KB: Phone not/barely charging from power supplies when keyboard is not full
03-18-2022, 05:19 PM
I really love the keyboard case so far, now my Pinephone truly is the perfect pocket-sized computer! This is what “phones” should be like! The only inconvenience I’m trying to solve right now is the charging priority/behaviour with a power supply plugged in. (Just so you know: The phone’s USB port is covered up, the current limit of keyboard battery -> phone is set, the userspace driver is installed.)
Previously, I only owned a 5V 1A power supply. It charged the KB+PP fine (apparently at around 1.2A) and didn’t get too hot. The problem/inconvenience is that the priority of the keyboard charging chip seems to be keyboard battery > phone > phone battery. This means that the phone doesn’t receive any power from the power supply while the keyboard is charging, and only receives power once the keyboard is (almost) full, resulting in the phone battery percentage continuing to drop for the multiple hours it took my 1A supply to charge the keyboard battery. Since a 5V 3A PD supply is recommended, I thought that those might be enough to charge both batteries at the same time, which I believe the charging chip is supposed to be able to do.
To test this theory, which is mine, I asked around for 5V 3A supplies and C-C cables I could borrow. Apparently these are not common at all, at least among the people I know in real life. Two 5V 3A supplies I tested indeed resulted in both batteries charging and the phone’s percentage going up. According to /sys readings, the keyboard battery was getting 2.3A, the phone battery 0.8A and I assume the phone parts an additional 0.2A, which seems to be what it uses with the screen off and low workloads. However, shortly after plugging in one of these chargers, the keyboard charging chip started continuously shutting itself off and on again about once per second, as if I were spamming the side button. Since the current values seem to add up to more than 3A, I assumed that these power supplies were supplying too much power and the charging chip was shutting itself off as a safety measure. I noticed that neither of the supplies mentioned PD anywhere on them, which I guess means that PD is mandatory.
So I got this power supply and this C-C cable, which seem to charge the keyboard battery at around 2.4A, so twice as fast as the old supply which is great and also halves the time until the phone finally gets to receive power, but the phone still doesn’t get any power when the keyboard battery is not full yet. A last series of tests I made involved combinations of my new supply and cable with a USB C charging port in a car (that seems to support PD as it worked fine) and its C-C cable. These tests apparently showed that both power supplies have the same behaviour, but the cables differ. My cable charged with only 2.4A in the car as well, but the other person’s cable, in either supply, supplied the keyboard battery with 2.4A and additionally the phone battery with 0.4A, plus assumingly 0.2A for the phone itself (is there a way to read that value if it exists as I assume?). Which adds up to exactly 3A. I guess that means my cable for some reason only passes through 2.4A, even though it supposedly supports up to 60W (what cable would I have to get that works as intended?).
So 5V 3A PD supplies seem to work in general, but the keyboard still eats up 2.4A first with its highest priority, only leaving the phone with 0.6A. Which is enough to keep the battery percentage constant with the screen on at low loads, so that would at least mostly prevent the phone battery from completely discharging before it has reached its turn, as was the case previously. But of course it would be best if I could somehow influence how much the phone gets. The userspace driver seems to only allow toggling power supply input as a whole and keyboard battery -> phone charging, which is very useful but doesn’t help with this inconvenience. I also found /sys/class/power_supply/ip5xxx-charger/constant_charge_current (and _max) which is set to 2300000 (and 3100000). Would changing this value, say to 2000000 or 1800000, reduce the amount of power the keyboard battery takes and therefore pass through more to the phone? Or is there anything else I could do?
And as I said: This is only a minor inconvenience, especially with proper 5V 3A input (if I’m able to get a cable that works). I can live with it being like this. That’s a very small price to pay for this incredible device. Thanks to everyone involved, especially to Megi! But it would be nice to know if there is a way.
And sorry for this being so long, I tend to ramble on a lot. Maybe I shouldn’t have included so many details from my findings.
TLDR: Is there a way to change how much power the phone gets from a power supply while the keyboard battery is also charging and not full yet?
Previously, I only owned a 5V 1A power supply. It charged the KB+PP fine (apparently at around 1.2A) and didn’t get too hot. The problem/inconvenience is that the priority of the keyboard charging chip seems to be keyboard battery > phone > phone battery. This means that the phone doesn’t receive any power from the power supply while the keyboard is charging, and only receives power once the keyboard is (almost) full, resulting in the phone battery percentage continuing to drop for the multiple hours it took my 1A supply to charge the keyboard battery. Since a 5V 3A PD supply is recommended, I thought that those might be enough to charge both batteries at the same time, which I believe the charging chip is supposed to be able to do.
To test this theory, which is mine, I asked around for 5V 3A supplies and C-C cables I could borrow. Apparently these are not common at all, at least among the people I know in real life. Two 5V 3A supplies I tested indeed resulted in both batteries charging and the phone’s percentage going up. According to /sys readings, the keyboard battery was getting 2.3A, the phone battery 0.8A and I assume the phone parts an additional 0.2A, which seems to be what it uses with the screen off and low workloads. However, shortly after plugging in one of these chargers, the keyboard charging chip started continuously shutting itself off and on again about once per second, as if I were spamming the side button. Since the current values seem to add up to more than 3A, I assumed that these power supplies were supplying too much power and the charging chip was shutting itself off as a safety measure. I noticed that neither of the supplies mentioned PD anywhere on them, which I guess means that PD is mandatory.
So I got this power supply and this C-C cable, which seem to charge the keyboard battery at around 2.4A, so twice as fast as the old supply which is great and also halves the time until the phone finally gets to receive power, but the phone still doesn’t get any power when the keyboard battery is not full yet. A last series of tests I made involved combinations of my new supply and cable with a USB C charging port in a car (that seems to support PD as it worked fine) and its C-C cable. These tests apparently showed that both power supplies have the same behaviour, but the cables differ. My cable charged with only 2.4A in the car as well, but the other person’s cable, in either supply, supplied the keyboard battery with 2.4A and additionally the phone battery with 0.4A, plus assumingly 0.2A for the phone itself (is there a way to read that value if it exists as I assume?). Which adds up to exactly 3A. I guess that means my cable for some reason only passes through 2.4A, even though it supposedly supports up to 60W (what cable would I have to get that works as intended?).
So 5V 3A PD supplies seem to work in general, but the keyboard still eats up 2.4A first with its highest priority, only leaving the phone with 0.6A. Which is enough to keep the battery percentage constant with the screen on at low loads, so that would at least mostly prevent the phone battery from completely discharging before it has reached its turn, as was the case previously. But of course it would be best if I could somehow influence how much the phone gets. The userspace driver seems to only allow toggling power supply input as a whole and keyboard battery -> phone charging, which is very useful but doesn’t help with this inconvenience. I also found /sys/class/power_supply/ip5xxx-charger/constant_charge_current (and _max) which is set to 2300000 (and 3100000). Would changing this value, say to 2000000 or 1800000, reduce the amount of power the keyboard battery takes and therefore pass through more to the phone? Or is there anything else I could do?
And as I said: This is only a minor inconvenience, especially with proper 5V 3A input (if I’m able to get a cable that works). I can live with it being like this. That’s a very small price to pay for this incredible device. Thanks to everyone involved, especially to Megi! But it would be nice to know if there is a way.
And sorry for this being so long, I tend to ramble on a lot. Maybe I shouldn’t have included so many details from my findings.
TLDR: Is there a way to change how much power the phone gets from a power supply while the keyboard battery is also charging and not full yet?
