I’ve had my battery limit set to 60% for over half a year now, and prior to the update to BIOS 3.07, this seemed to work fine. If the battery was below the threshold, the battery would charge, and if it was above the threshold, it would discharge to it.
I am seeing an issue where, with the same BIOS settings (Battery Extender off and a manual limit of 60%), my laptop does not charge or discharge the battery when below 60%. Currently, it is at 58% and “Not Charging.”
The problem is that with powerprofilesdaemon, the laptop acts as if it is not connected to AC power and sets power limits accordingly.
Has anyone else noticed this or can anyone else replicate this (even with a different OS)?
The charging has hyterisis. If you set the limit at 60, it will charge to 60, then discharge, and only start charge below 55. Then charge back up to 60.
The details are in the bios release notes.
You can use “ectool chargecontrol …” to change those limits and also force it to charge again.
Thanks. I read the release notes and saw the notes you mentioned regarding hysteresis. Should this affect package power limits, though? I’ve noticed that I now have a 40 W limit when connected to external power, which appears the same as the on-battery “performance” mode limit. Previously, I was able to achieve 60-65 W, so this is extremely disappointing
EDIT: Just ran a quick test, and the power limit is higher on battery. 44 W vs 40 W on AC
As an update, when using the 180 W charger, even when the battery isn’t charging, I am able to hit the full power limit. On my USB-C monitor (which I believe supports 65 W PD), however, I see the limit to 40 W package power. It seems there’s some new logic based on the power source capability that affects power limits
The power limits based on charger capacity and battery presence/charge level thing has been in all versions of the ec code I have seen so far, don’t think that is new.
This one seems new as the exact same scenario (60% with 65 W charging) went from allowing a maximum of 65 W (60 W sustained thermal limit) to a maximum of 40 W. This lower limit is below the limit when operating on battery alone (44 W) and seems to be balanced to prevent discharge. At least when using the “performance” power profile, I think discharge should be allowed.
I was able to confirm that BIOS version 3.07 does change the CPU package power limits under the conditions of my setup. I rolled back to 3.05 (which was a bit of a process…) and I immediately saw the missing performance return. Average CPU clock improved by ~800 MHz and it hit 60 W, up from 40 W.
Did you also have some issues during the rollback, like the TPM warning it would need to be reset?
Also, I opened a bug ticket. If you can, it might be helpful to leave a comment that you also experienced this. The lack of feedback from others was making me feel like either no one else had this issue, or that no one else cared…
My TPM reset, yeah. Since I mostly run linux that was a non-issue, but I believe I had to do some stuff to recover Windows such as manually enter my bitlocker and re-activate Windows using my product key.
I’ll do some due-diligence testing and then add my own comment to the github ticket.