Homepage
Forum
Drivers
HMM
MTM
Feed
Some related information about my system: I have a T420 with coreboot, a quad-core i7, and I'm running Linux.
Skip to so now the problem if you don't want to read me ramble about my project and how I ended up here.
I am having a very strange issue after re-celling an aftermarket 45N1007 battery with a BQ8030 with SANYO firmware. I bought the thing a year ago but I wanted to upgrade the capacity substantially to a terrifying 189 Wh with some 27100 Samsung E50 cells I bought. This will give my T420 (theoretically) up to 12 hours on battery if I'm willing to be conservative. I have yet to find out in reality, but it did survive playing Half-Life 2 for an hour with minimal drain.
I am using my VGA port to interface with my battery over i2c and wrote my own tools to dump the battery and read its status before realizing that someone else already did that exact thing: https://github.com/noolex/lenovo_battery_repair/. I took dumps of the EEPROM before re-celling the battery but there was a bug in my dumping tool that made me lose every 32nd byte (rookie mistake). I fixed this and now my dumps are 1 to 1 with the other tool, but critically I don't have the original dump.
During the re-cell the BMS must have set a "permanent failure flag" despite my best efforts, but it wasn't a huge issue as I was able to unlock it no problem using someone else's EEPROM dump since I'm lazy. There were oddly no fuses blown, but I shouldn't complain.
So now the problem
Well, when I plug the battery in to my laptop it will run on battery power, but when I open the xfce4 power manager I get this:
This screen should be filled with useful information like the battery technology, capacity, health, manufacturer, serial number, etc. like it was before I re-celled it. This was still problem before I ever touched the EEPROM, but after re-celling. I assumed this was because of the "permanent" failure flag, but it would appear not.
I get the same amount of information when listing it with upower as well.
So the BMS must be toast right?
Well, no... Using the tool I made (and lenovo-battery-repair gives the same result):
ignore the concerning temperature, it's actually in degrees Kelvin, not Celsius. also it didn't charge at 65.52 amps, it charged at 6.552 Amps, that's another bug because apparently I can't math correctly.
The BMS is reporting everything exactly as it should, even reporting that it charged a little bit! (Remaining Capacity was at 0 before, and I verified with a multi-meter that the voltage did indeed go up).
As you can see, I even left the capacity and everything at its default value just in case either the EC or BIOS was seeing something that was making it nope out.
So it must be a problem with my beloved T420 right?
Well this is the confusing part...
This is a picture of the same xfce4 power manager screen, but with my unmodified 6 cell LGC battery. As you can see, it works.
So at this point I assumed it must be an issue with the connections. Maybe jamming my wires into the battery's connector caused it to bend out of shape. Well, I checked with a multi-meter and the pins appear to have perfect continuity between the laptop and the battery. I even tried bending the pins inward to make the connection stronger, but it didn't have any effect. Furthermore, I'm fairly certain it is able to make some form of communication to the battery, since before unlocking the "permanent" failure flag it would also say "waiting for discharge" below the generic "Battery" text.
Does anyone know what's going on here? What could be the cause? Has anyone ever encountered this before? Is the EC somehow seeing through my lies and deceit and refusing to give me my percentages? At this point I've ordered another battery so that I can try again with a fresh BMS, but I'd like to save this one if possible. Even if I end up fixing this one before the new one arrives, it would be nice to have a spare "normal" 9-cell battery that doesn't look like a bomb.