Mensagem original de: yassan ,
'''Answer on iFixit''' [br] Jay writes: [quote]Is the AMD chip the real problem anyway? Could I have a different problem?[/quote] I think you got it right here. I got the MacBook Pro on which I am now writing from a friend. He got it from the university where he works, and some time ago the machine started having glitches like you describe, but with only its own display in use, and also with very light graphics load, so it could not be the discrete GPU. Furthermore, my friend noticed that the problem occurred mainly while moving the machine or when awakening it from the sleep state. After some inquiry I found [post|458833]. (the editor doesn’t get the permalink to the answer of santiagodelatorre, but that is what I’m pointing at). It seemed very logical to me that the problem was the RAM slots. But for the university the only solution was to change the motherboard; (un)luckily the MacBook was too old and not worth that expense, so my friend, under my warm suggestion, just kept it. And, if I would have managed to repair it, I could have had it for me, as a compensation for being his IT man for free (anyway he had already bought for himself a well beefed up 2017 15” MBP :) Surprise surprise, when I tried the trick at the link above, it didn’t work. I inquired further, and the RAM remained the only suspect. I was already thinking of an home made reflow when I found [https://ifixit.org/blog/6882/why-i-drilled-holes-in-my-macbook-pro-and-put-it-in-the-oven/|this], so I went for it, trying to heat up with a hot air gun only the soldering of the RAM slots, shielding the surrounding with little wood pieces. The first two times I was too mild, and it didn’t work. The third time I was too hot: the plastic of the slots became soft and deformed a bit (I should have taken a photo… :-), but one of the two slots came back to life. Since the laptop had 2x 4GB RAM, I bought a 1x 8 GB module. I also drilled a few holes in the bottom lid, a bit like in the in the link above, but larger and less in number. Furthermore, with some duct tape I fixed two pieces of a vacuum cleaner bag on the inside of lid, in front of the holes, so to keep the internals of the machine as dust-free as possible. (I have to admit that I am a bit scared that the duct tape gives up and the “filters” jam the fans, but to put it outside would have been aesthetically worse and would have required more holes to have the same air flow. Last but not least; since the original rubber feet where a bit worn out, I made new thicker ones, out of soft grippy foamy material, to lift the bottom a bit more from the resting surface. And, when I use it on soft surfaces, I place it on a wooden panel, which I already used with the same purpose under an iBook G4. [image|1516836] The machine is working flawlessly since about one month. Even with a room temperature of about 30°C, it stays below 50°C (at the left fan sensor) with normal tasks (web browsing, video watching, audio editing…). When I put under load (mainly exporting the mixdown of some long audio recording), I preventively spin the fans to 3500 rpm with smcFanControl; in this way it usually stays below 60°C. Should it fail again, I will try to reflow the RAM connectors with a 200 W soldering gun, coming very close to the motherboard but without touching it, so to have a more local heating. I also kept one of the 4 GB modules; if I would manage to revive both of the slots, I could have a 12 GB RAM setup :-) Just to give the whole thing a frame. Before this MBP, I had a 2010 white MacBook, bought used on eBay. It is a great machine, but the display is a bit small for the audio editing work. Now it is one of the two common laptops of the commune where I live, but with the clause that I can take it back if the MBP would fail. And yet, I didn’t go for the reflow of the RAM slots soldering until no other possibility was left, knowing that sometimes it works and sometimes it is the last nail on the coffin. So, Jay, I would understand you if you wouldn’t follow this road until there is another chance :-) If I remember correctly, at some point in the question of the first link, there is also an hypothesis on why the problem seems to be related to the sleep/startup phase, and maybe also to Chrome, which is a memory hog and may warm up the ram modules and surrounding. Hope this helps someone.