Fix (reduce) FP2 overheating

[Patrick1]

According to this topic I just recently replaced the pink thermal pad as my six year old FP2
overheated more and more often just by normal operation like browsing. It was that worse already, that charging while using it often wasn’t possible for when the core temperature exceeds 42°C charging is automatically interrupted by the charge controller. So the phone had to be left aside to cool down a bit so the charge process would continue. This behaviour got more uncomfortable for me.

As these overbred pocket computers do dissipate a lot of heat just as conventional computers plus by their compact SoC design and often having to struggle with being kept in close garment pockets the thermal regulation is very demanding.
So in mobile devices there is not much space for clunky heatsinks let alone fans.
In case of the FP2s design the best magic is relying on the thermal pad between the main chip and any type of heatsink (heatspreader) available. In this case the silver cover with the black foil glued ontop as the pictures 7+8 show.
Aging of the thermal pad will reduce its efficiency (-> high thermal resistance) causing the chip to faster reach higher temperatures and according to the temperature limit at 42°C for charging the device the process will be interrupted. I have not experience crashes or reboots because of this with my modell (one of the first 17.418 and first SoC revision), but others obviously have. But lagging occoured.
The relevant parameter for a good heat transfer is the ratio between W(atts)/mk(elvin), higher is better but also more expensive (non-linear unfortunately). Also the individual thickness needed to sufficiently fill a gap.
I think for the FP2 case a thickness of 1mm is suitable.

Many pads offer values between 1-10W/mk which may be ok for giving thermal support to a ssd chipdrive or not so powerfull circuits operating in open environments or bigger compartments than typical mobiles. But for the latter these values seem not to provide enough support to keep the operation flawlessly going.
I at first tried thermal pads I had left from external IcyBox M2 enclosures as I assumed everything should bring an improvement compared to the old spend pad of six years age and use. It did improve the thermal shift slightly, but not that much as from the beginning of the new device. (Btw. I have no idea about the parameters of the provided pads for these enclosures. They are mainly shipped along to fill a gap, but ssds also have an internal temperature management throttling down their operation speed should temperatures go to high.)
Still I had the necessity to place a better thermal pad to catch the high power dissipation in this use case. After a long search I surprisingly came accross very strong pads with an astonishing ratio of 17W/mk. (I could also find some with 20W/mk though)
I’ve choosen a type of [Alphacool] (Suchergebnis für kühlpad | Alphacool - the cooling company). Yes, that’s their price tag for overaverage components.
FP2 needs an about 1,5cm² small sized pad. I couldn’t find one manufacturer offering such tiny pads. Usually there are stripes or much larger quares on sale which makes it all expensive for just needing such a tiny fraktion (almost waste size)

Also there is fraud going on with these components as no one later can actually visualize or easily measure if that’s the type of pad he initially have ordered. There are plenty of grey, blue, pink types between 0,5 - 1,5 mm thickness in the market, so how to determine the correctness of its thermal ratio?

I think all that can be done is trusting/hoping that anyone selling fractions of these original pads from manufacturers are serious resellers that stick to their word. Sure, it turns out being a quite profitable business too.

At last what can/should be done is a “burn-in” test using an app e.g. cpu throttling test and the usual charger to see how temperature values change while the handset is under heavy load for some time. If it’s operating (much) better for a long time and there are no crashes and the charging is going on for a long period of time (which implies not overheating too quick) operation under regular conditions should be ok again.

I have put my trust into this reseller, but of course there are plenty more. I couldn’t find any better pad at my favorite electronics dealer Reichelt. And as results look up to now I think the pad I have received fulfills what was stated and operation is ok so far.

I believe since 2015 remaining FP2s in use these days often suffer from these aging troubles. Proper handling for replacement should take 30-60 mins for those brave wanting to take it on.

[Martin_Anderseck]

Thanks for the research you’ve done. Still I believe that if your phone suffers from so much heat, there may be some software keeping the phone way more busy than what it should be. My phone tends to get extremely hot when using navigation apps, Teams, when watching videos and that kind of stuff. The fact that I can almost burn my hands at the phone then indicates that the heat dissipation is more or less okay, doesn’t it? Interestingly enough my phone restarts completely independent from its temperature, if it restarts at all.
And all in all: Even if the phone is able to transport the heat from its processor to this tiny piece of metal, where should it go then if all that stuff is encapsulated in plastic? I don’t want to say it doesn’t matter if the heat remains at the core. Cool components are supposed to live longer. However I have the feeling that the heat dissipation is not the biggest issue of the FP2.

@Patrick1

It was that worse already, that charging while using it often wasn’t possible for when the core temperature exceeds 42°C charging is automatically interrupted by the charge controller. So the phone had to be left aside to cool down a bit so the charge process would continue.

Could that behavior be relaxed by what you’ve done? My phone just didn’t get as warm in the beginning as it gets nowadays when charging and I believe this is pure software.

[pmn]

The aforementioned apps are computationally intensive - hence dissipating more waste heat.

I believe numerous factors lead to this result:

1. Some (mainstream) apps require more compute resources than a couple of years ago
2. Thermal pads age
3. Devices accumulate garbage data over the years (e.g. logs) that increase resource use

[Patrick1]

@Martin_Anderseck
Yes, navigating as such always caused a thermal problem right from the start, I knew I couldn’t get this extreme case reliably cooled.
Yes, almost burn my hands was the case in rare situations so imagine how damn hot the internals must have been. Surely not life extending for any component.
The heat must be spread more not only stay at one spot, that’s all that can be/is being done.
Background processes need to be reduced too, but running Google Android…
The plastic at last does heat up too spreading some of the heat to the surrounding area (which may be ones pocket)
Yes I agree, comparing my FP2 with the FP4 there is quite a difference in so many ways not only heat dissipation. (It’s hard to get my FP4 to sweat even with a cpu stresstest)

Using a decent thermal pad at least shouldn’t make things worse. But I also agree, software does have
an impact. There wasn’t much of a change of use in the near past. I got the apps I needed and at some point there wasn’t much change anymore. I’m taking software load of the FP2 to fully shift to my FP4 with /e/ OS. I’m keeping an eye on my cache and usually force stop apps I surely use just once in a while (in weeks or months) by occasion.
To reduce traffic I have a simple firewall to block unwanted traffic limited, but Google does know how to get past if it wants. I think it’s causing most of the load and traffic. I never had anything installed for emailing or FB or Wha…crap. The System monitor lite app usually list at least the Google Play Store or Google Play Service tasks jumping around among the top five. Once in a while I do surprisingly find another new service installed like “Earthquake alerts”, don’t know when this got on my device.

I did notice however shortly as battery saver kicked in how all of a sudden smooth and responsive swiping and stuff went. So what’s all going on in the background over time has reached a noticeable level.

@pmn
Yes, clearing the entire cache once in a while may reduce crashes and reduce load a bit. Automatic app updates in the background are comfortable, but one never know when and how troubles are being introduced. At some point users realize that the device or app act strange, operate unstable or crashes occur. Finding the culprit is hard then. I perform manual updates per app, clear the cache and maybe once per week reboot.
“1.” good point here, I can see how app .apk archive filesize has grown over time when scrolling through my backup folder.