(yet another) framework 16 won't power on

Community Support
,
1

Framework Laptop 16 (Batch 14) AMD Ryzen 9 7940HS with Radeon RX 7700S. Has been very lightly used for about a year. Most recently unplugged for a few months while I was away traveling. It started up fine with wall power; the battery was drained but it actually appeared to be charging OK. I left it on overnight (with lid open) to download a big file. The next morning, the screen was dark, lights were off, and it wouldn’t turn on no matter what I did.

I did open a support ticket, but I read reports of long delays in reply so I thought I’d ask for advice here in parallel.

I first re-plugged it in to wall power, the indicator LED initially was red/amber, and after several hours it became steady white (suggesting good charging). Meanwhile I was trying many things to troubleshoot. Reading the community forums and reddit I gleaned ideas including the following:

-Resetting the mainboard (holding the chassis intrusion switch down for 2 sec x 10 times, with the case open and the charger plugged in.

-Swapping/removing/rotating RAM sticks

-Removing all expansion cards except the one I was charging with (and that one I swapped left-right)

-Removing the battery, holding down power button for >30 seconds.

During this process I would occasionally get the following sequence of LED lights:

W-R-G-R-G-G-G-R-R-G-R-G-G-R-G-G-G-G-G-G-G-G

(Red lights correspond to: Battery connect check, Power Good VCCIN_AUX, Audio board Detect, Thermal sensor detected, CPU reached S0 state, if you go by the FW 13 chart)

However it would come on several minutes (2:04 on one of my videos) after a power-on attempt, never immediately after pressing the power button.

Sometimes, a power-on attempt would result in the white indicator light changing momentarily to red then blue once, or sometimes flicker blue first then red, before going back to white. Then for a while, unplugged, all LEDs are off and pressing power shows zero signs of life. When subsequently plugged in, the indicator LED is white, and pressing power seems to impact nothing whatsoever (no change in color). I then removed the keyboard and disconnected the mid-plate ribbon press-connector again, and the power button once again elicits a R-B or B-R color flash (seems random which order it occurs in, but it’s always just one flash of each color).

Differential diagnosis: Considering the battery - the charging light went from red to white after an appropriate amount of time which is somewhat reassuring. The red flashing light during POST corresponds to a bad battery connect check however. Yet, behavior is different if I fully disconnect the battery before power-on attempt (couldn’t get a POST with battery removed). I know lithium batteries are susceptible to damage from deep discharging, but I wouldn’t expect a failure to look like this– I dunno? I don’t think it’s RAM- I tried swapping/removing. I don’t think it’s interference from an expansion card- I removed all I could. I considered an ECU malfunction, but I hoped that resetting the board (many times) would clear it.

That leaves me with few theories besides a mainboard failure. I’m hoping the community has a few more reasonable tricks or tips to suggest, alternate theories, or interpretations that strongly suggest a specific diagnosis.

Thanks in advance!

2

It sounds to me that you have a failed battery.

But, to fully test try:

  1. power off the laptop.

  2. unplug the psu for 2 minutes. ( this will reset the EC)

  3. see if the laptop will power on. First try without psu connected, then try with psu connected.

  4. if not 3, open the laptop and physically remove everything except the ram. I.e. remove Battery, ssd, wifi card. Then try power on with psu connected and see if anything at all displays. (The display will not work unless ram is present)

  5. get an EC CCD and use it to get info from the battery without needing to power on the laptop.

2 Likes
3

Thanks for these suggestions!

Steps 1-3, no improvement in situation (I did that several times yesterday, and did it again this morning)

Step 4, I will work on.

Step 5- Can you clarify what an EC CCD (What’s “CCD” in this case?) is and/or which one is compatible with the FW 16 battery? I can build a custom tool if there’s a schematic (but I suck at designing my own circuits).

Thank you again- I appreciate the community support greatly

M

4

The EC CCD is this:

It is “Closed Case Debugging” card.

You plug it into slot 4 of the FW16 and then plug a usb cable into another laptop, and you then can view the console command line of the Embedded Controller Chip.

Essentially, you can do all the ectool commands while the FW16 is powered off and failing to boot.

But, hopefully doing (4) will find out what is faulty because any of those being wrong can prevent the laptop booting.

1 Like
5

James,

I did step 4. With PSU connected but SSD, wifi, and battery disconnected, there were no signs of life. With battery connected, it then resumed the original faulty behavior. Does that suggest anything to you?

Thank you for the CCD explanation and the link to Howett’s work. I’ll study the material and reach out to the project maintainer. I’d love to have UART access to the EC, one way or another, regardless of the outcome in this case. In the meantime, hopefully my support ticket will yield further info.

I’ll keep watching this thread in the meantime and update it once I have more info. Thank you James, and to anyone else who is reading this and considering offering suggestions!

M

6

Update: Framework customer support engaged me quickly, and have been very professional and helpful. They advised the following for diagnostics:

  1. Disconnect all expansion cards, charger, RAM, SSD, input module, touchpad module, spacers, Mid Plate, GPU module, and main battery. Press and hold the power button for 20 seconds, then leave them disconnected for 15 minutes to allow the mainboard to power drain. Reassemble and attempt to boot. (It did not)
  2. “Try booting up the laptop without all the expansion cards while charging directly into the expansion ports and observe for any changes. This is for us to isolate possible expansion card issues.” (no changes)
  3. To reset the mainboard: Power off the laptop and remove mid plate kit. Plug the laptop into AC power. Press the chassis open switch in the center of the mainboard 10 times. (“press it slowly, so press for 2 seconds, release, and wait for the red blink on the mainboard LEDs and then press again.”)
    You should see flashing red LEDs when the chassis open switch is not pressed, otherwise you don’t have power to the mainboard. The Mainboard reset will NOT work if the mainboard does not have power. (no changes)
  4. “Remove both the SSD and the Wi-Fi module, then attempt to power on the system. This helps rule out peripheral-level faults that could interfere with POST or power sequencing.”
  5. “Disconnect the main battery entirely and try powering the laptop using only the Framework USB-C power adapter. This will help determine if the battery or its connection is contributing to the issue.” (causes constant red-blue flashing only. The familiar LED POST sequence flashes ~1 minute after pressing power button)

Observations:

On wall power with battery disconnected, LED continuously flashes red-blue. Pressing power button causes POST sequence to flash after 1 minute.

With wall power disconnected, attaching battery causes two red flashes. Subsequently pressing power also causes two red flashes. There is another single red flash after 30 seconds, and then 30 seconds later the POST sequence flashes.

In this configuration (wall power disconnected, battery attached), depressing the chassis intrusion switch changes behavior: Instead of 2 red flashes, I get a single: red-blue flash, or a blue-red flash, or just a blue flash, or just a red flash (seems unpredictable which one I get. POST flash presents after 1 minute.

It proved helpful to take notes after every test and be patient. I only noticed some LED behavior after setting a stopwatch and watching the board for at least a full minute. Taking a video was also helpful. Slowing down the video to 25% speed helped reliably check the POST sequence, because it only takes 8-10 seconds for all 21 flashes and my brain is not that fast.

Other news: I have been able to source an “EC Tool 2” expansion card for UART-over-USB closed case debugging (CCD). Support advised me that this capability is not yet documented in the FW16 (it’s designed for the FW13 mainboard) so I will hold off on applying it. But these custom expansion cards are in rare supply and if anyone needs one in the future, please reach out to me. Hopefully I can safely put it to good use, eventually.

I want to emphasize that the support folks have been excellent, responsive, informative, prompt, and patient!

Malcolm

2 Likes
7

Hi @Malcolm_S,

Sorry to hear about the issues with the lightly used Framework 16. From your description I would suspect a possibly bad battery (one of the cells might be bad). Though from the christmas tree of colors on the POST something else has to be afoot. Unfortunately the Framework Laptop 16 does not like standalone mode if I remember right which would be the condition to power on sans the battery altogether.

The EC might have some stuck/corrupted values it is reading too and this is possibly preventing it from getting up to a running state.

You are definately on the right track to leave literally everything else disconnected except the RAM. (SSD, Wifi, Expansion cards, Expansion slot video card) to get the best chance of it coming to life.

It would be nice if the engineering team had some addtional documentation to share on the red/blue codes the boards sometimes are showing. These are likely state codes that correspond to other subsystems on the board. Some of this might be proprietary to the vendor whose components they are though which makes it tougher. Just in the off chance, have you tried another power source? I believe even a 30W PD power supply is enough to initiate a POST. I know on the Framework Laptop 13 my little Google 30W PD charger will bring it up just fine. The Framework Laptop 16 is a little hungrier though, it should have provisions for detecting what voltage and current is available. Just another thought to help rule out unknowns.

Ideally a mainboard reset should reset the EC to a known working state; sorry to hear of your struggles. One last ditch thought would be to take the mainboard out of the case and just look over it for anything that might have “cooked” itself. If there is physical damage like a circuit jettisoned its black smoke then likely the only repair would be a new mainboard.

As a last hint, if you are out of warranty (over 12 months) and you purchased this with your credit card; you might check and see if they have an automatic extension of factory warranties. This would be done through your credit card company; this is worth a shot so you are not out the cost of a new mainboard if indeed you are outside the warranty and Framework elects not to cover it.

2 Likes
8

Hi @pkunk ,

Thank you for your words of support and fresh perspective on possible causes. I agree, it seems likely to either be a battery, EC problem, or a mainboard power (control or sensing) problem. I have isolated the mainboard from virtually every connected piece of hardware (except the graphics module) without influencing the problem.

I would like to learn the battery pinout and the expected voltages between pins to confirm nominal voltages on my multimeter. This doesn’t seem to be published anywhere I can readily find.

When I receive the EC Tool 2 CCD device, that will be very helpful in understanding the EC state.

I did infrared-image the board this evening. No obvious hot spots.

~m

9

Update: I provided some UART diagnostics to customer service and they are going to RMA the board back to their repair facility for me. This is a good outcome and I feel relieved. I can’t think of another brand that would have worked with me as diligently as FW did, and I want to recognize their support team.

I encourage Framework to publish more EC and diagnostic code information as a next step for this product line.

1 Like
10

Got the laptop back, 100% working in a reasonable time frame. Highly satisfied with Framework’s service and outcome!

2 Likes
11

This topic was automatically closed 2 days after the last reply. New replies are no longer allowed.