the 'culprit' here.

Maybe see if you can track a pattern to them of some kind? Are they always ON, for example, do they shift power-strength at certain times of the day/night, do they coincide with some neighbor-activity, or any pattern you may be able to discern from them that may help you pinpoint what is causing them.

This most likely won't be easy though.
And even if you do find a source, it may not be one you can 'control' (A guitar-amp is fairly easy, but if it's something essential to your neighbor's house, for example, something related to light, heating, home-automation, some TV-box, or something along those lines, then you could face an interference that is there to stay.)

But as said; at this point you need input from somebody with more knowledge/experience on this kind of interference/situation than myself.
(I also don't know anything about chimneys, even though I once sat in Santa's chair at the local mall :lol: )

If nobody else chips in, then I do hope you will keep us posted on your progress though. Could be a learning-experience for all :)

LAH
Posts: 13
Joined: Tue May 26, 2020 5:27 pm

Re: Is this interference i see at 137MHz?

Post by LAH » Sun Jun 21, 2020 4:13 pm

Thanks a lot for all the input and the warm welcome :D
There are so many variables in a radio receiver setup, i understand that it is next to impossible for anyone to give any precision advise unless the noise is a common occurrence and could be recognized immediately from the waterfall images.
JEL wrote:
Sun Jun 21, 2020 12:11 pm
I want to show you one thing though, which I battled for a while some years ago: A neighbors guitar-amp
I remember getting radio through my guitar amp but seems like it works both ways :lol: Poorly shielded pickups can receive radio beautifully as well as mains hum. Speaking from experience, a good shielding can work wonders but then again radio is a completely different beast in itself. I bet what your neighbour had was a cheap wireless system from aliexpress which was never tested for EMI in the first place.

I wonder if a tube amp would emmit much more than a transistor amp both through the air and the mains. You know high voltages, analog rectifiers and lots of transformers. Speaking of analog circuitry i believe what i have in my waterfall is power circuit related, maybe a voltage regulator? I would love to try with the SDRTouch if we ever have a blackout here. I know it is there as long as there is sunlight, i will keep the antenna up tonight and confirm if its "on" at night as well.

I agree, this signal is here to stay. I just have to learn to live with it and focus on figuring out workarounds. I think it was always there but my previous antenna location was somewhat luckily shielded from noise since it had a rather limited view of the sky. Now that it's up there in the open it can pick up much more.

I know, my methods are not exactly controlled experiments so far, i feel like i'm running around hopelessly trying anthing that comes to mind. But with radio, it's hard to predict what will or will not work. I will definitely keep posting as i go further.

LAH
Posts: 13
Joined: Tue May 26, 2020 5:27 pm

Re: Is this interference i see at 137MHz?

Post by LAH » Mon Jun 22, 2020 7:27 am

Minor update,: during routing of the new USB cable the other day, everything went suspiciously easy, no cable streches even though according to my calculations it should have been a "tight fit" so to speak. I had to bring the antenna down due to more rain recently and had the opportunity to measure the lenght. Found out it was 20m, well not exactly 20 but more like 19.5, i can't complain given the price point though. I purchased it as 15m, the box it came in reads 15m and i didn't bother to check at the time.

For anyone wondering how long and cheap they can go with the active cable, i can confirm the cheapest no-brand generic cable i could find with no ferrites and a flimsy looking gauge did work with the RTL-SDR blog dongle flawlessly. No drops what so ever.

Obviously i used it only with the RTL-SDR blog dongle and no LNA (yet) so can't speak about how it would handle the increased current draw via bias tee or another dongle with more current draw. I need to get one of those plug in USB tester meter things to see how much voltage drop is going on.

The two large ferrites at each end was added by me, the cable has no built in ferrite filters. Also you can see my shielding attempt at the female end. The first active box is at 10m mark and the aluminum tape on the sides of it is only to prevent moisture ingress through the shell joint at the sides.
Attachments
USB 20m.jpg
USB 20m.jpg (262.22 KiB) Viewed 1041 times

JEL
Posts: 177
Joined: Wed Feb 07, 2018 2:08 am

Re: Is this interference i see at 137MHz?

Post by JEL » Tue Jun 23, 2020 1:05 pm

You mentioned you're getting a band-pass filter.
Which frequency-range is that letting through?
This will probably be your best bet at removing those unwanted spikes.

You also mentioned the broadcast-FM (BCFM) stop-filter, which I find is brilliant (If it's RTL-SDR's version you're getting, which is the type I use)

At the very beginning of my own SDR-use I was being drowned by BCFM stations that mirrored all over the airband (Which is my primary focus)
Attaching the BCFM stop-filter made that problem go away :)

I also have the broadcast AM (BCAM) High-pass filter from RTL-SDR, which I also find is great.

I have those 2 permanently attached to my VHF/UHF-receiver :)

They generally clean things up a lot for me, so I never detach those (They didn't stop the guitar-amp though, so may not be the 'magic-bullet' for the noise you're battling either though. But if you're going for sat-signals around 137 MHz I think you would like both of them, but especially the BCFM filter which you are getting :) )

Your extension-cable is twice as long as the one I tested.
I'm surprised 20 meters work, but if it works it works :)

LAH
Posts: 13
Joined: Tue May 26, 2020 5:27 pm

Re: Is this interference i see at 137MHz?

Post by LAH » Tue Jun 23, 2020 7:18 pm

The bandpass filter is for satellite APT and LRPT reception, claims to have 137Mhz center and pass the 130Mhz-150Mhz band.
In fact i received it today (actually had to go pick it up myself since the postman failed to deliver yesterday) I even made a quick test run. I had the Meteor pass at low elevation and right after that NOAA 15 with a fairly high elevation pass. NOAA 15 also happens to have the strongest signals for me among the 3 APT broadcasting NOAA satellites which was a good benchmark. Well the bad news is i got no signal from the Meteor (i wasn't expecting a lock but thought i should at least see the signal in the waterfall which did not happen) and NOAA failed to reach SNR 22dB which is the minimum for a clean picture or a picture at all maybe. My peak SNR was around 20dB for about 10 degrees or so when the satellite was at peak elevation. I would expect around 30dB peak SNR at this elevation without the filter. The signal was there down to 7 degrees elevation though rather weak. I usually get it down to around 3 degrees without the filter.

I really don't know what to make of the results at this point. I have no means to test it on a bench with a test signal and see how well it actually performs. But one thing i know for sure is that it will not work without an LNA for me.

First thing i did was to check the BCFM band. I think it did a nice job supressing. Attached picture shows how it looks like at max gain. This should have been off the charts without the filter. To me it looks like a good 40dB attenuation at least. Also attached are the pictures of the block and the insides.

I removed the cover to inspect the pcb after i tested and now i'm not sure if i put it together the right way :lol: I might have reversed the in and out. The markings could have worn out at some point in the future anyway. Is there an easy way to tell the input from the output? If not i will simply plug it in either way an see what changes.

Apart from this little mishap my noise floor is now at -52dB at max gain with no signs of any interference at the frequency of interest. I used to have the floor at -40's without the filter, 12dB loss is a lot but i don't want to come to conclusions without trying it with an LNA first. It is a joy to look at such a buttery smooth noise floor though :D but its obviously no use. And some of the interference might come through once i put the LNA in line.
JEL wrote:
Tue Jun 23, 2020 1:05 pm
I have those 2 permanently attached to my VHF/UHF-receiver

They generally clean things up a lot for me, so I never detach those
Good to hear the traps are working nicely for you! Yes both the BCFM trap and the LNA is the RTL-SDR's version. I quite like their design philosophy and they tend to not break the bank so it was an easy decision. I'm not sure if i need the AM filter though. Maybe i will see after i get the LNA working. Speaking of which, looks like it will take while untill i receive the LNA and the BCFM trap. I made a strategical mistake planning which address to have them shipped to and now i have to wait more untill they're forwarded to me.

Airband is fun! I used to listen to VHF communications at home with the SDR in an effort to get used to radio chatter to help with my PPL training. A tuned dipole is such a great simple antenna. Then experimented with some ADS-B, all of which was back when the airport nearby was active. Now they moved most operation elsewhere and downsizing the airport. Such a pity.
JEL wrote:
Tue Jun 23, 2020 1:05 pm
Your extension-cable is twice as long as the one I tested.
I'm surprised 20 meters work, but if it works it works
It was surprising indeed :D I might get a better quality 15m one just to see how it compares with the rest. I also don't like the idea of extending 5 extra meters more than i need. But eventually i might ditch the cable setup completely and go for a spyserver. The only downside is that i need a stronger raspberry pi than the zero i have running pihole for me at the moment. The version 4 in kit form costs around 55 USD. Which is not little, i might as well invest it towards a better SDR dongle like the airspy mini I've been eyeing for a while now. Anyhow, i still have a lot to play with for now.
Attachments
FM broadcast through bandpass filter_.png
FM broadcast through bandpass filter_.png (433.95 KiB) Viewed 932 times
Bandpass inside_.jpg
Bandpass inside_.jpg (137.64 KiB) Viewed 932 times
Bandpass_.jpg
Bandpass_.jpg (180.84 KiB) Viewed 932 times

JEL
Posts: 177
Joined: Wed Feb 07, 2018 2:08 am

Re: Is this interference i see at 137MHz?

Post by JEL » Thu Jun 25, 2020 1:52 pm

The minimum noise-floor you can expect with the RTLSDR is probably not much lower than around 50 or 60 dB, so it does indeed sound like an LNA is necessary for your needs.

I saw a video on youtube where somebody tests the BCFM and BCAM filter and concluded they lower the signal by about 50 dB, so seems to match your findings.
(Here's the video if you want to see it for yourself: https://www.youtube.com/watch?v=I2yNLNXd1io)

I have not done any PCB-work since I was a young man ready to conquer the world and have become a happy illiterate on that issue at this stage of my life :oops:

But the BCAM and BCFM filter are supposedly palindromes, so can apparently be injected either way. I would imagine, but that is just a pure unsubstantiated guess, that the 137 MHz sat-filter will have a similar design.

Either way, the BCAM and BCFM filter are shown on the shop-website has having 4 dots at the bottom-center (And 2 at the top-sides) of the PCB when you have the input (The antenna side) on the left and the output (The radio side) on the right.

I do not know if this is something to go by, but at least those 2 filters have that design (Kind of looks like a bowl or a dip or the letter U) so I would not be surprised if the 137 MHz filter is similar (I do not have that filter myself though, so again; I am just guessing here. I have also not experimented much with sat-reception yet (We need a day to be more than 24 hours long to get through everything :lol:))
You can see the image of one of those filters here: Image
Or the entire page here: https://www.rtl-sdr.com/buy-rtl-sdr-dvb-t-dongles/
Also, you will notice the printed text on the PCB is readable when you have flow going from left to right (The 414077A... text)

So my guess is the input is where it reads 414... etc and output is where it reads ...106 (But I could be completely wrong here, or it may work the same regardless of direction) :?

I do not know how to tag RTLSDR in a post so they may read this and perhaps be able to give you the correct answer.

LAH
Posts: 13
Joined: Tue May 26, 2020 5:27 pm

Re: Is this interference i see at 137MHz?

Post by LAH » Sat Jun 27, 2020 5:05 pm

-50dB floor is fair enough but the band pass section loss is around 12dB which i think is too much. It's either a poor design or this is how band pass filters work in general. I cannot compare since this is my first ever band pass filter experience. The insertion losses are noted as non existent for the RTL-SDR blog BCFM trap. It would be nice if someone with the equipment can test this filter also. BTW i love how the oldschool hardcore equipment has those dot matrix screens on the video :D
JEL wrote:
Thu Jun 25, 2020 1:52 pm
But the BCAM and BCFM filter are supposedly palindromes, so can apparently be injected either way.
This seems to be confirmed with the layout of the BCFM trap, it looks symetrical on the PCB. Judging from the traces, the signal seems to first pass through a capacitor then an inductor and on to ground. Of course i don't know the values, they might not be symmetrical but it might not even matter.

Since the blue tinted SMD's are marked as inductors on the RTL-SDR blogs trap i assume the same to be true for my bandpass filter. However looking at the traces after the weird V shaped SMD bridge, one connector tip goes through an inductor first and the opposite tip goes through a capacitor first which makes me think it is directional. If it does not immediately reveal itself when connected to the antenna i might ask the supplier about it. Though very few suppliers in aliexpress actually know something about what they sell but it's worth a shot.

I haven't noticed how similar the boards were on both units untill you pointed out the dots. I think your theory is totally plausible. This way the orientation of writings on the shell and the PCB also matches. I think this will be the first orientation i will go with.
JEL wrote:
Thu Jun 25, 2020 1:52 pm
We need a day to be more than 24 hours long to get through everything
Well said :D in fact this pandemic is the reason i had the time to actually play with SDR again. Otherwise it's been collecting dust with a lot of other stuff in the drawer for years. I completed so many postponed projects in the last couple of months, it made me feel young again :lol: Anyhow things are getting better over here and next month could be back to work close to full time already. I highly recommend you look into satellite signals. It is very rewarding and if you're in to space in general it's double the fun :) as a bonus i managed to expand my knowledge about both radio and satellites/orbits along the way.

Post Reply