Starlink dishes go into "thermal shutdown" once they hit 122° Fahrenheit:
A Starlink beta user in Arizona said he lost Internet service for over seven hours yesterday when the satellite dish overheated, demonstrating one of the drawbacks of SpaceX's broadband service. When the user's Internet service was disrupted, the Starlink app provided an error message saying, "Offline: Thermal shutdown." The dish "overheated" and "Starlink will reconnect after cooling down," the error message said.
[...] The user, named Martin, posted a screenshot of the error message on Reddit. He contacted Starlink support, which told him, "Dishy will go into thermal shutdown at 122F and will restart when it reaches 104F." Martin decided to give the dish a little water so it could cool down. He pointed a sprinkler at Dishy, and once it cooled enough to turn back on, "I immediately heard YouTube resume playback," he wrote yesterday.
But the Internet restoration was short-lived, Martin told Ars in a chat today.
"The fix was temporary," he told us. "When I stopped the sprinkler, [the dish] heated back up and would cycle back on for a few minutes and go back down for thermal shutdown.
(Score: 2) by driverless on Wednesday June 16 2021, @08:10AM (4 children)
Friend of mine had that, Dave's Syndrome I think it was called.
(Score: 2) by mcgrew on Wednesday June 16 2021, @03:26PM (3 children)
Indeed. 120 degrees is pretty damned hot. Visiting a friend in Arizona once, 115 melted a cassette that was laying on the car seat, but most places won't get that hot. However, if it's painted black, anything in the nineties would likely heat it up to 120 pretty easily.
And solid state electronics can't take much heat or current, unlike their analog counterparts.
mcgrewbooks.com mcgrew.info nooze.org
(Score: 0) by Anonymous Coward on Wednesday June 16 2021, @04:58PM
Starlink dishes are white plastic.
(Score: 0) by Anonymous Coward on Wednesday June 16 2021, @05:04PM (1 child)
This sentence is a little bit confusing because even though the meaning of "solid state" has changed over time it has never meant "not analog".
But anyway, silicon-based semiconductor components most commonly have a maximum rated operating temperature of 125°C or more (175°C is also common). Usually in electronics other components (such as capacitors) will have lower operating limits, with 85°C being typical. This is well above 50°C (122°F).
Of course, operating limits for individual components doesn't mean that the overall product will work at those temperatures. Most characteristics of an electronic device will vary depending on temperature. Good designs work properly over a wide temperature range. Bad designs don't. Probably someone goofed in the design stage, they found problems at >50°C in testing and it was too late to fix it properly if they wanted to meet their delivery timelines.
The fact that they did work to add this thermal shutdown suggests the problem was likely discovered while they were trying to pass regulatory compliance testing (EMI, perhaps?)
(Score: 2) by Reziac on Thursday June 17 2021, @02:43AM
Anything sitting out in the desert sun will easily reach 160F.
Guess Arizona is not their market.
And there is no Alkibiades to come back and save us from ourselves.
(Score: 5, Funny) by Rosco P. Coltrane on Wednesday June 16 2021, @08:36AM
There, FTFY.
(Score: 4, Insightful) by Runaway1956 on Wednesday June 16 2021, @08:43AM (12 children)
Dude probably could have driven to his favorite convenience store, bought two (or ten) bags of ice, and cooled his dish quicker than that. Water through a garden hose probably isn't all that cool, with the sun shining on it.
Shading the dish is an obvious solution. Hope he thinks carefully about the material he uses.
The ham antenna tower idea? Not sure about that. Is it common for the air at ground elevation to be significantly warmer than 20, or even 40 feet above ground? Not being from Arizona, I wouldn't know about that. There's certainly not much shade up in the air, so he better plan on some kind of canopy or something that will mount on the tower.
Musk needs to re-engineer that dish. 122 degress is not at all uncommon in direct sunlight. Just lay a black steel framing square on a shingle roof in direct sunlight, and try to pick it up. You'll put on a pair of gloves before reaching for it a second time.
(Score: 2) by DannyB on Wednesday June 16 2021, @01:55PM (4 children)
I don't think a couple bags of ice would last very long at all with such high outdoor temperatures.
Shading the dish is a good idea, as long as you can shade it with something that is transparent to the RF the dish and satellites use. Even shading might not work. If outdoor temp is very high, and dishy generates its own heat from operation, it might still go into thermal shutdown.
I don't know about the ham antenna tower, but it sounds interesting. Maybe combined with a shade or "dome".
The lower I set my standards the more accomplishments I have.
(Score: 1, Insightful) by Anonymous Coward on Wednesday June 16 2021, @05:14PM (3 children)
Dishy uses a 180W power adapter [hackaday.com] and typical draw is reportedly around 100W. That is a lot of heat to get rid of on a hot day. If Starlink can't get high temperature rated versions of their electronics then places like Arizona will need active cooling in the summer. There is also a worldwide silicon shortage right now so even if the components exist they might not be able to get them.
(Score: 2) by DannyB on Wednesday June 16 2021, @06:12PM (1 child)
It sounds like Starlink's product may be unsuitable for large rural areas in a number of Southern US states.
I wonder how Starlink will do in Australia?
The lower I set my standards the more accomplishments I have.
(Score: 0) by Anonymous Coward on Wednesday June 16 2021, @07:10PM
Until they fix the overheat problem they aren't going to be able to offer reliable service in hot climates. On the other hand a lot of potential customers in those areas would still gladly fork over because it is still better than what they currently have. "Better than Nothing" seems about right. I do notice that they started in middling latitudes and have been gradually expanding both north and south. Expected thermal issues may have played into that decision since it lets them focus on getting the system up and working before needing to deal with thermal hardening.
(Score: 0) by Anonymous Coward on Wednesday June 16 2021, @07:47PM
is there some reason why the electronics have to be in the dish itself. it seems to me the only thing exposed to the heat should be the actual dish and a cable going to the electronics that are in the building a/c. what am i missing?
(Score: 0) by Anonymous Coward on Wednesday June 16 2021, @04:24PM (2 children)
You run your hose for a minute to get the hot water out, then it goes cold.
(Score: 2) by DannyB on Wednesday June 16 2021, @06:13PM
Aren't those hot Southwestern states currently experiencing a major drought and trying to save water?
The lower I set my standards the more accomplishments I have.
(Score: 0) by Anonymous Coward on Thursday June 17 2021, @03:18AM
> ... then it goes cold.
Not in College Station/Bryan TX when I was there one summer. The local water treatment plant had their evaporative chiller (giant swamp cooler) go off-line. When I came in from 100F+ outdoor temps and tried to take a cold shower, the water from the "cold" tap was warm, and stayed that way.
(Score: 5, Insightful) by PinkyGigglebrain on Wednesday June 16 2021, @04:45PM (2 children)
Ice would actually have been less effective at cooling the dish. it's a solid that would have had limited contact area with the dish's surface. Plus you would have to set up some kind of holder, and once it melted it would just flow off the dish without absorbing more heat than water from a garden hose. Flowing Water, even if it was at 100F, would have been able to carry away more heat more efficiently than the ice just sitting in contact with the dish, because it could contact more of the dish.
Something being colder doesn't always mean it's better at cooling, you have to also look at how well the heat can transfer to it.
"Beware those who would deny you Knowledge, For in their hearts they dream themselves your Master."
(Score: 2) by DannyB on Wednesday June 16 2021, @06:14PM (1 child)
Is ice transparent to the frequences used in Starlink?
(I honestly do not know.)
The lower I set my standards the more accomplishments I have.
(Score: 0) by Anonymous Coward on Wednesday June 16 2021, @09:04PM
It depends upon how thick, of course. It operates on either side of a water absorption peak around 25 GHz.
(Score: 3, Informative) by Reziac on Thursday June 17 2021, @02:53AM
28 years a desert rat... the higher you are off the ground, the cooler it gets, because you're away from the reflected heat coming off the ground, and usually catch more breeze. But when the ambient air temperature is over 100F (hardly unusual in the desert) you won't gain much even if your dish is on a long pole. And anything in the sun can easily heat up to 160F; your shade needs to be insulated, or it'll just get a little less cooked. Also, roofs reflect heat just like the ground, so atop your roof is not a solution.
And there is no Alkibiades to come back and save us from ourselves.
(Score: 0) by Anonymous Coward on Wednesday June 16 2021, @08:54AM (4 children)
My first thought, 122F what an awkward number to shut down. Seems like Starlink uses the Celcius scale.
Anyway, the big question is why? Never heard of this problem with TV satelite dishes and I guess the important electronics aren't build into the dish, right? And 50 degrees C... a cpu runs hotter than that.
(Score: 5, Interesting) by driverless on Wednesday June 16 2021, @09:28AM (2 children)
Not Starlink but their suppliers. Something in there has a max.operating temperature of 50 degrees C. Commercial is usually 0 - 70, but since they claim 22 below zero Fahrenheit up to 104 Fahrenheit which is -30 to +40 they're using components rated on a different scale.
(Score: 1, Interesting) by Anonymous Coward on Wednesday June 16 2021, @07:23PM (1 child)
I posit that a component with a 50°C max operating temperature is not a likely problem. The existence of a software shutdown like this makes no sense if the problem was just the designer picked an unsuitable part (or worse, the manufacturing team susbsituted a cheaper part). Such a scenario would more likely lead to the problem being ignored completely, or just swapping in a different part with an appropriate rating.
This shutdown smells like an "oh fuck" kind of 11th-hour quick fix. Some serious design problem was discovered late in testing and this was the only way they could get the finished product out the door in time. Probably either a thermal design issue (e.g., something running so absurdly hot normally that a slightly elevated ambient temperature leads to damage) or a regulatory compliance issue (e.g., the product failed EMI testing at elevated temperatures).
(Score: 2) by driverless on Thursday June 17 2021, @09:11AM
Sure, I was just looking at the temperature range and noticed it was the commercial 70 degrees shifted down a bit. I know that when we've tried to source devices it's been relatively easy to find stuff that will run at higher-than-commercial temps but once you get to lower-than-commercial temps it gets tricky. I know that some passives can do odd things at low temps but once you get to below about -30 it's really hard to find devices that are rated for that.
You're right though, this one seems like a patch for something...
(Score: 5, Interesting) by ElizabethGreene on Wednesday June 16 2021, @12:52PM
With a "normal" satellite dish the electronic components are in the LNB, a small light-colored device with lots of surface area that is isolated from the main area of the dish. Dishy mcflatface is a panel antenna with a bunch of electronic kit built into the main solar collecting area of the dish.
It's a requirements problem really; Designing the hardware in Redmond Washington they didn't think to include "Needs to work at temperatures high enough to scramble eggs/bake cookies."
Thermal issues for RF gear is not uncommon. I had to move a point-to-point access point out of an attic once for the same reason. It would go kaput in the heat of our Tennessee summers.
(Score: 1) by venn on Wednesday June 16 2021, @08:59AM (8 children)
With all the chips and circuitry glued to the back of the dish, I wonder how much heat that's adding to ambient. Might be time for some liquid cooling via radiators in warmer parts of the globe.
(Score: 2) by canopic jug on Wednesday June 16 2021, @09:06AM
It doesn't have to be all that warm a part of the globe, just painted black and sitting in the direct sunlight [trb.org] during the summer. They really picked some poor temperature limits and placement of temperature sensitive components. It is a mistake worthy of a product recall,
Money is not free speech. Elections should not be auctions.
(Score: 0) by Anonymous Coward on Wednesday June 16 2021, @09:38AM (6 children)
Sitting on the ground like it is, that dish is easily 40 degrees warmer than the measured ambient air temp (which is taken at a higher level and in the shade), depending on the exact composition of the soil below. In the sun, it can be even more. That dish is toast without active cooling of some kind in that sort of weather. Without it, the apparent temperature will actually increase as the air flow increases over the exposed surfaces! I know it sounds weird compared to our usual experience, but instead of a bubble of warm air being blown away causing things to cool faster, the bubble of "cold" air surrounding it gets blown away causing things to heat faster.
(Score: 0) by Anonymous Coward on Wednesday June 16 2021, @10:05AM (3 children)
Theory relies on existence of a bubble of cold air.
Proof of existence not supplied.
A resistive load cooling the air around it violates all laws of electronics and thermodynamics I can think of.
(Score: 0) by Anonymous Coward on Wednesday June 16 2021, @10:47AM (2 children)
That is why cold was in scare quotes. It isn't actually cold by our standard sensibilities, but it is cold relative to the other air masses of the system and its low thermal conductivity acts as insulation from the warm air masses while present. Without it in the hot apparent temperature, you will actually add heat to the system faster than the rest of the system can get rid of it, even with the electronics off.
(Score: 2) by FatPhil on Wednesday June 16 2021, @12:21PM (1 child)
So we conclude that now you've admitted the bubble of cold air is bogus, your theory is false.
Great minds discuss ideas; average minds discuss events; small minds discuss people; the smallest discuss themselves
(Score: 0) by Anonymous Coward on Wednesday June 16 2021, @09:50PM
https://time.com/5644737/fans-can-make-you-hotter/ [time.com]
Relative measures. Plenty of other sources too.
(Score: 2) by Reziac on Thursday June 17 2021, @02:58AM (1 child)
I experienced this one summer in Reno ... around 100F, near-zero humidity, and blowing about 25mph.... felt hotter out in the wind than standing behind shelter. Quite unexpected.
And there is no Alkibiades to come back and save us from ourselves.
(Score: 0) by Anonymous Coward on Thursday June 17 2021, @10:36AM
Yes, high heat can have unexpected effects due to how the apparent temperature can change. Another good example is that covering yourself in multiple layers with a shell of dark color can make you cooler than few layers of light colored clothing.
(Score: 5, Insightful) by Mojibake Tengu on Wednesday June 16 2021, @09:30AM (1 child)
We already discussed this poor design in the past.
Just for a reminder, the contraption is designed and manufactured by STMicro.
https://wccftech.com/starlink-user-terminal-apple-supplier-teardown/ [wccftech.com]
https://hackaday.com/2020/11/25/literally-tearing-apart-a-spacex-starlink-antenna/ [hackaday.com]
Currently, they wanted a new scapegoat:
https://www.linkedin.com/jobs/view/starlink-manufacturing-and-test-engineer-electromechanical-at-spacex-2493157558 [linkedin.com]
Respect Authorities. Know your social status. Woke responsibly.
(Score: 2) by RS3 on Wednesday June 16 2021, @11:06AM
I wonder if a small Peltier cooler would be good enough to fix it?
(Score: 3, Funny) by Gaaark on Wednesday June 16 2021, @10:27AM (9 children)
Paint it, paint it, paint it.... paint it white.
--- Please remind me if I haven't been civil to you: I'm channeling MDC. ---Gaaark 2.0 ---
(Score: 2) by RS3 on Wednesday June 16 2021, @11:01AM (5 children)
Maybe. Some things that look white to you and me look very black in infrared light. But yeah, use some kind of paint that's infrared reflective.
(Score: 2) by ElizabethGreene on Wednesday June 16 2021, @01:21PM (4 children)
Soylent ran an article a bit ago about a cheap superwhite that was IR reflective. Barium Sulfate? i think? It's not a bad idea.
(Score: 0) by Anonymous Coward on Wednesday June 16 2021, @05:22PM (2 children)
Is it transparent to RF?
(Score: 2) by RS3 on Wednesday June 16 2021, @07:01PM
Great question big AC! Not much time to research (have to fix a cranky robot) but most IR reflective coatings reflect RF too.
https://www.milpaints.com/infra-red-reflective-paint.html [milpaints.com]
https://www.sciencedirect.com/science/article/abs/pii/S092583881832214X [sciencedirect.com]
The problem is more of scattering the RF, so rather than focus it on the antenna target, you'd splatter it on everything else and lose your signal.
I think a Peltier cooler https://en.wikipedia.org/wiki/Thermoelectric_effect [wikipedia.org] , and maybe some kind of shade, if possible, would be the fix.
Well, really a better design that can tolerate the temperature would be the correct fix.
A big issue: heat increases noise in electronic components, and in many things electronic and signal amplification and detection, like especially RF reception, noise is the enemy. In many extremely sensitive sensors and amplifiers, the sensors and "front end" amplifiers have to be super-cooled to reduce the internal "shot noise" of electron movement. https://en.wikipedia.org/wiki/Shot_noise [wikipedia.org]
(Score: 2) by ElizabethGreene on Wednesday June 16 2021, @09:58PM
I genuinely don't know, and it's a good question. I assumed it was transparent based on it being a thin film much less than the wavelength and largely non-conductive, but I shouldn't do that. Even small losses matter when you are talking about a dish that uses technology indistinguishable from magic to receive a signal transmitted from a fancy camp lantern 500 km away.
(Score: 2) by RS3 on Wednesday June 16 2021, @07:04PM
Great point. No time to research (yikes) but as I commented below, if it absorbs or scatters RF it would defeat the purpose. My gut feel is that it would work pretty well, but I still like my Peltier cooler idea in addition to IR reflective paint. :)
(Score: 2) by c0lo on Wednesday June 16 2021, @02:34PM
Bad taste. Not gonna match with the color of those satellites, now that they were told those were too reflective.
https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford
(Score: 2) by ChrisMaple on Wednesday June 16 2021, @02:44PM
Almost half of the energy of sunlight is in the visible portion of the spectrum, and much of the remainder is near-infrared. The ideal coating would be reflective in UV, visible, and near IR, black in the far infrared.
(Score: 2) by bussdriver on Wednesday June 16 2021, @11:08PM
Obviously, paint them white! With a week of temps in that range you need to do more than worry about the sunlight hitting it, especially as the number of hot days increases.
(Score: 0) by Anonymous Coward on Wednesday June 16 2021, @11:55AM
just place it in a bucket of water?
(Score: 0) by Anonymous Coward on Wednesday June 16 2021, @01:00PM (1 child)
Go find shade when it's too hot, sun when too cold.
I wonder if the dish would work under a sun umbrealla?
Cloth might reflect sun and still be transparent to RF.
(Score: 0) by Anonymous Coward on Thursday June 17 2021, @01:52AM
50C is not enough for outside telco equipment, even with countermeasures like shade.
The number is 60C at altitude, where the air doesn't cool as well.
Alternatively, test to 70C at low altitude and then use countermeasures in hot climates.
Building stuff for 70C requires attention to detail in part selection and thermal management.
Dishy's purpose is to learn what they need to do.
Many lessons in talking to birds.
But looks like basics as well.
BTW, the other end is -40C ambient with warmup permitted.
(Score: 1, Insightful) by Anonymous Coward on Wednesday June 16 2021, @01:49PM (2 children)
that things that sit in full sunshine all day on a hot summer day get *very very warm*.
This would be a design fail on starlink's part. They did not design for the true thermal extremes the antenna was going to operate within. And it also sounds like they forgot to consider radiant heating from direct sunlight as well.
(Score: 2) by c0lo on Wednesday June 16 2021, @02:36PM
Not the first time they forgot.
Remember those reflective satellites that were giving grief to astronomers?
https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford
(Score: 2) by DannyB on Wednesday June 16 2021, @06:16PM
In those warm Southern states experiencing extended drought, one of the things you don't see a lot of is: shade
No trees. Maybe Palm trees.
No grass. Rock lawns.
The lower I set my standards the more accomplishments I have.
(Score: 2) by looorg on Wednesday June 16 2021, @02:33PM (1 child)
According to Starlink they are supplying broadband internet access to meet the needs of customers across the globe in under-served areas of the planet. Clearly it doesn't appear to be a device designed for those under-served areas of the globe. They have a tendency to either be to hot, to cold or to humid. Which is why people doesn't really live there, or they are shitholes without proper infrastructure. So the device should be tailored for those conditions and not for conditions at the lab in Redmond, I guess they might have solved the humidity problem there.
In some regard it's actually even worse then at first glance described, since the device will naturally be or start at the ambient temperature and go from there. So during the summer, in most places, or in the deserts of Arizona there is not a lot of room before the thermal shutdown. I'm surprised it hasn't happened more often, or I guess it has happened often enough since they included sensors and an error message but they just don't want to talk about it or mention it. If left in the sun long enough it probably doesn't even have to be powered on to reach the temperature and then never being able to power on. Just the sun, without the components running should be more then enough to reach the temperature. Seems like a clear design failure that they have not provided ample ventilation or cooling for the device to keep it in operating range. Still just vent holes and a fan might not be enough to keep this on for long in a desert.
I do wonder if there is a lower thermal shutdown, sort of what would happen in the arctic during the winter. Will the machine freeze to death?
(Score: 1, Informative) by Anonymous Coward on Wednesday June 16 2021, @03:47PM
Many years ago, while in college, the car I had at the time had a black top.
As well, I also had an AC service thermometer in the AC ducts (largely to make sure the AC was working properly).
Sitting in the college parking lot, in full sun for much of the day, when I'd return at 3pm after classes, that thermometer sometimes read 140-150 F (60-65.5C). Which is well above the limit starlink has set.
So someone, somewhere, at starlink was not at all aware of just how hot something sitting in full sun all day long can become.
(Score: 1, Insightful) by Anonymous Coward on Wednesday June 16 2021, @06:28PM (4 children)
Outdoor electronics that can't work at temperatures higher than 50C is defective by design...
(Score: 2) by RS3 on Wednesday June 16 2021, @07:16PM (1 child)
Agreed. It's sadly a cost-cutting thing- better electronics generally cost a lot more. Hopefully the negative PR over this will cost them a lot and they'll once again learn a hard lesson about cheapening something to the point of uselessness.
(Score: 0) by Anonymous Coward on Wednesday June 16 2021, @09:26PM
Cost cutting is very likely the reason. Proper, fully heat resistant, devices would have been much more expensive.
But it is also very possible they just neglected to consider the temperature rise that results from a full day in cloudless full sun on the summer solstice.
One would have to be a fair bit "up in the north" before "temp. rise from full day of cloudless sunshine" would not risk hitting 122F inside a closed box. 122F is not hard to obtain from full sun on a cloudless summer day in a great many locations within the US.
(Score: 0) by Anonymous Coward on Wednesday June 16 2021, @07:18PM (1 child)
That depends entirely on where it will be used. In Nevada heat matters, but in Alaska cold resistance is the critical factor. In the middling latitudes where most of their customers live their current temperature range works just fine. I expect extreme heat and cold rated dishies to hit the market sometime during the next year or so.
(Score: 0) by Anonymous Coward on Wednesday June 16 2021, @09:03PM
Even in those 'middling' areas, a day of full sunshine could easily push the actual electronics past this heat cutoff point, even if the ambient is still within normal temp ranges (i.e., not record breaking heat).