Framework 13 7840U Not charging

Battery no longer charges. Laptop turns on when plugged in but cat /sys/class/power_supply/BAT1/status shows “not charging”. I tried multiple chargers and USB-C ports, still would not charge.

Possible diagnostic info:

When plugged into 100W (Minix) charger the LED blinks alternatively in white and amber with a sort of pattern, two slow, two fast. Was unable to find a way to decode this, is there any way to find out what various blinks mean? This page was not useful as it is both unclear and didn’t include the code I’ve seen myself.
When plugged into a portable power bank (SWIT Omni 99S) the light remains consistent amber as it normally does when charging.

I tried running some battery diagnostics. Output of inxi -B is

Battery: ID-1: BAT1 charge: 0.5 Wh (0.9%) condition: 54.3/60.6 Wh (89.6%) volts: 14.69 min: 15.48

When I ran inxi -B again about 1h later (as it was plugged into the SWIT), I noticed the change in charge from 0.9% to 1%.

Here’s the output of framework_tool --power -vv

Charger Status
AC is: connected
Charger Voltage: 17800mV
Charger Current: 80mA
0.02C
Chg Input Current:2700mA
Battery SoC: 1%
Battery Status
AC is: connected
Battery is: connected
Battery LFCC: 3510 mAh (Last Full Charge Capacity)
Battery Capacity: 35 mAh
0.514 Wh
Charge level: 0%
Manufacturer: NVT
Model Number: FRANGWA
Serial Number: 042E
Battery Type: LION
Present Voltage: 14.691 V
Present Rate: 0 mA
Design Capacity: 3915 mAh
60.604 Wh
Design Voltage: 15.480 V
Cycle Count: 499
Battery charging
Battery level CRITICAL!

Things I tried:

  • different chargers
  • different USB-C ports
  • plugging it into the USB-C port directly without an expansion card
  • unplugging the battery and plugging it back again

Maybe I should mention that it’s been very hot this summer (30 celsius in my office) and I’ve had some work to do that required my laptop to stay plugged in for hours and think real hard which got it to overheat a bit. That being said, this was 1-2 weeks ago and I haven’t abused it much since.

Looking at the framework_tool output the capacity being shown as 35mAh is really sus… I only had this machine for less than 2y I can’t believe I killed the battery already…

Thank you for any suggestions!

What’s the output of framework_tool --smartbattery and ectool chargestate show?

Turned it on to run the diagnostic but sadly it wouldn’t stay on anymore. The power button led turned on but went back off after about 10 seconds while the charging led went chaotically blue red and green. Any other ideas?

I’ll leave it for a bit and try again.

Something is wrong with the battery or, possibly, the connections to it.
If I’m interpreting correctly the on charge battery voltage is shown as 17.8V which is too high, as in dangerously overcharged, so it would be unsafe to attempt to charge it. While the off charge voltage is shown as the much lower 14.691V which is the expected voltage at somewhere around 50% charged.
I guess it’s oscillating back and forth between finding the open circuit voltage is low so it needs charging and finding that once it starts charging the voltage is way too high potentially indicating dangerous overcharging.

1 Like

This is very SUS, should be 3915 or 2740mA, not 80. The EC or BMS is blocking charging

Not really, 17.8V is the correct charging voltage and it won’t overcharge the battery. The actual voltage in the battery would slowly rise from 14.691V to 17.7-ish V, not instantly.

For wherever reason when you press the power button the USB-C will momentarily shut off, this makes the computer to drain battery for 1 second. If the battery is only 1% doing so will drain it, so it’s better to only switch on when the battery is above 5%. Since you did manage to get the battery from 0.9% to 1%. I recommend using a 100W charger and plug your computer in for like 20-50 hours. If available, use a USB meter to monitor the charging wattage (still leave the computer off), it should slowly rise from 1.xW to about 40W when the battery is above 5%

Your battery has extended warranty so if troubleshooting fail you can replace it for free

1 Like

I imagine it has something to do with this:
Battery level CRITICAL!

Theres a lot of broken things about the readout.

AC is: connected
Charger Voltage: 17800mV
Charger Current: 80mA 0.02C
Chg Input Current:2700mA

Your USB-C charger is getting your laptop 17.8V (20V with some voltage drop), at 2.7A. The battery charger is doing 80mA to the battery.

Battery LFCC: 3510 mAh (Last Full Charge Capacity)
This sound reasonable. remember framework use a 4S, so voltage is much higher.
Battery Capacity: 35 mAh 0.514 Wh
If the battery itself think it only have 35mAh of capacity, thats wrong.
Present Voltage: 14.691 V
This is very low for a 4S pack. If we assume 3.7V per cell, 4x 3.7 is 14.8. Framework uses liHV I think, so its actually 3.85 per. Which 4x 3.85 = 15.4, close to the Design Voltage: 15.480 V

try a different charger. it might be struggling to negotiate USB PD.

something is wrong with the charging. it probably think it is full (white), and then not. (amber).
If it cannot negotiate, it will just alternate between off (no charger negotiated) and amber (charging).

I say. perform a CMOS reset. which usually just involve removing all sources of power (and the coin cell), and wait a bit.
I cant find a guide, idk. I did find this outdated Framework 13 11th gen intel CMOS reset, which is still mostly relevant because they dont update the mainboard designs.


TL;DR I think mate your battery cooked

1 Like

Too early for the conclusion. If framework_tool --smartbattery shows the battery cells are balanced (low voltage spread) the battery is far from cooked. 3.85V is the nominal voltage, which is the unloaded voltage when the battery is about 50% charged. 14.691V is a bit low but not “cooked” low. For the battery to be “cooked” the voltage “should” be lower than 12V. So I suspect the EC or the battery management is bugged, causing the charger to charge at a pathetically low 80mA. And since the laptop’s main rail circuit topology is narrow voltage direct charging (NVDC), peak load (such as booting up) is provided by the battery’s burst power. An 80mA of current from the charger will cause the battery to drain during bootup, which may ultimately “cook” the battery.

I can only recommend plugging the laptop to a powerful USB-C charger and leave it there for 20 hours so the battery should be at least charged to 5% before even trying to boot and typing the terminal commands

1 Like

A lot of modern battery BMC will lock you out below 3.6V per cell. I found some lenovo / Dell packs that lock out after a single “low voltage event”. And those are traditional chemistry, not LiHV (3.85V per cell) like Framework, newer Lenovo, or DJI.

Its for “safety”, but. its for guarantees. They dont mess around.

1 Like

Ah, planned obsolescence. Modern hardware users worst nightmare. For DJI, there are well documented methods to unlock the BMC. Not sure about Lenovos

I wouldnt draw such a conclusion. If you want 0 out of your 2 million batteries to blow up, you cap a pretty tame limit. Plus, battery is not exactly expensive. Lenovo/Dell sell replacements for like, $60.

I just rip them out and use the otherwise functional cells in my projects. It is what it is.

Now, none of this is Framework. So. I dont know. But given his descriptions, I say its a BMC lockout.

A single low voltage event(if you call 3.6V is low) is not going to blow up a battery. Stop safewashing(used in a similar way to “whitewashing” or “greenwashing” XD)

2 Likes

oh yeah it mostly wouldnt. and. fun fact. Old Dell packs from like, 2002. wouldnt lock you out. it will trickle. battery was expensive back then, hundreds of dollars. I guess at some point we draw a line.

I mean, what am I saying. this is the west. You scrap GPUs with broken PCIe fingers. We literally desolder GPU dies and put them on new PCBs. Mobile ones, too.

3 Likes

I did a test by playing a video game on battery

$ sudo framework_tool --smartbattery

Device Name:   FRANGWAT01
Manuf Name:    NVT
Serial Num:    0035
Manuf Date:    2023-08-21
Chemistry:     LION
FW Version:    Device=0x0045 FW=09.03 Build=0x4900
Mode:          0x6001
Temperature:   47.1C
Voltage:       13.523V
  Cell 1:      3.371V
  Cell 2:      3.381V
  Cell 3:      3.385V
  Cell 4:      3.387V
Current:       -3.773A (avg -3.728A)
Charge:        27% (870 / 3337 mAh)
Design:        3915mAh @ 15.480V
Bat Status:    0x00C0 [DSG (Discharging), INIT (Initialization)]
Cycle Count:   346

=== Battery Health Analysis ===

Status: HEALTHY
  No issues detected. Battery is operating normally.

Summary:
  Cycle count: 346
  Capacity: 3337 / 3915 mAh (85%)
  Current cell balance: 16mV spread

Then I plugged in USB-C it charged, so I guess lower than 3.6V isn’t going to “lock out” the BMC.

However if OP’s battery is unbalanced (i.e. one cell 2.5V another cell 3.7V) then locking out is justified and the battery should be replaced.

1 Like

first, thank you all so much for the ideas and responses. I’m in Europe so I woke up to this and I am very grateful.

For wherever reason when you press the power button the USB-C will momentarily shut off, this makes the computer to drain battery

that is quite unpleasant given how before people used their laptops as desktops once their batteries died. but maybe it’s a limitation of USB-C charging. would it turn on if I disconnected the battery?

There is no CMOS battery in the 7040 series, so I went for disconnecting the battery for 2m. (in my previous attempt I didn’t do it long enough to count as mainboard reset I think).
Now the LED is back to being amber when plugged into the 100W power supply, and I shall leave it to charge for some time.

I tried to turn it on after about an hour, just to see if it could, clearly it is charging a little as now it made it as far as stage 1 boot and showing the list of NixOS generations before shutting back down. So I shall leave it for some hours as you all recommend and come back with the smart battery diagnostic as soon as I can.

Thank you again to everyone who replied!

1 Like

After getting a reliable solid amber LED, do not rush to boot up the computer. The booting up process takes about 36W and it can easily drain the little energy put into your battery by charging at 80mA for an hour within seconds.

The charger has a significant steady-state error, the measured charging current is always below than requested current. According to my previous tests, if the requested current is 3915mA the actual current is about 3894mA. If the requested current is 783mA the measured current is about 759mA. If you do the linear regression 80mA will leads to 55.3mA. The “net gain” current is probably even lower because the same quiescent current of your EC drains a higher percentage when the input current is very low.

If you keep charging nonstop for 20 hours, the battery should be charged to at least 25%. However

If one hour means 0.1% then 20 hours is only 2%. Lets just hope the charger (or the smartbattery) snaps out of 80mA and charge normally.

If you have a USB power meter or a power socket power meter, you can observe the input power. If it increase to at least 30W, you can switch on your computer after waiting for another 5 minutes instead of 20 hours.

1 Like

Then why are those two voltages exactly the same on my fully working Framework 13 (AMD 7040 series) ?


PS C:\> framework_tool --power -vv
Battery Cutoff:     Unknown
Charger Status
  AC is:            connected
  Charger Voltage:  16344mV
  Charger Current:  0mA
                    0.00C
  Chg Input Current:2924mA
  Battery SoC:      71%
Battery Status
  AC is:            connected
  Battery is:       connected
  Battery LFCC:     3971 mAh (Last Full Charge Capacity)
  Battery Capacity: 2781 mAh
                    45.430 Wh
  Charge level:     70%
  Manufacturer:     NVT
  Model Number:     FRANGWA
  Serial Number:    014F
  Battery Type:     LION
  Present Voltage:  16.336 V
  Present Rate:     0 mA
  Design Capacity:  3915 mAh
                    60.604 Wh
  Design Voltage:   15.480 V
  Cycle Count:      76
PS C:\>

When the battery is not charging, the charger voltage reduced to “one notch” above battery’s voltage

Hey, have you created a support ticket? If so, please DM me the ticket number so I can bring this to the team now to look at

1 Like

Just as an update. I left it for 20h on my trusted 100W charger. Tried to turn on but it was exactly the same, only lasted until the list of NixOS generations. I am starting to think that the 0.1Wh is more likely. To test further, I plugged into another power bank which has power delivery and a screen: the screen says

Speed: 001mAh/min …

Any other advice? Anything I can check or change maybe in the BIOS? Can limtations be placed on voltages as a result of overheating as a safety measure?

I will be travelling for 2 days. I will leave it charge for 48h. However given i am in Europe, if this is looking more and more likely as it requires battery replacement (anything else?) I would like to know sooner rather than later so I can order it.

any advice on running without a battery? I need to work, I would like to make it boot

Now that is a cursed unit