Elon Musk's 'Next-Gen' Broadband Service Is Overheating In The Arizona Desert
from the reinventing-the-wheel dept
As we've noted a few times, SpaceX's Starlink service will be a massive improvement for the up to 42 million Americans that lack access to broadband. The ongoing 10,000 user beta is delivering speeds between 50 and 150 Mbps at low latencies, something that's much better than the expensive, slow, high-latency, usage-capped traditional satellite broadband service most people hate.
That said, early reviews have also showcased how the service isn't going to be quite the miracle some Elon Musk fans are expecting. And this week complaints bubbled up among Starlink beta participants that their service dishes have been overheating in the summer sun:
"But according to user complaints on Reddit, the Starlink satellite dish user terminal, affectionately dubbed “Dishy McFlatface” by SpaceX, is experiencing temperature issues for some users. Reddit user SocietyTomorrow stated that his broadband connection shut down at noon in the Arizona Summer sun, only to kick back on again after being sprayed with a hose.
"I did submit a ticket and they only said it will shut down at 122 [degrees],” the user wrote. “Sadly tomorrow will be 122, and Wednesday will be 123. Dishy is already out at 112 so [I’m] gonna be quiet at home while I work out a solution."
To be clear this is the kind of stuff betas are for. It's not clear that this is happening at any scale (SpaceX/Starlink wouldn't comment). I spoke with an engineer that has dissected Dishy, and he suggested that fixing the satellite dish's heat issues, while very likely, could take some time:
"Since they've got a lot of custom silicon in there—likely the limiting factor—the turnaround time on this would be very slow,” he added. “They could resort to some form of active heat removal like fans or thermoelectric cooling, but then they burn a ton of power which would make Dishy even more power hungry than it already is."
“This is a really tricky engineering problem with some insanely tight constraints,” Keiter said. “The good news is that the team is pretty sharp."
But even if engineers resolve the heat problem, Starlink isn't likely to have quite the impact many people think. Wall Street estimates indicate that limited capacity (aka the laws of physics) means the service will likely be able to serve around 300,000 to 800,000 users with its initial fleet of around 12,000 satellites. That's around 1% of the U.S. broadband market. Even on the very optimistic end, with 42,000 low orbit, upgraded satellites feeding 60 Gbit/s each several years from now, 6 million users is the max number of potential customers, and those will mostly be rural, boat, or RV users that can afford it.
And this ignores Starlink's negative impact on night sky research (which US regulators certainly are). And the fact that in a post net neutrality world, the service is likely to eventually impose odd caps and throttling restrictions to manage capacity. And the fact that at $600 for the first month, it's not really going to be a service for the tens of millions of Americans that lack access due to affordability. And some of the quirky line of sight issues indicated in early reviews.
Fiber networks of course don't suffer from any of these restrictions, which is why many in telecom circles argue that if you're going to subsidize broadband, you should be subsidizing future-proof broadband without odd caveats and constraints. It's also why consumer groups have been a bit disgusted at the fact that the FCC recently threw at $886 million at billionaire Elon Musk so he could deliver Starlink broadband to some traffic medians and a few already served airports. Again, Starlink will provide some help to the uncompetitive US broadband market, but a miraculous US broadband market disruptor it's not.
Thank you for reading this Techdirt post. With so many things competing for everyone’s attention these days, we really appreciate you giving us your time. We work hard every day to put quality content out there for our community.
Techdirt is one of the few remaining truly independent media outlets. We do not have a giant corporation behind us, and we rely heavily on our community to support us, in an age when advertisers are increasingly uninterested in sponsoring small, independent sites — especially a site like ours that is unwilling to pull punches in its reporting and analysis.
While other websites have resorted to paywalls, registration requirements, and increasingly annoying/intrusive advertising, we have always kept Techdirt open and available to anyone. But in order to continue doing so, we need your support. We offer a variety of ways for our readers to support us, from direct donations to special subscriptions and cool merchandise — and every little bit helps. Thank you.
–The Techdirt Team
Filed Under: overheating, satellite broadband, starlink
Companies: spacex
Reader Comments
Subscribe: RSS
View by: Time | Thread
I've been waiting for the day that i would finally hear, "I had to spray my internet connection with a hose".
This is indeed what beta testing is for, but how we didn't forsee this little fly in the ointment is hilariously beyond me. Wonder what the low temperature and humidity constraints are.
[ link to this | view in thread ]
Starlink is still better than what a lot of people can get
So even if it's not perfect, it's still a huge improvement for many.
And the night sky issues are WAY overblown.
[ link to this | view in thread ]
Re: Starlink is still better than what a lot of people can get
Spoken like a true fan-boy!
[ link to this | view in thread ]
This is a simple engineering problem that will be addressed. There are more valid criticisms.
Even if they have to supply active power to the thing and cool it with peltiers, it'll work. Or probably there's some crappy water-cooled system they can put on it... wouldn't that be a hoot if you had to go fill up your internet's radiator once or twice a year.
[ link to this | view in thread ]
Re: Re: Starlink is still better than what a lot of people can g
Please refute with facts if you don't agree.
[ link to this | view in thread ]
Re:
Sorry, my internet froze on me overnight, I have to restart the download...
[ link to this | view in thread ]
Déjà vu
Sounds like Tesla's temperature issues with the screens in some of their cars because they were too cheap to use automotive grade components. Otherwise, any respectable electronics circuit engineer should know better. That's what component datasheets (analogous to SDSes for chemicals) are for.
https://www.thedrive.com/tech/27989/teslas-screen-saga-shows-why-automotive-grade-matters
[ link to this | view in thread ]
solar powered fan?
Wire up a small solar panel to a small cooling fan. Anytime the sun is shining, the fan is spinning... might not be "active cooling", but perhaps some airflow might help?
[ link to this | view in thread ]
Heat and energy dissipation
I live in Arizona and this is a normal consideration for all electronics. For a while I founded and ran two ISPs. Outdoor installations often are NOT as exposed as SL dishes. Sometimes they are in an in-ground "vault" and sometimes an above-ground "pedestal." In either case these have passive cooling due to a variety of factors.
These factors include a lack of space/power to place a direct gas expansion cooler. (think A/C unit size being bigger than the dish and making noise. Add in residential considerations like neighbors and HOAs and CCRs and this is a moot point.)
You can do non-active cooling -- energy exchangers -- but you typically are limited to both space (must be able to exhale energy) and technology (typically reduces temperatures 15-20°F at best.)
Even for devices without analog RF radios, it's tough to survive in the Arizona desert. ToughSwitches and Ubiquiti WiFi cams are two examples of "how soon will this die."
SL is no different in that regard. It's hardware that takes in power (energy) and in operating converts it to heat. It can't remove that heat in the presence of similar heat it's encased in (the air). Differential heat exchange requires a useful difference.
To summarize: ALL power-consuming devices of less than 100% efficiency must have a way to remove the generated heat. SL is just one more example.
Ehud
Tucson, AZ
110°F in the shade right now.
[ link to this | view in thread ]
Having $100 in your bank account is better than having $0, but it ain’t the same as having $100,000.
[ link to this | view in thread ]
Re: Déjà vu
there were no automotive grade screens available at the time.
[ link to this | view in thread ]
Re: Re: Starlink is still better than what?
Musk should just crawl back in his hole and let the telcos do their job. One more market with to many products chasing a fixed customer base, hurting all the players.
[ link to this | view in thread ]
Policy vs Practicality
I understand that Techdirt is very much focused on the big picture and policy and I believe that that is a great thing.
But I cringe a little bit at these articles that in my opinion are doing what you so often criticize others for doing, setting up a straw man to knock down.
You treat Starlink in every article as if it is supposed to be the solution to broken broadband and the criticize it for not being that solution.
And look I am not blind to the policy and industry efforts that point to any new technology and say look we do not have to improve or fix our issues because of whatever the flavor of the day is let it be Starlink or 5G and I appreciate that you push back on those efforts.
Starlink has never been promoted as far as I can tell as the or even a fix for the broken broadband industry, except by those that want to detract from the industry failures. If you want to criticize Starlink for its flaws, and it clearly has some but don't put it up on a pedestal and take cheap pot shots at it.
But for those of us that cannot get decent internet any other way it is a godsend. Is it better then fiber clearly not but is it better than fiber that I realistically am never going to get Hell Yes it is.
I think a little bit more nuance in these Starlink articles would be nice.
[ link to this | view in thread ]
Re: Re: Re: Starlink is still better than what?
You seem to be working under the delusion that telcos will do their jobs or that they have done them in the thirty odd years.
[ link to this | view in thread ]
Re: Re: Starlink is still better than what a lot of people can g
Spoken like a true, irrational, hate-driven bigot!
[ link to this | view in thread ]
Re: Policy vs Practicality
Elon, that you?
I don't think Techdirt treats Starlink like it is supposed to be the entire solution for the broken braodband industry, just a partial solution. This seems like a foreseeable whoops. A roof, in places like Arizona get very hot. Usually, a lot hotter than the air temp. I've been on a roof with the shingles getting soft under my feet from the heat.
[ link to this | view in thread ]
Re: Re: Déjà vu
And on top of that, nobody was going to sell him milspec ones .... well, because they didn't want to.
[ link to this | view in thread ]
Re: solar powered fan?
Sadly thermodynamics says no.
If the ambient air temperature is at or above the threshold then no amount of air glow is going to help.
Evaporative cooling is viable as long as you don't have water issues.
[ link to this | view in thread ]
Re: Re: Policy vs Practicality
That is one of things that raises the question.
What the heck are people thinking using dark colors and shingles for roof coverings in a hot dry desert environment?
A little while ago one of my technical articles mentioned the whitest white with an extraordinary level of reflectivity which would lower the heat load on structures.
[ link to this | view in thread ]
Re: Re: Re: Policy vs Practicality
I totally agree. I see so many dark colored shingles in high heat areas. It's just crazy. A switch to white/light colored shingles alone would save huge amounts of energy on a large scale.
[ link to this | view in thread ]
Re: Re: Re: Starlink is still better than what a lot of people c
The facts are that Starlink / Musk fans bring up irrelevant shit any time Starlink is mentioned.
We are quite aware that it will provide a service. It does not, by any stretch of the imagination, change the nature of the market in the area of competition, or affect the urban underserved, who generally couldn't afford it anyway. But if that is mentioned, look out for the fan squad to throw some irrelevant if sometimes true facts into the conversation. It is clearly only for shipping, as a counter to something they think is begative toward their idol.
In this case it is still that, plus some blatantly forseeable (imo) issues in their shakedown beta testing, which is sort if fine, and yet of interest. But, "Waah it's better than nothing!"
We know that. Starkink knows it. The only ones who cannot seem to come to grips with this are those subscibed to their self-generated Starlink or Musk cults.
Also the sky issues are not "overblown". That's my opinion against another opinion. If you don't care for the facts given by astronomers and company, i guess you just need to assume they all joined the anti-Starlink sect because reasons.
[ link to this | view in thread ]
Re: Policy vs Practicality
You treat Starlink in every article as if it is supposed to be the solution to broken broadband and the criticize it for not being that solution.
That isn't techdirt's strawman. Earlier on, Starlink made claims, but mostly the claims are made by politicians, wonks, and industry groups because "look there is healthy competition, nothing to see here, there are no probkems with this sector, move along." Also claims are made by the faithful because some people need to belive in things and defend them from ... something.
So that is who is being addressed, and there is a point to it. Karl, techdirt, and the commentariat don't just pull these things out of thin air.
[ link to this | view in thread ]
I strongly dislike Elon, but he's still far more tolerable than a temperature of 123. I would just use a 56k connection to buy the supplies to off myself.
[ link to this | view in thread ]
'Competition'. The thing you're arguing against is 'competition'
Yes, because they've been so very good at doing that before his entry into the market...
[ link to this | view in thread ]
Welcome to beta testing...
Users realize the service is still in beta, correct? Beta testing is for finding issues like this. I hope their engineers will start testing at these temperatures and see if they can up the shutdown limit in firmware, or use wider temperature range components in future revisions, or increase heat dissipation in future revisions.
[ link to this | view in thread ]
Re: Heat and energy dissipation
I live in northern MN, and keeping equipment working at -30F can be a challenge also. Many times you have to add heaters to cabinets.
[ link to this | view in thread ]
Re: Re: Re: Starlink is still better than what?
"...and let the telcos do their job."
Their "job" is to abuse their monopolies and provide the worst service for the highest expense...in the US, at least, where the regulatory body of telecommunications has been the well-trained pet of the telco lobby for decades.
I'm no fan of Musk and his Starlink "solution" doesn't scale, rendering it a gimmick for a few rather than an aspect of infrastructure for the country. It certainly isn't the answer to abusive US telcos. It is, however, a successful proof of concept and as a business idea, may end up profitable.
Assuming his engineers get around to proof his receivers against the type of temperature you might expect in the places he intends to serve.
[ link to this | view in thread ]
Re: solar powered fan?
"... but perhaps some airflow might help?"
Only if the air in question is cooler than the gadgetry you want cooled. There are only three possible solutions;
Active cooling. Turn the casing of the device into a fully functional refrigerator and use heat-transmissive coils to lead the heat away from sensitive components.
use components capable of toughing out the type of heat and cold the device is likely to be exposed to.
This is an old problem every enthusiast assembling their own home computer runs into and has to solve - and it beggars belief that Musk's engineers apparently never considered the fact that their device intended for rural areas would in fact not operate in an air-cooled environment...
[ link to this | view in thread ]
Re: Heat and energy dissipation
"To summarize: ALL power-consuming devices of less than 100% efficiency must have a way to remove the generated heat. SL is just one more example."
Even enthusiasts assembling DIY computers know this. What shocks me is that Musk's actual engineers missed the bit about their device having to work well away from AC of any kind.
[ link to this | view in thread ]
Re: Policy vs Practicality
"You treat Starlink in every article as if it is supposed to be the solution to broken broadband and the criticize it for not being that solution."
...because yes, that's what Starlink is hyped as. The solution for the absence of a viable broadband solution in rural areas. That's literally what's written on the subsidy application Musk's filed with the FCC.
"I think a little bit more nuance in these Starlink articles would be nice."
I don't think Techdirt has denied, at any point, that for a select few Starlink will be a boon. What it does is raise conservative evaluation about an application of technology which does not conform to the hype. Will Starlink be worth a tax-dollar subsidy, for instance? Does it mean the US can stop trying to roll out landlines? Those questions are fairly important and both have to be answered in the negative if we're factually fair. And yet the body politic appears to lean towards answering both of those questions to the contrary.
[ link to this | view in thread ]
Re: Welcome to beta testing...
'It's beta testing' only excuses so much though, it's not like extreme temperatures were an unknown factor that's only now evidently something they'll have to deal with, that's something that should have been planned and designed for from the outset.
[ link to this | view in thread ]
Not surprising to be quite honest. The best solution is often one that doesn't require you to shoot stuff into space. In this case, I'm wondering why Musk just doesn't promote the use of directed wireless relays with last mile hub/spoke wired internet? I mean the last mile wired part is expensive but you could make your wireless business the main part and have cooperatives sell the last mile to folks in rural communities. This all seems to be another grift to subsidize SpaceX.
[ link to this | view in thread ]