What is this stuff
- Thread starter Waylorn
- Start date
- Status
I cannot tell how much band space you are looking at, the spectrum.
Those could be birdies, or stray clocks/oscillators you are picking up.
Sent via Tapatalk
Those could be birdies, or stray clocks/oscillators you are picking up.
Sent via Tapatalk
USB noise I would guess.
kermit_t_f
Member
- Joined
- Jul 13, 2012
- Messages
- 29
It would be helpful, if you show us the waterfall display, to take a look on the variation of the signals. You could upload a recording of the raw data, too.
Attachments
br0adband
Member
Long post here so be warned... 
The one thing you need to learn as a new SDR user, at least as I've gathered that status from your posts lately, is that not everything you see on the spectrum, the waterfall, or even what you might actually hear when tuning in various frequencies is actually some kind of a radio broadcast. SDR hardware like the ubiquitous RTL-SDR ones (powered by the R820T tuner and the newer R820T2 along with the RTL2832U controller chip) as well as even better hardware like Airspy, BladeRF, and hackRF all suffer from the traditional things like spurious signals (generated by the actual electronics of the devices themselves even if you're not using a transmitter) aka "birdies" which are ghost frequencies, basically. They'll show up as signals you can receive (peaks on the spectrum) but when you tune them directly what you get is just a tone and nothing more, hence the name "birdies" as in a bird stuck whistling one audible frequency or tone.
Looking at the images you posted that would be my guess considering the bandwidth we're staring at - if you ARE using a traditional RTL stick (R820T or R820T2 powered) then here's my suggestions:
- decrease the working bandwidth to 2.56, 2.4, or even 2.08 MHz (using the SDR# config options - the gear icon). The reason for this is because it's a fairly well known issue with the RTL-based hardware that pushing it to provide the full theoretical bandwidth of 3.2 MHz causes problems like increased spurious emissions (the birdies you're most likely seeing) as well as a lot of IF images (duplicates of a given signal that show up more than once on the spectrum - you can easily tell if this is happening by shifting the receiving frequency - if the peaks on the spectrum appear on both sides of the center frequency, symmetrically speaking, then you're having image reception which can be alleviated by two methods: either back down the RF gain to something more useful like 30-36 dB instead of keeping it maxed out at 48+ which a lot of people do at the beginning of their SDR journey not realizing that the internal gain on the RTL devices ain't so great which is actually a bad thing, or decrease the bandwidth as I just noted (2.56 is about the highest you'll get without problems, I use 2.4 MHz myself and get no issues, each configuration is different so you just have to work at finding your most optimal setup).
When pushing an R820T-based stick at the max of 3.2 MHz (again, your second image shows a bandwidth of about 3.2 MHz from approximately 133.300 MHz to about 136.500 or so) it's really putting a severe strain on the hardware and RTL sticks pushed that hard have a known issue of dropping samples and images being created - so bring that bandwidth down to the 2.56, 2.4, or 2.08 MHz window and it'll work better, seriously.
Here's a link to some decent info about birdies with respect to radio scanning/monitoring (even if you're not using a "police scanner" all-in-one type device):
Radio Scanner Birdies-Interference by Design - Scanner Heaven
- alter the step size because frequencies in the range you're exploring there (around 133 MHz) use a step of about 25 kHz (in the US, sometimes you may find transmissions using a 12.5 kHz stepping as well) so that means they're always going to be on a given "step" like 118.000 then the next is 118.250 and so on - newer requirements are actually 8.33 MHz so even more narrow iirc. If you look at typical airport frequencies in use you'll note a familiar pattern of stepping which can also simply be referred to as spacing so the frequency usage falls onto given nicely rounded ones: 119.400, 125.100, 122.950, and so on - they're all falling on frequencies that are easily divisible by .025 MHz aka 25 kHz stepping. Given frequency bands use a typical stepping size - a modern high end scanner like the Uniden Bearcat 436HP has a lot of potential capabilities regarding stepping sizes and you can get an idea of just what the given frequency bands use for stepping at this link:
Easier to Read BCD436/536HP Digital Scanner Manual - default band stepping sizes
Since a typical RTL-based SDR stick has pretty much the same working range of reception (25 MHz up to about 1.7 GHz) that table should give you a great idea on what stepping size to use in a specific frequency band to ensure you'll receive something (as long as the people transmitting are following the rules, that is).
I am not telling you that you can't use 10 Hz as your stepping size anytime you want - even always if you desire - but what I am telling you is that if you're looking for transmissions of most any legal kind (the vast majority of what's out there, of course) then following the stepping sizes used in the given band is a great way to know that yep, you're gonna pick something up on a given frequency as long as your PPM is adjusted correctly after your RTL stick has warmed up - unless you happen to own one of the more recent TXCO (Temperature-Compensated Crystal Oscillator) enabled devices which means regardless of how warm/hot the device gets, the frequency it's tuned to doesn't "drift" which is when it slides off frequency a bit (PPM = parts per million in terms of Hertz) and it doesn't quite tune dead on. I have two RTL sticks myself, one has a PPM of 56, the other is 61 so they don't tune solidly until after they've been "on" for about 15-20 mins most of the time but after that they're pretty solid.
In the second image you provided you're trying to a) use the max theoretical bandwidth of 3.2 MHz so it's putting a strain on that SDR stick you're using which can create the images I mentioned, it can also make the RTL stick run much warmer/hotter which causes more PPM drift and other issues too (again, if you have a TXCO enabled device, ignore the temp warnings because they won't matter) and b) using a stepping size - 10 Hz - which is so small you're rarely going to be directly on-frequency so try and stick with the given steppings used in given frequency bands because it makes tuning that much easier. Yes, we all know how cool it is these days to "see" the spectrum and spot a peak above the noise floor and just click on it but unless you're using the given stepping sizes you're going to have to do more manual fine tuning than is really necessary. Anything above 25 MHz will typically always be "on frequency" with respect to the stepping sizes, anything below that will get into the 1 kHz area with respect to tuning, especially in the shortwave and HF bands (4891 kHz, 7777 kHz, etc).
Having said all that, my first impression is still the same: the bandwidth is too high (drop it back down to 2.56, 2.4, or 2.08 MHz - yes you see less of the spectrum but the RTL sticks work better at these settings overall) and pay more attention to the stepping size depending on the frequency band you're listening in on.
As far as noise that's not being caused by the RTL stick itself, well, modern electronics - especially computers - generate a lot of EMI and RFI so, shielding your RTL stick in some manner (even wrapping it in tin foil can help as long as you allow for a few holes for airflow). There are just too many things to offer in terms of advice and again, every configuration of hardware and every physical setup is unique and has its own particular idiosyncrasies that you'll have to work on to get the most optimal performance from your monitoring hardware.
With respect to noise and getting rid of as much as possible, I discovered this page long ago and made use of many of the tips - the most notable one being the one about using a USB extension cable (I have two, one for each RTL stick) and pulling off the external shielding on the RTL stick end (where the stick is plugged in) - that one thing alone dropped my noise floor nearly 9 dBi which is rather significant so, at the bare minimum look at the page and try to adopt some of the suggestions. RF chokes are a must these days, long USB cables, mounted the actual RTL sticks away from the computer(s) as far as possible, shielding the sticks (with aluminum foil or even metal enclosures making sure you ground the shielding to the ground on the USB stick), etc etc. Too many suggestions for a post here so check this link out:
Software Defined Radio for Mariners: Reducing electrical noise
That guy's site is loaded with info and he put out an eBook about "RTL on the cheap" too which is recommended.
Hope some info here proves useful...
The one thing you need to learn as a new SDR user, at least as I've gathered that status from your posts lately, is that not everything you see on the spectrum, the waterfall, or even what you might actually hear when tuning in various frequencies is actually some kind of a radio broadcast. SDR hardware like the ubiquitous RTL-SDR ones (powered by the R820T tuner and the newer R820T2 along with the RTL2832U controller chip) as well as even better hardware like Airspy, BladeRF, and hackRF all suffer from the traditional things like spurious signals (generated by the actual electronics of the devices themselves even if you're not using a transmitter) aka "birdies" which are ghost frequencies, basically. They'll show up as signals you can receive (peaks on the spectrum) but when you tune them directly what you get is just a tone and nothing more, hence the name "birdies" as in a bird stuck whistling one audible frequency or tone.
Looking at the images you posted that would be my guess considering the bandwidth we're staring at - if you ARE using a traditional RTL stick (R820T or R820T2 powered) then here's my suggestions:
- decrease the working bandwidth to 2.56, 2.4, or even 2.08 MHz (using the SDR# config options - the gear icon). The reason for this is because it's a fairly well known issue with the RTL-based hardware that pushing it to provide the full theoretical bandwidth of 3.2 MHz causes problems like increased spurious emissions (the birdies you're most likely seeing) as well as a lot of IF images (duplicates of a given signal that show up more than once on the spectrum - you can easily tell if this is happening by shifting the receiving frequency - if the peaks on the spectrum appear on both sides of the center frequency, symmetrically speaking, then you're having image reception which can be alleviated by two methods: either back down the RF gain to something more useful like 30-36 dB instead of keeping it maxed out at 48+ which a lot of people do at the beginning of their SDR journey not realizing that the internal gain on the RTL devices ain't so great which is actually a bad thing, or decrease the bandwidth as I just noted (2.56 is about the highest you'll get without problems, I use 2.4 MHz myself and get no issues, each configuration is different so you just have to work at finding your most optimal setup).
When pushing an R820T-based stick at the max of 3.2 MHz (again, your second image shows a bandwidth of about 3.2 MHz from approximately 133.300 MHz to about 136.500 or so) it's really putting a severe strain on the hardware and RTL sticks pushed that hard have a known issue of dropping samples and images being created - so bring that bandwidth down to the 2.56, 2.4, or 2.08 MHz window and it'll work better, seriously.
Here's a link to some decent info about birdies with respect to radio scanning/monitoring (even if you're not using a "police scanner" all-in-one type device):
Radio Scanner Birdies-Interference by Design - Scanner Heaven
- alter the step size because frequencies in the range you're exploring there (around 133 MHz) use a step of about 25 kHz (in the US, sometimes you may find transmissions using a 12.5 kHz stepping as well) so that means they're always going to be on a given "step" like 118.000 then the next is 118.250 and so on - newer requirements are actually 8.33 MHz so even more narrow iirc. If you look at typical airport frequencies in use you'll note a familiar pattern of stepping which can also simply be referred to as spacing so the frequency usage falls onto given nicely rounded ones: 119.400, 125.100, 122.950, and so on - they're all falling on frequencies that are easily divisible by .025 MHz aka 25 kHz stepping. Given frequency bands use a typical stepping size - a modern high end scanner like the Uniden Bearcat 436HP has a lot of potential capabilities regarding stepping sizes and you can get an idea of just what the given frequency bands use for stepping at this link:
Easier to Read BCD436/536HP Digital Scanner Manual - default band stepping sizes
Since a typical RTL-based SDR stick has pretty much the same working range of reception (25 MHz up to about 1.7 GHz) that table should give you a great idea on what stepping size to use in a specific frequency band to ensure you'll receive something (as long as the people transmitting are following the rules, that is).
I am not telling you that you can't use 10 Hz as your stepping size anytime you want - even always if you desire - but what I am telling you is that if you're looking for transmissions of most any legal kind (the vast majority of what's out there, of course) then following the stepping sizes used in the given band is a great way to know that yep, you're gonna pick something up on a given frequency as long as your PPM is adjusted correctly after your RTL stick has warmed up - unless you happen to own one of the more recent TXCO (Temperature-Compensated Crystal Oscillator) enabled devices which means regardless of how warm/hot the device gets, the frequency it's tuned to doesn't "drift" which is when it slides off frequency a bit (PPM = parts per million in terms of Hertz) and it doesn't quite tune dead on. I have two RTL sticks myself, one has a PPM of 56, the other is 61 so they don't tune solidly until after they've been "on" for about 15-20 mins most of the time but after that they're pretty solid.
In the second image you provided you're trying to a) use the max theoretical bandwidth of 3.2 MHz so it's putting a strain on that SDR stick you're using which can create the images I mentioned, it can also make the RTL stick run much warmer/hotter which causes more PPM drift and other issues too (again, if you have a TXCO enabled device, ignore the temp warnings because they won't matter) and b) using a stepping size - 10 Hz - which is so small you're rarely going to be directly on-frequency so try and stick with the given steppings used in given frequency bands because it makes tuning that much easier. Yes, we all know how cool it is these days to "see" the spectrum and spot a peak above the noise floor and just click on it but unless you're using the given stepping sizes you're going to have to do more manual fine tuning than is really necessary. Anything above 25 MHz will typically always be "on frequency" with respect to the stepping sizes, anything below that will get into the 1 kHz area with respect to tuning, especially in the shortwave and HF bands (4891 kHz, 7777 kHz, etc).
Having said all that, my first impression is still the same: the bandwidth is too high (drop it back down to 2.56, 2.4, or 2.08 MHz - yes you see less of the spectrum but the RTL sticks work better at these settings overall) and pay more attention to the stepping size depending on the frequency band you're listening in on.
As far as noise that's not being caused by the RTL stick itself, well, modern electronics - especially computers - generate a lot of EMI and RFI so, shielding your RTL stick in some manner (even wrapping it in tin foil can help as long as you allow for a few holes for airflow). There are just too many things to offer in terms of advice and again, every configuration of hardware and every physical setup is unique and has its own particular idiosyncrasies that you'll have to work on to get the most optimal performance from your monitoring hardware.
With respect to noise and getting rid of as much as possible, I discovered this page long ago and made use of many of the tips - the most notable one being the one about using a USB extension cable (I have two, one for each RTL stick) and pulling off the external shielding on the RTL stick end (where the stick is plugged in) - that one thing alone dropped my noise floor nearly 9 dBi which is rather significant so, at the bare minimum look at the page and try to adopt some of the suggestions. RF chokes are a must these days, long USB cables, mounted the actual RTL sticks away from the computer(s) as far as possible, shielding the sticks (with aluminum foil or even metal enclosures making sure you ground the shielding to the ground on the USB stick), etc etc. Too many suggestions for a post here so check this link out:
Software Defined Radio for Mariners: Reducing electrical noise
That guy's site is loaded with info and he put out an eBook about "RTL on the cheap" too which is recommended.
Hope some info here proves useful...
Last edited:
And this is why I am starting to really like these forums. I will have to reread this post when I get home and digest it a bit before I can make a coherent post. Thank you for taking the time to school a noob. It is appreciated.
FYI: The dongle I am using is the Nooelec R820T2
Here is a sample taken this morning and you can see there is much less "birdies". This leads me to think they are actual trasmissions.
FYI: The dongle I am using is the Nooelec R820T2
Here is a sample taken this morning and you can see there is much less "birdies". This leads me to think they are actual trasmissions.
Attachments
Last edited:
br0adband
Member
Welcome to the wonderful world of RFI, sunspots, solar cycles, neighbors turning on their microwaves, and a whole slew of potential candidates for wreaking havoc with your radio frequency monitoring hobby.
What isn't there now could come back tomorrow, or later today, or whenever and it can even come back with higher levels so, as long as you realize this is a hobby where nothing remains the same minute to minute and certainly not day to day because of entirely too many variables outside your control then you'll be fine.
Some days when I'm monitoring the civilian air band (118 to 136 MHz, AM mode) I'll have a nice flat noise floor with either of my R820T sticks, then the next day I'll see (literally) a 10-15 dB increase with peaks and "something" that just comes out of nowhere and I never can figure it out - the next day it'll be nice and flat again, the day after the noise is back, and so on. It can be frustrating at times but that's how it goes.
Anyway, you'll figure it out over time learning as you go along.
What isn't there now could come back tomorrow, or later today, or whenever and it can even come back with higher levels so, as long as you realize this is a hobby where nothing remains the same minute to minute and certainly not day to day because of entirely too many variables outside your control then you'll be fine.
Some days when I'm monitoring the civilian air band (118 to 136 MHz, AM mode) I'll have a nice flat noise floor with either of my R820T sticks, then the next day I'll see (literally) a 10-15 dB increase with peaks and "something" that just comes out of nowhere and I never can figure it out - the next day it'll be nice and flat again, the day after the noise is back, and so on. It can be frustrating at times but that's how it goes.
Anyway, you'll figure it out over time learning as you go along.
- Status
