Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Unintended electromagnetic radiation from Starlink satellites (aanda.org)
99 points by Podgajski on Jan 3, 2024 | hide | past | favorite | 55 comments


Don’t hold your breath on anyone doing anything about it. Regulators wont even go after intentional interference.

See: The guys in the southern U.S. running 10,000-100,000W AM transmitters with dirty amps that splatter 6 CB channels at a time across half the planet. And all they do is yell gibberish and try to drown each other out 16 hours a day, making swathes of the 11m band unusable when they get bored of 27.025 and start moving to higher channels.

They’ve been doing it for 30-40 years, and people have been complaining and reporting it the entire time.


SpaceX voluntarily has initiated brightness reduction in response to concerns from the astronomical community. They did a similar thing for radio astronomy (although there is at least some limited regulatory teeth here, unlike the brightness thing which has no regulatory teeth). https://api.starlink.com/public-files/BrightnessMitigationBe...

I don't know why people just assume nothing will be done. SpaceX probably just wasn't aware.


> SpaceX probably just wasn't aware.

They filed a Supplier's Declaration of Conformity without actually measuring unintended emissions? Seems a bit hard to believe, and rather disturbing if so. It's not like Part 15 Subpart B is in any way obscure!


It is not uncommon for this to happen.

It is also not uncommon for a version of a device to pass emissions, being modified during the testing process to do so, but those changes never make it back into the production BOM. Or if they do, the production build is never re-qualified.


> SpaceX voluntarily has initiated brightness reduction in response to concerns from the astronomical community.

SpaceX knew their satellites would cause issues and didn't do anything with the early starlink satellites. After extensive outrage they claimed they had "fixed" things but the reductions were trivial and have done little to address the problem.

https://www.scientificamerican.com/article/spacexs-dark-sate...

https://www.astronomy.com/science/starlink-satellites-disrup...


Those links don't say what you claim they say. Big red flag alert!

Your first link is ancient (September 2020) so it doesn't reflect any of the latest-and-greatest brightness mitigations. The study just looked at VisorSat, and SpaceX never claimed VisorSat "fixed" the issue completely.

Your second link has nothing to do with visible light reflections, and is actually about the same RF emissions as the current article (emissions which SpaceX is already in the process of mitigating[0] btw).

You're not even lying with statistics here. You're just lying with.... lying. :-\

  >SpaceX knew their satellites would cause issues
Citation needed. I'm also 100% sure you can't provide one.

All the reports from SpaceX and from the astronomical community alike both say the impact was surprising and unexpected. Do you have a (reliable, documented, and critically non-misrepresented) source that says otherwise? No?

[0] https://news.ycombinator.com/item?id=38863424


I'm totally fine believing SpaceX/Starlink did not expect it, the astronomical community on the other hand as far as I'm aware expected this and has been raising issues with SpaceX/Starlink the whole time (source: am an astronomer). As for things getting significantly better, that'd be news for me as well (https://arxiv.org/pdf/2309.14152.pdf seems to be the latest review I found on the issue, with Post-VisorSat being worse than VisorSat, and worries about Gen 2 sats).


The astronomical community, in fact, did not expect it. (Go ahead and do an internet search before the first large Starlink batch was deployed... there was no outcry; in fact, there's a public comment period for such things, and if there HAD been concerns, they would've been raised there, but they were not. After the first launch, it was a shock to the astronomy community... And SpaceX was also surprised.) The astronomical community was just as surprised as SpaceX. SpaceX acted remarkably quickly in inventing several different brightness mitigations and deploying them within months.

As far as your claim that the latest sats, are worse, that contradicts what your link says in the abstract: "The first of these [new V2 satellites], called Minis, are dimmer than Gen 1 Starlinks despite their greater size. Photometric observations verify that brightness mitigation efforts employed by SpaceX reduce spacecraft luminosity substantially"


> the astronomical community on the other hand as far as I'm aware expected this

Huh I didn't realize that. Do you have a source handy?

I couldn't find it in my brief literature search.[0] I found a couple references to Starlink brightness from 2019, but they're all after the first operational launch.

> https://arxiv.org/pdf/2309.14152.pdf seems to be the latest review I found on the issue, with Post-VisorSat being worse than VisorSat

Yes SpaceX needed to delete the visors to add laser hardware, so they developed a novel space-rated dielectric mirror as the new mitigation method. That was their first generation mirror, with the current second-gen mirror having better optical performance. As far as I know they haven't been characterized yet.

Iterative development takes time, it's true. It also takes money, which is why Starlink competitors are investing essentially zero resources in dark sky stewardship and R&D. Why bother? They know skywatchers will call anything and everything Starlink, so Bad Billionaire will soak up the blame regardless of who really caused it.

The competitors are astronomy's real 'enemy at the gate,' but most folks are too distracted by the shiny Elon Musk bouble to notice.

[0] for example https://scholar.google.com/scholar?q=megaconstellations+brig... and https://scholar.google.com/scholar?q=starlink+brightness&as_...


https://ui.adsabs.harvard.edu/ works much better for astronomy than google scholar, my search query was "starlink" and limited to astronomy papers.

My research is theory (not directly affected), so most of my background on this has come from grumblings over beers and the occasional talk (so I'm not really in the loop). Satellites (and other forms of interference such as RFI) have always been a problem for some groups (basically, anyone with long exposures, wide exposures or working with faint sources, it all comes down to the probability of a satellite appearing in the field of view mixed with how bad that is), but it's the massive increase in numbers plus the move over the last ~10 years to looking for transients (LSST is the poster child here, but there are lots of others) that have made it a more community-wide issue.

Other groups have been mentioned (I recall Amazon being mentioned), but the "focusing on Starlink" thing might be a reflection of what pop-science reports, and what's actually able to be measured being published.


Par for the course, unfortunately.


Could you use KiwiSDR to time difference of arrival triangulate?

https://www.youtube.com/watch?v=9CrKTm0wbRU

https://www.rtl-sdr.com/kiwisdr-tdoa-direction-finding-now-f...

https://news.ycombinator.com/item?id=17414699 ("HN: Using KiwiSDRs to locate HF radio transmitters and numbers stations")


One of the properties of shortwave is extensive multi-path propagation. Much of the signal goes the shortest route - and a small amount of it goes all the way around the world and arrives the long route, tens of milliseconds later. And at every point in between too. This is why shortwave AM radio sounds so hollow/resonant/unique - RF echoes stacked on top of each other, all slightly differently timed and distorted, arriving by many paths.

This can be adapted for, somewhat. For that sort of application (triangulating from multiple passive fixed receivers scattered over the continent) confidence to less than within 50 miles would be hard to pull off. Though if you do get that kind of fix with some confidence, you can drive around -- when you're fairly close to a transmitter of that power (within line of sight), it's hard to miss it.

Practically, few of these types transmit continuously. (10 kW is quite the power bill.) If they're only on the air for a few hours, with changing locations, it's going to be difficult to find them.


Sounds like https://www.he360.com/products/rfgeo-radio-frequency-signal-... might be a better fit in this use case.


That may work as a starting point for groundwave, but anything skywave is WILDLY inaccurate.

That “tool” needs to go away. Time and time again people get accused of causing interference because their station falls in the center of the heat map and the person using completely fails to understand what it is they’re seeing.

I was so annoyed at this going public that I pulled my Kiwi off the internet. Nice experiment, interesting code problem of synchronization, failure to let the public use it.


Not buying they’re running 100,000 watts, nor much over a few kilowatts. Which, isn’t in any way acceptable for a service that limits AM to 5 watts.

Splatter isn’t indicative of large amounts of power, just an RF amplifier going very non-linear.


Apparently this is the kind of hardware they are using.

http://www.xforceamps.com/item/Straight-64Pill---Krazy-Xtrem...

> Straight 64Pill - Krazy Xtreme Duty - 6.4KW-M

> Together they can draw up to 1280amps total off your mobile.

> The combined output power of the two units is +16000w PEP when supplied with enough electrical power.

edit: Larger one from 10 small ones with a maximum ampere draw of 980.

http://www.xforceamps.com/item/Straight-480Pill---Krazy-Xtre...

> Straight 480Pill - Krazy Xtreme Duty - 48.0KW-M

> Maximum Recommended Dead Key Output 12000w

> The combined output power of the Ten units is +130000w PEP when supplied with enough electrical power.

edit1: Also https://www.worldwidedx.com/threads/if-your-128-pill-amp-isn...

>Puts 48kW amp in car, gets effective radiated signal no better than a couple of kilowatts does due to the fact the car is a crap ground plane to start with.

>Someone commented that commercial AM stations are limited to a couple of kilowatts more. If you look at the antenna ground system AM stations use, there's a whole lot more metal involved than is in the body of a car.

edit2: Video from a smaller car https://www.youtube.com/watch?v=aOT63GUdQDc

edit3: More specialized cars https://www.youtube.com/watch?v=daKDu6FT2Xs

edit4: Truck that claims 60kw dead key output https://www.youtube.com/watch?v=S1YWke8cseg


100kW ERP would make more sense. A transmitter that actually consumes 100kW of electricity would be a monster.


> A transmitter that actually consumes 100kW of electricity would be a monster.

Before we switched to DTV from analogue, the TV transmitter tower near me was broadcasting FIVE 870kW channels. Admittedly these were between 600-700mhz and I'm not sure if that's ERP or not, but "monster" sized RF transmitters were surprisingly common.


Oh it's common, just not something you could casually hook up in a residential setting. You'd need access to 480v/3ph service to do this realistically. Moving 100kW at residential voltages to a single device is not happening. Your wires would have to be cartoon-sized.


This is a 50kW AM tower in some amount of detail — https://youtu.be/Aax-ehkRTnQ?si=Hy0YklPl06XM4tfR


Would like to read more about these menaces and the reasons for the government's refusal. Do you have any articles or blogs or even other comments you could link me to to read more about this?


Google or YouTube CB Channel 6 Superbowl. At one point it was sort of an unwritten rule that they kept to Channel 6, but now you’ll find them everywhere. Earlier today I observed one of them completely saturating the part of 27 MHz reserved for RC cars/planes/drones using my SDR… from Canada.


Your fellow canadian VE7KFM is one of the worst offenders:

https://ve7kfm.com/profiles.html

I've been yelled at by him before on 14.275, about 15 years ago. He's still at it.


That blog is something. Almost as deranged as the people it's profiling (doxxing?)


Its like the chris-chan phenomenon: Who is worse, the person the blog is focused on or the people obsessed with them?


by saturating, do you mean they deliver so much power that they effectively create a dynamic noise floor below which normal users fall, and are thereby lost signals for most receivers?

i don't have equipment to fiddle with but i'm fascinated, and not shocked at the behavior.


Exactly. I'm told there is at least one listening station that measures their power at distance, and they are trying to make sure their signal reaches these listening stations and nobody else's. I've heard at least 5 going at any given time. They play looped samples of their own voices with a few audio effects added on, then run that through the most powerful amplifier they can manage to build.


According to this they are guys driving around in cars: https://www.reddit.com/r/cbradio/comments/2fpwsk/so_what_do_...

Found a video of it: https://m.youtube.com/watch?v=B3uX8Ks5tBo


It’s like RF nerd narcissism with those people. Bizarre.


I did, and it's not the football Superbowl, but a term coined for a group of people who seem to just shout nonsense 24/7 on a CB channel?


correct. it's the CV superbowl.


Yes, but has a major purchaser of spectrum complained? No, because its CB and they're not interfering with AM/FM broadcast, first responders, aviation or cellular bands. Therefore, nothing will get done and I doubt the FCC is even monitoring it.


  >Don’t hold your breath on anyone doing anything about it. 
Maybe you meant, "don't hold your breath on regulators doing anything about it."

From the original study[0] that found unintended emissions from Starlink:

  >The authors are in close contact with SpaceX, and the company has offered to continue to discuss possible ways to mitigate any adverse effects to astronomy in good faith. As part of their design iteration, SpaceX has already introduced changes to its next generation of satellites which could mitigate the impact of the Swarm transponder emissions.
[0] https://arxiv.org/abs/2309.15672


I've heard very different stories about FCC enforcement. I wonder of its band specific?


The FCC regularly hands out five and six digit fines to violators.

https://www.fcc.gov/enforcement/orders


Excuse me, what? What is the point of doing this?


To argue over who has the bigger dick, basically


This is SpaceX, unintentionally or otherwise, making things worse for everyone else within range who uses those frequencies. The band in question (2 meters, around 144 MHz) is probably the #1 most commonly used band by the licensed amateur radio community.


It's Starlink.


SpaceX owns Starlink, it is a product of theirs.


Can you cite that? I thought they just operated it.


Switch mode power supplies in space. From the article:

> The broad-band features are with high probability caused by other means, such as switched-mode power supplies, communication signals internal to the satellites, or some other electronic or electrical subsystem.

As a radio amateur substandard switch mode power supplies are a major source of annoyance in the amateur radio bands.

Unfortunately you can't go and talk to your neighbours about it if the SMPS is in space :-(

I feel a lot of sympathy for the radio astronomers - radio is very hard in this electrically noisy world, and the radio astronomers have some extremely sensitive equipment (much more sensitive than radio amateurs).


That paper uses a nice unit I didn't know of; the 'Jansky' (Jy) ' 'non-SI unit of spectral flux density' https://en.wikipedia.org/wiki/Jansky


The movie Contact makes mention of Jansky - "Reading over a hundred Janskys". Thought it was a made up unit at the time


Nitpick: as a non-SI unit recognized for use in astronomy[0] "jansky" is always lowercase, not capitalized.

This is true for all SI units. Newton, watt, and joule are all lowercase, unless they appear at the beginning of a sentence or when you're referring to the original scientists.

People often get confused because the abbreviations for those units are capitalized.

[0] https://www.iau.org/publications/proceedings_rules/units/


Is it possible Starlink satellites are providing non-publicly known functionality at those (and other) frequencies to military and/or government agencies?


TFA won't load for me, does it list the frequency or show spectragraphs?


This seems rather amateur hour. Do satellites not undergo rigorous testing to ensure they're operating within constraints?

I've done plenty of consumer electronics in my day and leaking around 144Hz is frankly unimaginable at any scale above a toy. I'd love to see a post mortem on this one.


Ever check chinesium? This is possibly from swarm technologies


Tldr: lets make regulation to attack Musk specifically.


Thats not at all what this says lol. Its simply some research pointing out that Starlink likely has some unintended radiation at specific frequencies. Nobody is calling for regulations or anything, no need to instantly come to Musk's defense.


Sometimes when you do something wrong, there are consequences.


Yes but it hurts my feelings very very much and he's very very important to me, so much that, when a bunch of astronomers publish an academic paper with a cool RF investigation, the first thing I say is: "the government is At it Again! Regulatin!" -- no one talks bad about anything my Elon is near without showing what a shill they are.


Ha ha that's pretty funny but the article specifically states that there's no regulation around these types of rf emissions and that maybe there should be




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: