Lord.
1 Like
This goes way too far. It is also risky and running the laptop this way does constitute a serious risk for killing the board.
If I got asked this as a resolution step, this is where I would stop, go over to the Marketplace section of the forum and create a nice ad for my machine.
I hoped the Support process would be smoother, at least for known issues.
This might be a slight overstatement. As long as you aren’t getting it wet or sticking anything conductive in there (or removing parts) while it’s running, it should be fine. I know it’s a little different, because it wasn’t something support requested I do, but during my own testing while going through this, I did run the computer without any input modules or even the midplate installed. It was just sitting on a desk and I was using an external keyboard.
Obviously, this requires some extra care, and I’m not sure what support is hoping to learn, unless they are wanting to rule out uneven pressure due to a warped or out of spec midplate. So it’s a bit of a surprising request (it also assumes NEITG has external keyboard/mouse to use the computer). But I don’t see it as a major risk for anyone who is comfortable doing the kind of work required to pull the mainboard and such.
You might be right. I do find the prospect of probably having to replace my own motherboard a bit scary, but the fact that no electricity will be running though it is at least reassuring
1 Like
Framework does have a guide for swapping the mainboard. It’s not too bad if you follow the guide. But it’s also not something I’d have been comfortable with either, if I didn’t have a history of building and working on computers and other electronics. If you truly aren’t comfortable, you could explain that to support and see if they’d be willing to have you send it to their repair center.
1 Like
I do have some level of comfort with it, I had to do a complete disassembly and reassembly of my old Dell laptop. Thanks, traditional designs where everything is mounted on top of the input cover and the keyboard is riveted right in - very cool. It wasn’t the end of the world, but the laptop was old and cheap. Had it not booted again, I would have shrugged it off as “eh, whatever, it was time anyway”.
From the guide it doesn’t seem way too bad. I have already sent my laptop to the repair center and it has been a terrible experience: one month, no uninitiated communication, delayed replies, and the problem I sent it in for mostly unsolved. I think I’d much rather pick myself up, drink a calming drink before starting the repair, and swapping the board myself 
2 Likes
I’m like that with a lot of things. Not sure what I can do, but if I can do anything to help, let me know. Not that I’m an expert, but I’ve swapped my mainboard a few times now, lol. One tip is that you don’t have to disconnect the antenna wires from the Wi-Fi card. I just unscrewed and popped out the card, and then let it hang from the antenna wires while I swapped out the board.
1 Like
Ooh, heck yeah. That is nice to know. I am visually impaired and WLAN replacements are always the things that make me struggle the most. It can take a full half hour to connect those damn pins.
Thought the fan is ramping up more than a year ago and found this thread.
Seems like I’m affected too, although I still get 14500pts in CB23 with my 7840HS
1 Like
yep I get 88 on the coldest core and 100 on the hottest. Though Core 3 and 4 seem to be the “preferred” cores, seeing more activity.
And it does match the power usage
I seem to be having a similar range - between about 85 and 100. But I cannot hit 45 W package during Cinebench. Amusingly, I regularly hit or surpass 45W package during Overwatch 2 - probably boost.
After much back and forth with support providing more detailed videos around my particularly bad thermals/noise levels, they are sending me a replacement board and a shipping label to return my current one.
I hope they seriously dig into it and use it as a case study for what is going wrong with everyone’s boards and perhaps acknowledge this may be a more widespread issue than some non-power users may realize.
4 Likes
I wish they would do a “recall” on these boards. But my guess is recalls only happens when there are legal implications for a company otherwise too bad too sad you got a bad machine… unfortunately we may all be stuck with thse machines with the bad boards.
Yeah, recalls are often a calculation. Will it cost us more to do it or to not do it? And in this case, if they’ve even investigated this issue deeply, the math is probably on the side of not doing a recall if only a subset of users either have or at least notice an issue.
I’m not so concerned about a lack of a recall, but lack of a statement/acknowledgement of the issue. For all I know the replacement board they are sending me will exhibit the same issues, if not immediately then down the road. I want to know that they found the issue, have addressed it, and all future board shipments going out whether for new orders or RMAs will not have this issue.
6 Likes
I Asked about exactly this Topic during my Support Requests and they said there is only a Minor Percentage of People Requesting an RMA for the Thermals. It is highly unlikely that it will be aknowledged. Possibly with Pressure or Aknowledgement from Linus Sebastian as he is a Major Investor there will be a Step Forward, but this Topic has to be brought to him Personally somehow. I Try my best spreading about this Issue around Reddit X and Youtube. Thats one of the Reasons this Thread here is the one with the 4th most Replies on the FW16 Subforum, but only 7K Vews.
3 Likes
That’s a very valid point you bring up that there’s no guarantee if the replacement board doesn’t have the thermal issue. Please do reply back when you get the replacement board and you get the opportunity to test it in your machine. Thanks!
1 Like
The replacement boards they send out will degrade.
I am a batch 1 owner and on my third board (7940HS, no dGPU). I am currently going through the support process once again and they will probably ask me to send in my laptop only to agree there is an issue and replace the board one more time. It takes about 1-2 months for the liquid metal to degrade in my experience (depending on use of course). Since I have had 3 boards, and all have degraded in similar time spans, I believe this to be a factory defect with the cooling solution and the internal response is to keep replacing boards/heat sinks until our warranty expires and we have to fix it ourselves.
On my current board, the hottest core; #4, hits 100C within seconds of a benchmark, and will even get to 100C during games. My coldest core; #1, only reaches an average of 66-68C during benchmarks and similar during games. So a 34C difference between cores under load, it doesn’t get more obvious in my opinion that we have a thermal interface issue. I am only reaching 38W on the package power now because core #4 hits 100C so quickly. My first board dropped to 30W before I noticed and contacted support. My second board dropped to 33W, and that is when support wanted me to mail in the entire laptop to get it fixed. I estimate a few more weeks and this one will also be near 30W.
During my support interactions I have directly asked about the thermal interface degradation and cited this thread, but none of those questions are ever acknowledged. I have provided plenty of evidence, but they keep trying to “find” another cause. I believe they 100% know this is a widespread issue but won’t do anything because not enough owners are checking their thermals.
We need a large channel like LTT to do a “X months later” video and show the thermal issue to get a spotlight on this if we are to have any hope of a solution. The problem with LTT being an investor, he may already know about this and is not willing to harm his investment.
I will keep pushing this issue because a 2000+ USD laptop that is meant to be sustainable should not degrade after a few months. That goes against some of Framework’s core marketing.
6 Likes
Batch 13, original board, no mods or tinkering with the cooling.
Cinebench R23, score usually around 14200, to as low as around 12,200.
Core4 hits 100 almost instantly.
The rest take there time to get where they do.
Been testing for almost 2 months, similar values every time.
Thermal spread is a similar value for all cores, as shown in HWMonitor, in all tests.
Worst thermal spread, about 15 deg C.
Power has been as low as approx 30W, but generally around 35-40W.
Sometimes core4 only gets to about 90C and the other cores are upto 10C less.
Sometimes a restart of laptop helps, which has me wondering if software/firmware has an influence also.
Cheers
As you may have read.
As out of my Investigation it is 2. Errors Occuring.
- Liquid Metal Run off (Degradation over Time)
- Bonding of the Shim to the Vaporchamber (High Thermal Resistance Reducing the Overall Performance)
After i removed the Shim and replaced it with a full Coppershim and using PTM7950 as a TIM i am now Reliably above 16k Points in CBR23 and the Cores take Time to Heat up. No instant 100C anymore.
The TDP goes from above 60w to a sustained 54w staying at about 96C at the hottest Core during a Benchmark when i do not tinker with the TDP myself. When i use X86 Universal Tuning Utility’s Extreme Profile i am Riding the 100C from above 70w TDP down to a sustained 58w under full Load.
The System itself is capable, but those 2 Issues have to be solved.
3 Likes
Excellent results! I feel like this may merit you creating an iFixit guide for us other poor souls if this issue persists and goes unacknowledged…
3 Likes