Unrealistic charging temperatures?

Hi all,

I had my first brush with the blinking red light, while I had the L5 charging in the car with the sun on it. I came home and found this statement in the docs:

The charge controller will pause charging when the battery temperature exceeds a threshold that is slightly above 40°C. Use the Usage utility, selecting the Thermal tab, to check the temperature reported for the sensor named bq25890-charger. If the reported temperature is too high then if the ambient temperature is high, try moving the charger and phone to a cooler location.
https://docs.puri.sm/Librem_5/Troubleshooting/LEDs.html

So… I’m planning to go to the Mojave desert this summer. Daytime temperatures will be ~110°F (43°C). Does this mean I’m only going to be able to charge at night? Why is the max charging temperature set so low?

1 Like

Out of curiosity, what do you think is more reasonable?

I have gotten the blink on my sofa with an ambient temperature under 70F, while reading the forum with FF-ESR.

I work in the Mojave Desert, but I don’t expect it to be much of a problem, as I don’t plan to charge outdoors. I charge off and on at my desk, and all the time in the car. Both the office and the car (1.4 hours, each way) are cooled, and my car phone mount is literally in the vent.

I have had the same problem in Cuba. The first mistake was already carriing the L5 in a backpack wherevit gets hot by its own. Afterwards it was impossible to charge it.

Most Li-ion cells aren’t safe to charge above 40-45°C.

3 Likes

Then we do not know something… Although not exactly a dessert (and not far from it :grinning:), when in Athens Greece in the summer I cannot charge L5 because the ambient day temperatures are typically above 33, often 35 and can easily climb to 40 C. So for summer in Athens I use another phone. So the question is:

Why those other phones (android, kaios (nokia)) do not complain about these temperatures. Should we assume they do not care about our safety or their battery safety?

If yes then this is good to know.

They probably limit the charge current (making it take longer) when the temperature is approaching the safe limit, making it produce less heat due to the charging process itself. We currently don’t, we let it charge at full speed until the thermal threshold is crossed, then pause charging until it cools down.

I’m not exactly sure which approach makes you end up with a fully charged battery faster, although one of them makes the user quite obviously aware of this thermal throttling while the other does not :wink:

5 Likes

One important thing to note is that Librem 5 can produce quite a bit of heat while in active use (especially with high CPU/GPU usage) and this heat is spreading to the battery as well - therefore, putting the phone into suspend or turning it off while charging will certainly help when the ambient temperature is so high that this heat can’t be dissipated efficiently enough.

The main way this heat dissipates is through the phone’s metal body and screen, and I have noticed that even the tiniest airflow directed towards the phone helps tremendously with keeping it cool.

4 Likes

OK, I see. Thank you.

Just a note which might be useful for @bear454 I have noticed that you can charge the phone during daytime in hot areas when it is shutdown and placed on (the non-working) home heating radiators. In such temperatures the radiators still produce an upwards air flow that will keep L5 cool to complete the charging.

For the record, the charging current can be limited by the user manually:

echo 1000000 | sudo tee /sys/class/power_supply/bq25890-charger/constant_charge_current

This will set the maximum charging current to 1A instead of the default 1.5A. The kernel won’t allow you to set a value that’s higher than the one used by default, you can only make it lower.

(in fact, looking at the battery’s datasheet we should be able to safely bump the maximum to about 2.2A to make it charge faster, but this would of course produce even more heat during charging)

One thing to note is that this was a pretty recent kernel driver feature and I don’t remember whether it’s already in the 6.1 kernel that’s used by default now or whether it was only added in 6.2, so if it doesn’t work for you it likely will once the kernel is updated :wink:

5 Likes

Very nice to know that you can lower the charging current.

By the way since we discuss about such issues, when the phone is plugged for charging does it still draw current from the battery? Or it draws energy from the mains?

It only draws current from the battery when the current provided via USB-C is not high enough to power it (it takes the remainder from the battery).

1 Like

Ah, which means that if the phone is idle and is forgotten on the charger the battery does not really suffer from charging and discharging simultaneously. Right?

Yup, IIRC the only component that’s connected directly to the battery aside of the battery gauge is RTC, so it would be a matter of days or even weeks to drain it enough to start charging again (which you’d notice because the red LED would shine again).

Unless you fiddled with the charger’s registers manually to disable the indicator, you can be sure that if the red LED isn’t shining, the battery isn’t being charged (well, unless your phone is from the Birch batch, then it doesn’t shine when the phone is off and charging :slight_smile: )

2 Likes

This is one of the big benefits to suspend. Once the dynamic DRAM allocation is in place and Kasts fixes the no audio during suspend I’ll be using it all the time, the phone stays quite cool even when in your pocket.

1 Like

I haven‘t tested autosuspend for longer time. I did it now and the L5 didn‘t came up on a call.

it takes about 3 rings for it to wake up so that is the down side to using it…needs to be sped up somehow.

I let it ring 10-15 times.

Oh ok…yeah I think it’s still very experimental so I’m sure it’s not very reliable at the moment.

And it could depend on the ambient temperature.

Sure, 40°C is nothing here too but there is a difference between

  • the meteorological reported temperature (outside, in the shade, under ideal conditions e.g. using a Stevenson screen)
  • temperature inside without air-conditioning
  • temperature inside with air-conditioning

Wandering around outside in direct sunlight the temperature could be much hotter still but then you wouldn’t normally be charging your phone when wandering around outside (unless using a battery bank?).

Fortunately in the last few years here the summers have been relatively mild and so I haven’t had to test this issue in practice.

Maybe Purism needs to sell a fridge pack for safe (condensation free) charging in the fridge. :slight_smile:

1 Like