Results of G4 temperature experiments

[sswitzer]

OK -- I've done several (very non-scientific) experiments with the phone and I've come to these two conclusions:

1. With or without WIFI, regardless of location method, moderate phone usage (as opposed to sleep time) increases temperature by between 5C (9F) and 8C (14F).
2. Ambient temperature (prolonged exposure to a nice cool room of around 20C or to a stinking hot car of around 30C) has a nearly direct correlation to phone temperature changes.

During my experimentation, I played a non-graphics game that has Google ads and I also used Google Now to look at stuff. I did between 5 and 10 minutes of such phone usage in each of these circumstances, and the phone started off with a temperature of around 30C.

(Temperatures are approximate averages... not scientific averages by any means, and are given in Celsius. Celsius to Fahrenheit: C x 1.8 + 32.)

• Apps used in Hot car, LTE, location from networks/LTE: 46C
• Apps used in Hot car, LTE, location from GPS: 46C
• Apps used in Cool room, LTE, location from GPS: 36C
• Apps used in Cool room, LTE, location from networks/LTE: 36C
• Apps used in Cool room, WIFI, location from networks/LTE: 36C
• NO Apps used in Cool room, WIFI -- killed all running apps -- back down to 30C.

So my conclusion is that phone temperature when idling is around 5C higher than ambient temperature, and when doing moderate phone usage (using apps that are definitely using Google services) the phone goes to around 15C higher than ambient temperature.

As to why the temperature seems to rise so much, as compared to other phones (I'm comparing to my old Note 2 and my wife's iPhone 5c), I think that it's a combination of the high screen resolution, the makeup of the glass, and the performance.

The limitations to my experimentation are:

• I only used 4 different apps during this experiment: Fives (my fav game), Google Now, Boat Browser, and the GSam Battery monitor app
• I only tested each circumstance for between 5 and 10 minutes.
• I'm gauging room temperatures without a thermometer... just guessing by how the temperature "feels".

 

[Mikey47]

What are you using to measure the temp of the phone? GSam? If so, you do realize this is just the battery temperature, right?

Anything that uses more battery will cause the battery to be warmer. Your conclusion should really just be using apps causes the battery to be warmer.

Everyone has been reporting higher temps around the back buttons (presumably where the radios are located). You really should monitor the temp there with an infrared thermometer.

 

[sswitzer]

What are you using to measure the temp of the phone? GSam? If so, you do realize this is just the battery temperature, right?

Anything that uses more battery will cause the battery to be warmer. Your conclusion should really just be using apps causes the battery to be warmer.

Everyone has been reporting higher Temps around the back buttons (presumably where the radios are located). You really should monitor the top there with an infrared thermometer.

Yes, just GSam, and yes I realize that it's just measuring batter temperature... but I could also tell by the feel of the phone (both the back and the screen) that the phone was pretty well at that temperature. (Is there another easy way to measure the temperature of the phone... that doesn't involve scientific instruments?)

And yes, I knew that apps would cause the battery to get warmer, but wanted to get a feel for how much of a difference really simple app usage (not intensive gaming or video watching) would have on the phone. I also wanted to firmly conclude that it had nothing to do with wifi vs. lte, and that the more extreme temperatures (in excess of 45C... pretty darned hot) were only felt when I was in a hot ambient temperature environment.

The post was meant to answer concerns that people have had about the phone temperature; if you have not been concerned about it, then please ignore my humble attempts at experimentation.

 

[RedOctobyr]

(Is there another easy way to measure the temperature of the phone... that doesn't involve scientific instruments?)

Yes. CPU-Z (which is free) will show the readings of a bunch of different temperature sensors in the phone.

I think the infrared thermometer idea isn't bad, if you have one of those, and if you want to see which regions on the back of the phone get to what temperatures. But that's honestly not as useful as getting common/universal readings that everyone can observe for free with just software.

Do note, however: use CPU-Z in C, not in F. It has a temperature-conversion error in the Thermal tab, if you're displaying in F. The F readings are not equivalent to the C readings, and seem too-high by about 30F. I am ASSUMING that the C readings are correct, as they made more sense than what it was showing in F.

I appreciate your doing the testing, and sharing the results. This helps everyone learn.

One challenge I found with trying to do some temperature testing is that it's a bit tricky to put a controlled or consistent load on the phone and the CPU. If something is still running in the background unexpectedly, for instance, that can skew the results. For future testing, if you have the ability to measure the ambient temperature of the room/car, that would be good info to know.

 

[sswitzer]

Yes. CPU-Z (which is free) will show the readings of a bunch of different temperature sensors in the phone.

I think the infrared thermometer idea isn't bad, if you have one of those, and if you want to see which regions on the back of the phone get to what temperatures. But that's honestly not as useful as getting common/universal readings that everyone can observe for free with just software.

Do note, however: use CPU-Z in C, not in F. It has a temperature-conversion error in the Thermal tab, if you're displaying in F. The F readings are not equivalent to the C readings, and seem too-high by about 30F. I am ASSUMING that the C readings are correct, as they made more sense than what it was showing in F.

I appreciate your doing the testing, and sharing the results. This helps everyone learn.

One challenge I found with trying to do some temperature testing is that it's a bit tricky to put a controlled or consistent load on the phone and the CPU. If something is still running in the background unexpectedly, for instance, that can skew the results. For future testing, if you have the ability to measure the ambient temperature of the room/car, that would be good info to know.

I'll try it all out again using CPU-Z... if I'm feeling particularly ambitious about it. Frankly, it was a lot of fuss just doing the very non-scientific experiments, and I'm not sure if I want to do it all again a 2nd time.

As far as CPU load goes, I started off with a few "lingering" apps that I'd run and not closed, and then I used the same 4 apps, doing around the same amount of stuff, in all my experiments... but it was certainly not a precise endeavour.

I think what I was most surprised at is that the temperature went up 8C or more by just doing low- to moderate- app usage. We're not talking Youtube or even much data transfer.

I was also pleased to find out that it had nothing to do with what type of location tracking was being used or whether or not the phone was connected to WIFI.

 

[RedOctobyr]

Do any of the apps you were using rely heavily on location info? Like the way that navigation software would?

I suppose that enabling a certain type of location reporting, then using apps that don't require frequent location checks, may not show the full difference between different location methods.

 

[sswitzer]

Do any of the apps you were using rely heavily on location info? Like the way that navigation software would?

I suppose that enabling a certain type of location reporting, then using apps that don't require frequent location checks, may not show the full difference between different location methods.

I was using Google Now -- obviously uses location info, but not nearly as much as map navigation -- and a simple game, Fives, that uses Google ads that are targeted by location. I chose these two because they use typical low-usage of location. If I was going to do the experiments again I'd try apps with location services disabled/enabled.

 

[sswitzer]

I'll try it all out again using CPU-Z... if I'm feeling particularly ambitious about it. Frankly, it was a lot of fuss just doing the very non-scientific experiments, and I'm not sure if I want to do it all again a 2nd time.

As far as CPU load goes, I started off with a few "lingering" apps that I'd run and not closed, and then I used the same 4 apps, doing around the same amount of stuff, in all my experiments... but it was certainly not a precise endeavour.

I think what I was most surprised at is that the temperature went up 8C or more by just doing low- to moderate- app usage. We're not talking Youtube or even much data transfer.

I was also pleased to find out that it had nothing to do with what type of location tracking was being used or whether or not the phone was connected to WIFI.

CPU-z is very interesting... And it's showing the battery temperature 3c higher than gsam.

I can guess what battery and tsens are... but what are the others?

 

[dpham00]

The note 4 runs much cooler, and performs close to the g4.the only thing about the note 4 is it likes to kick apps out of memory quick. Note 4 has the same resolution, and way more sensitive touch screen than the g4

So either lg used a more power hungry display or a more power hungry soc, or both. Given the lower battery life in the s6(proportional to the battery size) and that it was on 14nm unlike the s805 and s808,I would say more likely a soc issue.

Sent from my Verizon LG G4

 

[sswitzer]

The note 4 runs much cooler, and performs close to the g4.the only thing about the note 4 is it likes to kick apps out of memory quick. Note 4 has the same resolution, and way more sensitive touch screen than the g4

So either lg used a more power hungry display or a more power hungry soc, or both. Given the lower battery life in the s6(proportional to the battery size) and that it was on 14nm unlike the s805 and s808,I would say more likely a soc issue.

Sent from my Verizon LG G4

I have found that doing a "clear all" of the apps drops the temperature down immediately; I kinda knew that it was a matter of over-processing (my own non-techy term), but at the same time I'm impressed by how fast the darned thing keeps running no matter what.

I guess I was doing the experiments more to prove to myself that the heating is "normal", can be predicted, and is something that you can expect -- it doesn't mean the phone needs to be returned, and it doesn't mean that you need to turn off a whole bunch of stuff, i.e., it's just the way the phone is and if you can't live with it then you need to go to another phone.

One thing I will do tomorrow, however, is try turning off location services; my logic (if that helps) is that if someone is overly concerned about the heat, they can turn off location services when it's the ambient temperature is hot. Again, I'm just guessing that this will cool things down a bit, and I may be out to lunch.

I'm also looking for recommendations of how to reduce heating issues (aside from what I've already experimented with.)

 

[dpham00]

Some heat can definitely be normal. It runs warmer than the note 4 and s6 though, in my usage. Clearing all is basically dumping all tasks out of memory, so yeah it would run cooler but then there would be a slight lag opening apps like on the s6 after the apps are kicked out of memory.

For stuff like location services which are always running in the background, you can charge it to over 90% then leave the screen off for several hours or overnight until the battery drops below 90% and run wakelock Detector.

Sent from my Verizon LG G4

 

[sswitzer]

Some heat can definitely be normal. It runs warmer than the note 4 and s6 though, in my usage. Clearing all is basically dumping all tasks out of memory, so yeah it would run cooler but then there would be a slight lag opening apps like on the s6 after the apps are kicked out of memory.

For stuff like location services which are always running in the background, you can charge it to over 90% then leave the screen off for several hours or overnight until the battery drops below 90% and run wakelock Detector.

Sent from my Verizon LG G4

So how much warmer do you think the g4 is compared to the note4 and s6?

I'm also thinking that case style or nakedness will play a part in phone temperature... thoughts?

 

[dpham00]

So how much warmer do you think the g4 is compared to the note4 and s6?

I'm also thinking that case style or nakedness will play a part in phone temperature... thoughts?

Note 4 ran way cooler. S6 ran a little cooler.


Case does have an affect for sure but I use similar types of cases on all 3.

Sent from my Verizon LG G4