5G Power Usage Is Making Phones Overheat In Warm Weather

As reported by ExtremeTech, the brand new 5G network is running into a major snag with mobile devices as Qualcomm 5G modems literally cannot handle the heat. After just a few minutes of use they’re going into thermal shutdown and falling back to measly 4G data rates. Reports by both PCMag and the Wall Street Journal (paywall) suggest that 5G-enabled phones consistently see problems when used in environments where temperatures hit or exceed 29.5 °C (85.1 °F).

The apparent cause is the increased power draw required by current 5G modems which make heavy use of beam forming and other advanced technologies to increase reception and perform processing on the received data. Unlike 4G and older technologies, 5G needs to have multiple antennas (three or more) to keep a signal, especially when you grab your shiny new smartphone with your millimeter-wave blocking hands.

The spin-off from all of this seems to be that perhaps 5G technology isn’t ready for prime-time, or that perhaps our phones need to have bigger batteries and liquid cooling to keep the 5G modem in it happy. Anyone up for modding a liquid cooling loop and (tiny) radiator into their phone?

52 thoughts on “5G Power Usage Is Making Phones Overheat In Warm Weather

    1. you know that LTE is 4g right? and there is a lot of substandards such as LTE CAT M1 that is kind of a replacement for GSM (low power, long range, low data rate). BTW LTE stands for Long Time Evolution, that hints that things has, and still develops over time. Also future releases of LTE Cat M1 will be classified as 5G, so LTE will be 4G and 5G ^_^

      1. Oh man I really wish they had thought a bit harder about the naming scheme. Outliers like the incredibly vague and useless “LTE” aside, naming a radio technology “x generation” doesn’t really say anything about it. Very obviously based in marketing, not engineering.

        I’m gonna bet 5G is going to be a hell of a mess. I mean it’s pretty blatantly a way for telecom companies to capture more regulation, write some more of their own laws, and sell more hardware under the mandate of some dumb race towards a badly-defined technology. The haste will make waste. And its branding is simply plastered with lame buzzwords like ai and iot and even things that probably won’t really come into existence for several decades like autonomous cars. Really sus.

        I bet they’re going to use the absolutely insane level of infrastructure that 5g requires to even further entrench natural monopolies in telecom, since no new player is ever going to be able to build their own 5g network that’s competitive. Just the whole thing reeks of the worst aspects of our awful telecom companies.

  1. i sense some extreme mix up of technology here. most operators want to use 5GNR in the 3.4-3.6GHz band, which is not mmWave. and it has great propagation properties – otherwise the same phone would not be able to get proper reception with regular sub-6GHz WiFi either.
    5GNR heavily relies on MIMO, has very good spectrum efficiency in terms of Mbps/MHz, but all this comes at a price: complicated modulation scheme that needs a lot of power.
    true mmWave (24-300GHz) is wastly different: it always requires LOS, and in many many cases is subject of high free-space path loss – parts of the WiGig v-Band (57-71GHz) is influenced by the siganl absorption of atmospheric oxygen (98% of the radiated power is lost here) – but here you have a lot more spectrum and you can use simpler modulation scheme with more bandwidth. but mmWave has quite some downsides: for example doppler frequency shift is noticeable even at walking speeds, so it is basically useless for true mobile usage – for now. and there were a handful of phones before the 5G hype train launched supporting 60GHz – which i wouldn’t refer to as 5G, cause it is NOT 5GNR – and it fid not set the world on fire.
    back to the 5GNR story: it really sucks for some things, and not just on the UE side. for the marketed throughput you need massive mimo with heavy and power hungry antennas on the basestation side, the marketed “sub 1ms” latency is only available on the air interface, so between the baseband and the UE – and since the whole IP datapath relies on the mobile core network components – esp in the NSA deployment model – it has to travel to the PGW and light ain’t that fast, and we can’t speed it up any further. UEs suck a tremendous amount of power and they gave dual radios: one for 5GNR (data only) and one for 2G/3G/LTE and they have to operate simultaneously if you want to make voice calls (mobile voice) because the ‘telco’ voice service relies on cell switched voice (2g/3g) or VoLTE – and 5G has no voice bearer capabilities right now – ie the battery will sweat, and thermal management is a real bitch.

    long story short: 5GNR is a really great thing, but we’re not there yet. it is overhyped, over-marketed, basically now useless thing that vendors and soon telcos will be dying to sell you with some absolutely ridiculous ‘use cases’ and overpriced mediocre implementation.

    the truth is that operators are just about to reclaim their investment into 4G, and 5G is a real money eating singularity now – so it is probably the next dot.com bubble.

    1. Thank you for the heaps useful info here.

      Seems like the marketing people once again imagined and promised more than could exist back here in the real world at present, where us engineers live. Just this time in a lovely spectacular fireworksey kind of way.

      1. As an engineer it is a bit annoying to see advertisement that is incorrect from technical viewpoint but “must” be used because of how consumers understand these functions/properties.

        Recently I discovered this kind of misinformation in phone/tablet specs when I needed to find devices that follow the USB specifications. Turned out that almost all manufacturers advertise devices the way consumers “understand” the functions (USB-C OTG for example which should be DRP). A small thing that becomes quite annoying when you realise that they are even manufacturing adapters that are unique to their devices and may not work with other products that are really build to the specification. *End of rant about how hard is it to find a device that is advertised as it is really build*

        Back to the subject, here in Finland the operators started marketing LTE as 4G even before ITU-R changed (lowered) the defining “requirements” for the 4th generation. Looking back now, unlike the USB example, this kind of clear separation from 3G was probably a good accomplishment from marketing because people really don’t need to know different standards used in mobile networks, but they should have an basic understanding of what they are paying for. Much simpler to remember that it’s okay to pay a bit more to get 4G service for added data speeds. Though they should really take a step back with all the 5G marketing, specially when they are actually selling phones and services with big “Get 5G now” -labels. (And yes checked, 1000Mbit/s, unlimited data transfer for 44,90€/month apparently from Saunalahti with really limited coverage in 5 largest cities in Finland)

    1. Until you try to use it for anything real. Also, 5g won’t change this. What is the point of 5g? so you can blow through your data cap in 10 minutes, instead of 30 minutes? ie: 5g is garbage until they get rid of caps…. oh, and stop it from literally ruining your phone’s SOC from burning it up. heat=shortened lifespan, every time. blah blah blah something about the first law of thermodynamics and idiot cell phone engineers being fucking retarded. Furthermore, the BEST way to increase bandwidth is to add more towers/access points. Something that the industry has not done for roughly a decade in any meaningful fashion.

      1. What is “anything real”? When my xDSL goes down (which is often…) I switch to hotspot on my phone and stream HD video, game, download applications, browse, anything I would normally do. Only limitation is 10GB before Verizon threatens to throttle, but they haven’t yet.

    1. higher speed is just part of it. lower latency and more rubustness is also a big factor.

      there is also a bunch of different standards for different applications that all are classefied as 5G. you do not use the same standard/modem for your 5G dishwasher or 5G weather station as you have on your 5G remote controlled truck.

      1. lower latency on the air interface _only_ – so a cutback from 15-20ish ms down to 1-2ms. but the rest is the same. i.e. the light from the basestation still will travel at the same speed it used to in the last 15bn years. that’s why the bullsh!t around 5G is toxic – most people totally believe the crap fed to them by the mainstream ‘tech media’ that right now we can circumvent the laws of physics. until you’re stuck with the traditional mobile core, things will not get faster… in many cases your packet needs to travel hundred-ish kms or even more to reach the gateway where the first IP routing decision can be made.

    2. Look at that z from the operator‘s side of the pipe and you realize that consumers are using more and more bandwidth – resulting in poor customer experience in dense areas. And yes – lower latency is a big winner as the current high and sometimes unpredictable latency in the upper 100s of milliseconds is what makes the browser on your phone appear not so snappy.

      1. Another way to say this is that the current setup is becoming more and more oversubscribed, so they want/need the consumer to pay for more, smaller, cells so that they can actually provide what you’ve already paid for. Because it’s them who doesn’t have enough bandwidth, effectively, as a current cell can only support so much in total.

    3. Traditional personal cell phone usage is NOT the primary driver for 5G. However, future phones will use 5G to eliminate the need for local data storage. You will be persuaded to put all your data on the cloud and access it via a high speed 5G link.

      The real benefits of 5G are aimed at government and corporate users. It provides the missing link for monitoring every aspect of human activity. For starters, imagine 24/7 high speed facial recognition and vehicle location. Now imagine that this vast repository of cloud data is instantly available to administrators and privileged users.

      Banking is another target industry. Digital currency cannot become dominant without the speed and throughput of a 5G network. The future of autonomous vehicles will also be gated by 5G communication between vehicles and to a high level monitoring and control system. etc., etc., etc.

      In the end, 5G has very little to do with your cell phone. And if you really want to see your 5G future, keep an eye on how China is using it.

    4. Capacity; LTE can provide acceptable speeds if there aren’t too many active devices in the same area. This is part of why you get congestion (slowdown) in densely populated areas. 5G can have 20x the bandwidth per channel and make use of it more efficiently and with lower latency.

      1. This. This is what many people forget. It isn’t so much about YOUR personal usage and specifically speeds but it is having the capacity to serve more devices simultaneously and also reduce chances of encountering congestion. With the future holding the potential for so many more devices coming online with mobile broadband access (self driving/autonomous vehicles, a bigger push for wireless home broadband if wireline providers stay on their current track as far as underserved areas, IoT devices, etc), it’s more pertinent that the aggregate bandwidth is there.

        Same reason why on the cable side there was the push for Docsis 3.x with the potential for up to 1Gbit/s download with some of the higher end modems. At least most of the relatively informed users understand the use case there.

        That said this is just 3G and LTE all over again. Marketing pushing it too hard too early. New chips are always going to be inefficient right out of the gate. Go back and look at the first devices supporting these technologies and you’ll see similar complaints.

        1. “This. This is what many people forget. It isn’t so much about YOUR personal usage and specifically speeds but it is having the capacity to serve more devices simultaneously and also reduce chances of encountering congestion.”

          Exactly: the 4G networks of major cities are *already* past capacity. Carriers have been essentially trying to paper over the problem by adding every bit of spectrum they can, but that’s basically over at this point.

  2. … you listening, Telstra?

    Maybe worry about getting 3G/4G working PROPERLY where it matters instead of wasting customers’ money on 5G networks that are available for only 3 blocks in the centre of capital cities. I consider 4GX a warning label rather than a feature, sounds like 5G should be treated with the same caution.

      1. Lets see, 2030, we are projected to hit the ‘tipping point’, with global warming, where it gets really, very fast. Lot of place should exceed 85 F, all the time. Guess 10 years is plenty of time to fix the problem, or come up with something newer, more expensive, and less reliable…

        Bummer for Florida, 5G would work most of the time, mostly at night. Makes no difference to me, I don’t use a cell phone, and plan to continue avoiding them. I just don’t like the idea of being put on a leash. Mostly, I see cell phones as toys, people play with them, more than actual communication. Even the communication, is stuff that could have waited. Basically, people use them for games, videos, and social media, the vast majority of the time. They are like a digital drug, people can’t set them down, even if it’ll cause them trouble. Texting while driving, is a primary offense in Florida, but should have been any phone use, while the car is in gear. Couldn’t guess at how many people step into traffic, even in front of trains, while zoned out on their phone screen.

    1. Oh yeah. 5g will have even less complete rollout in the US than 4g (I mean how could it possibly ever cover a large territory?) and 4g will remain inconsistent and forgotten even when these marketing cons are working on 6g and shoving it through public opinion and legislation as well.

      They should be required to offer 4g in a way that is more complete before they’re allowed to start working on zany new projects. These natural monopolies gotta be regulated very strictly (and ideally would be nationalized and made into government institutions). They aren’t providing the service people need to all, and their bad behavior and massive size makes competition unlikely to provide a solution.

  3. I live in a rich, western democracy where 4G is spotty at best outside of the CBD in major cities. I rarely have more than 1 bar of 4G (regardless of the phone flavour) in the middle to outer suburbs of any state capital. Despite the hype peddled by our 2 major carriers, I find the idea of a 5G rollout to any significant extent in the next decade laughable considering the abysmal 4G coverage by those same carriers nationally. The fact that the end user devices are apparently not yet practical confirms my suspician that 5G is currently a solution to a problem that very few people have or even care about.

    1. “I find the idea of a 5G rollout to any significant extent in the next decade laughable considering the abysmal 4G coverage by those same carriers nationally.”

      Except 5G is mainly needed *in* major cities. 4G has great coverage inside cities, but the effective bandwidth that each user sees is going down as the number of users keeps increasing – right now it actually presents as a factor of 2-3 drop in bandwidth during rush hour (hopefully by mass transit users!).

      “my suspician that 5G is currently a solution to a problem that very few people have or even care about.”

      Um, no. The 4G networks of almost all major cities are more congested than not across the globe: the “average” bandwidth is much closer to the minimum speed than the maximum speed.

      You won’t even need 5G outside of major cities.

  4. 5G is the new thing smartphone manufacturers and sellers hype up in order to sell people a smartphone because it has shiny tech instead of being sensibly designed, dependable and end-user serviceable.
    Too bad that 5G is only relevant for urbanites, while the rest is still fretting over the trend of stupid huge screens requiring two hands unless you’re Paul Bunyan, underdimensioned and yet not easily replaceable batteries because designers insist on a marketable “sleek” look, and in some cases also the mini-jack got cut for the same reason.
    And don’t give me the “but mini-jacks isn’t waterproof” because mil-spec waterproof versions exist, they’re just naturally incompatible with the anorexic inspired designers.

  5. I’m not a happy camper given that in the US providers want to gobble up C-Band to get the frequencies and expect Cable network providers to use “other means” to get their programming to cable head ends.

    Listen up, our industry built the infrastructure to distribute our programming AND the delivery of Terabytes of data to various places around the globe. Now because the “new shiny” has worn off you need an idea to get the public to purchase the latest $1000 toy. Not pleased. I know the FCC will cave in because these telecom companies offer the FCC bunch golden parachutes in the form of places on their board or executive positions.

    That the half baked 5G idea is getting any traction at all is insane. How about nudging a little more space near WiFi and putting up islands of higher speed at no cost or change to consumer technology?

    Oh, then you’re putting out money to improve your service instead of looking for a return on investment that’s meaningless. Got it.

      1. Take a look at FTA c band. We’ve come a long way since Videocipher. There are even 4k feeds. Anyway, most networks still use c Band as the primary method of distributing their feeds to local tv stations., cable providers, and small dsih satellite services because it’s still the most reliable and efficient way to do so. Even some webcasts actually originate from a satellite feed. And C band satellite technology is an effective way to link feeds from anywhere, even in areas where cell towers are scarce and fiber optic cables aren’t available. No matter where you are, as long as you have a view of the southern sky, you can uplink a c band satellite feed.

        Also, some people say that 5G will cause problems with weather forecasting as well.https://youtu.be/K2rY7JkwL2k

  6. What has happened before, will happen again…

    It took years for UMTS implementations to mature such that they didn’t burn users hands.

    LTE wasn’t so bad, but early LTE implementations sucked in terms of power management too.

  7. If 5G is wrecking havoc on phones as well as modems, what can be expected of wireless routers connected to the modems. And in mentioning modems & routers, what are the effects of any currently on 4G networks. Lastly, what about a computer directly connected to the (wireless) router?

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.