I hope that this will help someone. This is my experience, with all the usual disclaimers like YMMV.
I have identified two repeatable scenarios that will crash my Droid Incredible.
The short:
My phone will crash and repeatedly reboot when the temperature reaches about 38-40 degrees C.
My phone (not just the browser) will crash when I use the browser to go to
511.org
The long:
This is how I arrived at these conclusions. A few observations first:
1) The phone rebooting itself is usually an indication that a fatal hardware error has
occurred. As in, some illegal area of memory was accessed, some instruction in the
CPU was invalid, some high priority interrupt from one of the components, etc.. The
phone reboots itself to protect from further catastrophe, because after any of these
things happen, all bets are off.
2) I have found that certain apps that run in the background, which incidentally start
without you wanting them to, and restart when you've manually killed them, and still
start even though you've told them you didn't want them to start up with the phone,
and oh, by the way CANNOT BE UNINSTALLED, will use enough of the CPU to increase
your baseline temperature closer to the failure point. Yes, I'm talking about you, SKYPE
MOBILE! After reading what I just typed, this might be a big DUH fact. Oh well, guess
I'm just venting.
3) I went through all the pain of factory reset and restore. Twice. No resolution whatsoever.
For the heat issue:
* I downloaded an app that would tell me the battery temperature, which for lack of being
able to read a temperature diode on the CPU itself, was a good enough indicator of chip
temperature. For this I chose "Temp + CPU v2" by SanelS because it was free and simple,
and also showed CPU frequency and busy percent. No plug here, just use whatever you
think is useful to you.
* I plugged the phone charger in, and proceeded to use several apps to warm up the phone.
I played Angry Birds, used the GPS, used Latitude and whatever else I could think of. I
switched back and forth between these apps and the home screen so I could see the
temperature rising. When the temperature got to about 37 degrees C, I knew I was close.
A few more birds out of the slingshot, and BOOM! Phone crash. Again and again until I
removed the battery and allowed the phone to cool.
* Next experiment was to get a block of Blue Ice, remove the back cover of the phone
with the battery still in, and place the phone directly on the Blue Ice so that the battery
was in contact. The charger was plugged in. I proceeded to do everything I did before
to get the CPU going and the battery/CPU/phone hot. The temp came down below 20
degrees C and stayed there. I was able to stay in operation for hours with no crashes.
* I then removed the phone from the Blue Ice and snapped the cover back on, repeating
the first experiment. 30 minutes later, BOOM again. Repeated phone crash.
This is as controlled of an experiment as I could come up with to prove that the issue
is heat. I have worked with integrated circuits for 20 years now, and my best guess as
to why some people see this heat issue is because there is a variation in the process by
which chips are created. Basically, all chips, especially CPUs, have limits as to how fast
they can run. If a chip runs at 1GHz, that means that one clock cycle is 1 nanosecond.
A signal must make its way from one place to the next within this nanosecond of time.
If it doesn't make it, this is an unrecoverable scenario in which a hard system reset occurs.
There are three contributing factors as to whether the signal will make it:
1) Voltage. A chip can run faster with higher voltages (but within a margin that would
not destroy the device). You overclockers are well aware of this. I'm pretty sure that
all the smartphones have voltage regulators that bring down the voltage from either
the battery or charger such that the voltage to the components in the phone will never
droop, so this should never be an issue.
2) Temperature. The hotter a chip runs, the harder it is for the internal signals to make
it to their destinations in time.
3) Process. There is a variation in how the silicon of each chip behaves. This has the
effect of some chips failing at lower voltages and/or higher temperatures than another
exact copy of the device. Each chip is supposed to be screened at the factory or
"guaranteed" by design that it will work with certain given operating temperatures and
voltages. Some chips end up being closer to the edge than others as to when they'll
fail (the distribution is a bell curve). Qualcomm, did you EFF up here with the operating
specs?
My wife and I each got Droid Incredibles at the same time. I installed the temperature
monitor on both devices and ran similar experiments. No problem with my wife's phone,
lucky her.
For the 511.org issue:
* Holding the temperature down again with the Blue Ice, simply browsing to 511.org
crashes the phone, although not repeatedly. The same failure occurs when the
temperature is not held down. I find this one more perplexing as I am not a software
guy. I would expect maybe a browser crash, but not the whole phone crashing. Anyone
want to chime in here with theories?
So what to do? I am frustrated with this and have spent countless hours Googling for
solutions that will not void my warranty. Of course, saying that, I'm going to return the
phone to Verizon for a replacement and play my luck. I'll run the same tests on the
new phone and return that one too if I see the same thing. I guess the point of all this
is that the heat factor is DEFINITELY a problem and it is in itself, I believe, a reason to
return the phone.
good luck to all of you, I hope that this post has armed you with more justifications to
get in Verizon's (and HTC's) faces about this issue. I envision an army of us with blocks
of blue ice walking into HTC headquarters with our phones and chargers.