huh??
2 Likes
pretty sure ptm units are shipping now. it’s just they ALSO have liquid metal units in stock, which might take a while to run out completely, since some are on less active skus.
How can i get this readout?
Not so good with linux.
That program is S-tui
Sorry about the confusion, I was using an FL13 hence the lower power/frequency. I was trying to explain “the acpitz 3 seems not to be core 3” not trying to compare my laptop’s thermal
6 weeks after RMA. 11800 CB23.
Life has been busy and hectic lately so I haven’t followed thoroughly the manner. But got the FW newsletter stating they’ll send PTM to people asking. I’m not trusting myself to do the swap but I’d guess FW will create a step by step guide?
I never closed my ticket with support, just sent them the result of my most recent CB23. Will keep you updated on the response I get.
Hope you guys are doing well. Wish you health and happiness.
2 Likes
the acpitz values are various spots of the mainboard’s sensors. framework has an lm-sensors config in one of their repositories for that, but i found discrepancies in my testing on a fw16 about which value is which, i commented on this pull request months ago, no response yet
the fw13 values may or may not be right? i can’t check those, i don’t have an amd board for the fw13
the thing about the replacement thermal pad. I think a full RMA is better than having customer do this on their own. liquid metal is very dangerous to handle around electronics. if people are not doing it properly it will guarantee kill the machine. I already repasted with ptm7950 but I highly recommend people who are thinking about the thermal pad to reconsider an RMA with new motherboard instead.
Well the Phase Change Liquid Metal Framework used is pretty save to work with, as its hardened when not heated.
I’d agree with that. Compared to normal LM repastes, the hard pad fragments were pretty simple.
Also, not sure if it was the amazon seller and their pad sizing I used, but I found if doing a PTM sandwich, a 0.5 copper shim works too.
Probably less mounting pressure than the 0.8 but my temps have been steady and lower on average compared to the original.
I can lock the temp limit to 65 even and still pull 50W power, fans barely ramp at this config, I’m pretty happy with that result
1 Like
the thing is that you need to heat up the system to remove the heatsink. and when you heat up and remove the heatsink, the chance of it spill it to the motherboard still exist
Nope absolutly not. I didn’t even heat the System prior but i agree its easier. But the second you turn of the System its not even 10 seconds the LM cristalizes again.
Pull 50w steady at 65C or just spikes? If so the Rest of your Heatsink is a golden sample. My soldered Shim + PTM sits at about 71C when gaming with 30w TDP. I choose 0,8mm as i planned to lapp them to even them out. And to have about 0,1mm thicker than stock to raise the Mounting Pressure. PTM needs ample amount of pressure to expand its full potential.
1 Like
I ran r23 (getting to 100c) for 10 minutes
Powered off the laptop
Flipped the laptop
Let cool for 5 minutes
After this - it was solid, and easy to replace… no more difficult than any other time I have done a repaste with other materials.
Hardest part was removing the heatsink because when the LM is solid it kind of binds it to the TIM… but… patience and it lifted off.
50W steady at least running heavily modded skyrim.
I can run a prime 95 set when I get home to see what a full load would settle at over time.
I took an hour or so lapping the heatsink itself with a nail finishing brick, since there was a huge glob of solder left stuck from the original shim. By the end, my heatsink surface was almost a mirror, so maybe that helped in my case haha.
1 Like
Wow, i think i have to buy my 4th Heatsink to recreate your Results. I would love to hear the maximum Peak TDP you get with your Setup and what the Temps during a max performance CB Run are. Yours are so much better than my soldered Shim. I need moa powa! 
Don’t get too hot and bothered yet lol.
It could very easily be my skyrim modding mess causing weird power reporting.
I will note this is using a FM 180W power supply too. It sounds like if using a 240W weird results happen from time to time?
Or could be an extremely lucky heatsink as you noted previous.
I’ll report back with a few CPU only tests to confirm either way
1 Like
My FW16 Ryzen 7 7840HS arrived on Monday.
It should be using the new PTM9758 as I placed the order exactly a week ago today.
I’ve been doing some Cinebench R23 runs today and I’ve attached my results below. Hopefully this information will be helpful to others!
My OS is Windows 11 Pro and did the test with High Performance mode on to see what I could squeeze out of the machine.
Also using 2×16GB Crucial DDR5-5600 RAM and only the iGPU (couldn’t justify buying the dGPU for my needs!)
My laptop was plugged in throughout the test. I’m not using the Framework charger, instead using a UGreen 140W charger
First picture is a cold run. For this I got a Cinebench score of 15806 and the power was pretty steady between 51W-54W. (The timer for HWInfo was reset a couple of seconds after starting the test)
Then I let my pc cool down for about 10 mins and performed a second run. This time I got a score of 15713. The power mainly stayed between 50W and 52W.
(The timer for HWInfo was reset a couple of mins before the start of the test so the averages aren’t as high as the first run).
Are there any other tests I can do that may be helpful to others?
4 Likes
The 240w Charger Issue only affects the dGPU as of my Testing. Would be great to hesr back with comparstive Results.
@JBD thank you for the Result, they look promising, even better than the LM Models from Day One. Can you keep us updated over Time?
1 Like
So some weird/interesting results:
When limited to 65C, from cold start the system is actually able to stay this temperature without fan use… causing P95 from a cold start, limited at 65C, to never ramp fans. This results in the CPU always running at about 20-30W max
To get around this unexpected situation, I ran Heaven benchmark in the background, limited to 72fps. The GPU heat kicked the fans on, and at the 65*C limit, the system stayed locked to the temp, but CPU usage jumped to 60-65W. It held steady at this usage, since the temps didn’t budge.
I then tried turning the limit to 75*C, again with P95 only first
This increased usage to around 30-35W, but again, the system didn’t try ramping fans very much, it was basically self limited because it wasn’t triggering fans.
I then turned on P95 + Heaven. Now with the fans actually running, the system held steady at 75*C with an average usage bouncing around 70-80W on the CPU.
I did a last run using the UTU built in stress test for good measure, this was with the 65*C limit again, but results were similar to P95 above. CPU averaging around 60-65W.
So… conclusion… PTM is some great stuff? Maybe I have a very good heatsink sample? Maybe a 0.5mm copper shim is working better than the 0.8mm somehow?
I don’t think I have enough info to really say what at this time, but I’ll take it lol